Update "Further Documentation"

Remove stale pointers in the "Further Documentation" and just point to RocksDB Wiki
rocksdb_transaction_get_for_update now exports (#6293 )
2026-11-07 19:42:37 -05:00 · 2020-12-30 17:14:07 -08:00 · 2020-12-30 15:42:59 -08:00 · 2020-12-30 13:40:10 -08:00 · 2020-12-30 09:25:09 -08:00 · 2020-12-28 13:38:10 -08:00
319 changed files with 23468 additions and 5026 deletions
@@ -17,6 +17,13 @@ commands:
          command: |
            HOMEBREW_NO_AUTO_UPDATE=1 brew install pyenv

+  install-cmake-on-macos:
+    steps:
+      - run:
+          name: Install cmake on macos
+          command: |
+            HOMEBREW_NO_AUTO_UPDATE=1 brew install cmake
+
  increase-max-open-files-on-macos:
    steps:
      - run:
@@ -27,10 +34,14 @@ commands:
            sudo launchctl limit maxfiles 1048576

  pre-steps:
+    parameters:
+      python-version:
+        default: "3.5.9"
+        type: string
    steps:
      - checkout
-      - run: pyenv install --skip-existing 3.5.9
-      - run: pyenv global 3.5.9
+      - run: pyenv install --skip-existing <<parameters.python-version>>
+      - run: pyenv global <<parameters.python-version>>
      - run:
          name: Setup Environment Variables
          command: |
@@ -39,6 +50,11 @@ commands:
            echo "export SKIP_FORMAT_BUCK_CHECKS=1" >> $BASH_ENV
            echo "export PRINT_PARALLEL_OUTPUTS=1" >> $BASH_ENV

+  pre-steps-macos:
+      steps:
+        - pre-steps:
+            python-version: "3.6.0"
+
  post-steps:
    steps:
      - slack/status: *notify-on-master-failure
@@ -63,7 +79,7 @@ commands:
          name: Install gflags
          command: |
            sudo apt-get update -y && sudo apt-get install -y libgflags-dev
-  
+
  install-gflags-on-macos:
    steps:
      - run:
@@ -77,7 +93,7 @@ commands:
          name: Install gtest-parallel
          command: |
            git clone --single-branch --branch master --depth 1 https://github.com/google/gtest-parallel.git ~/gtest-parallel
-            echo "export PATH=$HOME/gtest-parallel:$PATH" >> $BASH_ENV
+            echo 'export PATH=$HOME/gtest-parallel:$PATH' >> $BASH_ENV

 executors:
  windows-2xlarge:
@@ -89,15 +105,27 @@ executors:
 jobs:
  build-macos:
    macos:
-      xcode: 11.3.0
+      xcode: 9.4.1
    steps:
      - increase-max-open-files-on-macos
      - install-pyenv-on-macos
-      - pre-steps
      - install-gflags-on-macos
+      - pre-steps-macos
      - run: ulimit -S -n 1048576 && OPT=-DCIRCLECI make V=1 J=32 -j32 check | .circleci/cat_ignore_eagain
      - post-steps

+  build-macos-cmake:
+    macos:
+      xcode: 9.4.1
+    steps:
+      - increase-max-open-files-on-macos
+      - install-pyenv-on-macos
+      - install-cmake-on-macos
+      - install-gflags-on-macos
+      - pre-steps-macos
+      - run: ulimit -S -n 1048576 && (mkdir build && cd build && cmake -DWITH_GFLAGS=0 .. && make V=1 -j32) | .circleci/cat_ignore_eagain
+      - post-steps
+
  build-linux:
    machine:
      image: ubuntu-1604:202007-01
@@ -141,13 +169,13 @@ jobs:
  build-linux-release:
    machine:
      image: ubuntu-1604:202007-01
-    resource_class: 2xlarge
+    resource_class: large
    steps:
      - checkout # check out the code in the project directory
-      - run: make V=1 -j32 release | .circleci/cat_ignore_eagain
+      - run: make V=1 -j8 release | .circleci/cat_ignore_eagain
      - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
      - install-gflags
-      - run: make V=1 -j32 release | .circleci/cat_ignore_eagain
+      - run: make V=1 -j8 release | .circleci/cat_ignore_eagain
      - run: ./db_stress --version # ensure with gflags
      - post-steps

@@ -158,11 +186,11 @@ jobs:
    steps:
      - checkout # check out the code in the project directory
      - run: make clean
-      - run: USE_RTTI=1 DEBUG_LEVEL=0 make V=1 -j16 static_lib tools db_bench | .circleci/cat_ignore_eagain
+      - run: USE_RTTI=1 DEBUG_LEVEL=0 make V=1 -j8 static_lib tools db_bench | .circleci/cat_ignore_eagain
      - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
      - run: sudo apt-get update -y && sudo apt-get install -y libgflags-dev
      - run: make clean
-      - run: USE_RTTI=1 DEBUG_LEVEL=0 make V=1 -j16 static_lib tools db_bench | .circleci/cat_ignore_eagain
+      - run: USE_RTTI=1 DEBUG_LEVEL=0 make V=1 -j8 static_lib tools db_bench | .circleci/cat_ignore_eagain
      - run: ./db_stress --version # ensure with gflags

  build-linux-lite:
@@ -181,21 +209,21 @@ jobs:
    resource_class: large
    steps:
      - checkout # check out the code in the project directory
-      - run: LITE=1 make V=1 -j32 release | .circleci/cat_ignore_eagain
+      - run: LITE=1 make V=1 -j8 release | .circleci/cat_ignore_eagain
      - run: if ./db_stress --version; then false; else true; fi # ensure without gflags
      - install-gflags
-      - run: LITE=1 make V=1 -j32 release | .circleci/cat_ignore_eagain
+      - run: LITE=1 make V=1 -j8 release | .circleci/cat_ignore_eagain
      - run: ./db_stress --version # ensure with gflags
      - post-steps

-  build-linux-clang-no-test:
+  build-linux-clang-no_test_run:
    machine:
      image: ubuntu-1604:202007-01
-    resource_class: 2xlarge
+    resource_class: xlarge
    steps:
      - checkout # check out the code in the project directory
      - run: sudo apt-get update -y && sudo apt-get install -y clang libgflags-dev
-      - run: CC=clang CXX=clang++ USE_CLANG=1 PORTABLE=1 make V=1 -j32 all | .circleci/cat_ignore_eagain
+      - run: CC=clang CXX=clang++ USE_CLANG=1 PORTABLE=1 make V=1 -j16 all | .circleci/cat_ignore_eagain
      - post-steps

  build-linux-clang10-asan:
@@ -243,33 +271,63 @@ jobs:
      - run: CC=clang-10 CXX=clang++-10 ROCKSDB_DISABLE_ALIGNED_NEW=1 CLANG_ANALYZER="/usr/bin/clang++-10" CLANG_SCAN_BUILD=scan-build-10 USE_CLANG=1 make V=1 -j32 analyze | .circleci/cat_ignore_eagain # aligned new doesn't work for reason we haven't figured out. For unknown, reason passing "clang++-10" as CLANG_ANALYZER doesn't work, and we need a full path.
      - post-steps

-  build-linux-cmake:
+  build-linux-cmake-no_test_run:
    machine:
      image: ubuntu-1604:202007-01
-    resource_class: 2xlarge
+    resource_class: large
    steps:
      - checkout # check out the code in the project directory
-      - run: (mkdir build && cd build && cmake -DWITH_GFLAGS=0 .. && make V=1 -j32) | .circleci/cat_ignore_eagain
+      - run: (mkdir build && cd build && cmake -DWITH_GFLAGS=0 .. && make V=1 -j8) | .circleci/cat_ignore_eagain
      - post-steps

  build-linux-unity:
    docker: # executor type
      - image: gcc:latest
-    resource_class: xlarge
+    resource_class: large
    steps:
      - checkout # check out the code in the project directory
      - run: apt-get update -y && apt-get install -y libgflags-dev
-      - run: TEST_TMPDIR=/dev/shm && make V=1 -j16 unity_test | .circleci/cat_ignore_eagain
+      - run: TEST_TMPDIR=/dev/shm && make V=1 -j8 unity_test | .circleci/cat_ignore_eagain
      - post-steps

-  build-linux-gcc-4-8:
+  build-linux-gcc-4_8-no_test_run:
    machine:
      image: ubuntu-1604:202007-01
    resource_class: large
    steps:
      - pre-steps
      - run: sudo apt-get update -y && sudo apt-get install gcc-4.8 g++-4.8 libgflags-dev
-      - run: CC=gcc-4.8 CXX=g++-4.8 V=1 SKIP_LINK=1 make -j4 all | .circleci/cat_ignore_eagain # Linking broken because libgflags compiled with newer ABI
+      - run: CC=gcc-4.8 CXX=g++-4.8 V=1 SKIP_LINK=1 make -j8 all | .circleci/cat_ignore_eagain # Linking broken because libgflags compiled with newer ABI
+      - post-steps
+
+  build-linux-gcc-8-no_test_run:
+    machine:
+      image: ubuntu-2004:202010-01
+    resource_class: large
+    steps:
+      - pre-steps
+      - run: sudo apt-get update -y && sudo apt-get install gcc-8 g++-8 libgflags-dev
+      - run: CC=gcc-8 CXX=g++-8 V=1 SKIP_LINK=1 make -j8 all | .circleci/cat_ignore_eagain # Linking broken because libgflags compiled with newer ABI
+      - post-steps
+
+  build-linux-gcc-9-no_test_run:
+    machine:
+      image: ubuntu-2004:202010-01
+    resource_class: large
+    steps:
+      - pre-steps
+      - run: sudo apt-get update -y && sudo apt-get install gcc-9 g++-9 libgflags-dev
+      - run: CC=gcc-9 CXX=g++-9 V=1 SKIP_LINK=1 make -j8 all | .circleci/cat_ignore_eagain # Linking broken because libgflags compiled with newer ABI
+      - post-steps
+
+  build-linux-gcc-10-cxx20-no_test_run:
+    machine:
+      image: ubuntu-2004:202010-01
+    resource_class: xlarge
+    steps:
+      - pre-steps
+      - run: sudo apt-get update -y && sudo apt-get install gcc-10 g++-10 libgflags-dev
+      - run: CC=gcc-10 CXX=g++-10 V=1 SKIP_LINK=1 ROCKSDB_CXX_STANDARD=c++20 make -j16 all | .circleci/cat_ignore_eagain # Linking broken because libgflags compiled with newer ABI
      - post-steps

  build-windows:
@@ -336,25 +394,102 @@ jobs:
  build-linux-java:
    machine:
      image: ubuntu-1604:202007-01
-    resource_class: 2xlarge
+    resource_class: large
+    environment:
+      JAVA_HOME: /usr/lib/jvm/java-1.8.0-openjdk-amd64
    steps:
      - pre-steps
      - install-gflags
      - run:
-          name: "Build RocksDBJava"
+          name: "Set Java Environment"
          command: |
-            export JAVA_HOME=/usr/lib/jvm/java-1.8.0-openjdk-amd64
-            export PATH=$JAVA_HOME/bin:$PATH
            echo "JAVA_HOME=${JAVA_HOME}"
+            echo 'export PATH=$JAVA_HOME/bin:$PATH' >> $BASH_ENV
            which java && java -version
            which javac && javac -version
-            make V=1 J=32 -j32 rocksdbjava jtest | .circleci/cat_ignore_eagain
+      - run:
+          name: "Build RocksDBJava Shared Library"
+          command: make V=1 J=8 -j8 rocksdbjava | .circleci/cat_ignore_eagain
+      - run:
+          name: "Test RocksDBJava"
+          command: make V=1 J=8 -j8 jtest | .circleci/cat_ignore_eagain
+      - post-steps
+
+  build-linux-java-static:
+    machine:
+      image: ubuntu-1604:202007-01
+    resource_class: large
+    environment:
+      JAVA_HOME: /usr/lib/jvm/java-1.8.0-openjdk-amd64
+    steps:
+      - pre-steps
+      - install-gflags
+      - run:
+          name: "Set Java Environment"
+          command: |
+            echo "JAVA_HOME=${JAVA_HOME}"
+            echo 'export PATH=$JAVA_HOME/bin:$PATH' >> $BASH_ENV
+            which java && java -version
+            which javac && javac -version
+      - run:
+          name: "Build RocksDBJava Static Library"
+          command: make V=1 J=8 -j8 rocksdbjavastatic | .circleci/cat_ignore_eagain
+      - post-steps
+
+  build-macos-java:
+    macos:
+      xcode: 9.4.1
+    resource_class: medium
+    environment:
+      JAVA_HOME: /Library/Java/JavaVirtualMachines/jdk1.8.0_172.jdk/Contents/Home
+    steps:
+      - increase-max-open-files-on-macos
+      - install-pyenv-on-macos
+      - install-gflags-on-macos
+      - pre-steps-macos
+      - run:
+          name: "Set Java Environment"
+          command: |
+            echo "JAVA_HOME=${JAVA_HOME}"
+            echo 'export PATH=$JAVA_HOME/bin:$PATH' >> $BASH_ENV
+            which java && java -version
+            which javac && javac -version
+      - run:
+          name: "Build RocksDBJava Shared Library"
+          command: make V=1 J=8 -j8 rocksdbjava | .circleci/cat_ignore_eagain
+      - run:
+          name: "Test RocksDBJava"
+          command: make V=1 J=8 -j8 jtest | .circleci/cat_ignore_eagain
+      - post-steps
+
+  build-macos-java-static:
+    macos:
+      xcode: 9.4.1
+    resource_class: medium
+    environment:
+      JAVA_HOME: /Library/Java/JavaVirtualMachines/jdk1.8.0_172.jdk/Contents/Home
+    steps:
+      - increase-max-open-files-on-macos
+      - install-pyenv-on-macos
+      - install-gflags-on-macos
+      - install-cmake-on-macos
+      - pre-steps-macos
+      - run:
+          name: "Set Java Environment"
+          command: |
+            echo "JAVA_HOME=${JAVA_HOME}"
+            echo 'export PATH=$JAVA_HOME/bin:$PATH' >> $BASH_ENV
+            which java && java -version
+            which javac && javac -version
+      - run:
+          name: "Build RocksDBJava Static Library"
+          command: make V=1 J=8 -j8 rocksdbjavastatic | .circleci/cat_ignore_eagain
      - post-steps

  build-examples:
    machine:
      image: ubuntu-1604:202007-01
-    resource_class: medium
+    resource_class: large
    steps:
      - pre-steps
      - install-gflags
@@ -364,6 +499,29 @@ jobs:
            OPT=-DTRAVIS V=1 make -j4 static_lib && cd examples && make -j4 | ../.circleci/cat_ignore_eagain
      - post-steps

+  build-cmake-mingw:
+    machine:
+      image: ubuntu-1604:202007-01
+    resource_class: 2xlarge
+    steps:
+      - pre-steps
+      - install-gflags
+      - run: sudo apt-get update -y && sudo apt-get install -y mingw-w64
+      - run: sudo update-alternatives --set x86_64-w64-mingw32-g++ /usr/bin/x86_64-w64-mingw32-g++-posix
+      - run:
+          name: "Build cmake-mingw"
+          command: |
+            sudo apt-get install snapd && sudo snap install cmake --beta --classic
+            export PATH=/snap/bin:$PATH
+            sudo apt-get install -y openjdk-8-jdk
+            export JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64
+            export PATH=$JAVA_HOME/bin:$PATH
+            echo "JAVA_HOME=${JAVA_HOME}"
+            which java && java -version
+            which javac && javac -version
+            mkdir build && cd build && cmake -DJNI=1 -DWITH_GFLAGS=OFF .. -DCMAKE_C_COMPILER=x86_64-w64-mingw32-gcc -DCMAKE_CXX_COMPILER=x86_64-w64-mingw32-g++ -DCMAKE_SYSTEM_NAME=Windows && make -j4 rocksdb rocksdbjni
+      - post-steps
+
  build-linux-non-shm:
    machine:
      image: ubuntu-1604:202007-01
@@ -391,7 +549,7 @@ jobs:
            sed -i 's/[[:space:]]*$//; s/ / \.\//g; s/.*/.\/&/' /tmp/test_list
            cat /tmp/test_list
            export TEST_TMPDIR=/tmp/rocksdb_test_tmp
-            /usr/bin/python ../gtest-parallel/gtest-parallel $(</tmp/test_list) --output_dir=/tmp | cat  # pipe to cat to continuously output status on circleci UI. Otherwise, no status will be printed while the job is running.
+            gtest-parallel $(</tmp/test_list) --output_dir=/tmp | cat  # pipe to cat to continuously output status on circleci UI. Otherwise, no status will be printed while the job is running.
      - post-steps

 workflows:
@@ -419,9 +577,6 @@ workflows:
  build-linux-lite-release:
    jobs:
      - build-linux-lite-release
-  build-linux-clang-no-test:
-    jobs:
-      - build-linux-clang-no-test
  build-linux-clang10-asan:
    jobs:
      - build-linux-clang10-asan
@@ -434,9 +589,6 @@ workflows:
  build-linux-clang10-clang-analyze:
    jobs:
      - build-linux-clang10-clang-analyze
-  build-linux-cmake:
-    jobs:
-      - build-linux-cmake
  build-linux-unity:
    jobs:
      - build-linux-unity
@@ -460,6 +612,9 @@ workflows:
  build-java:
    jobs:
      - build-linux-java
+      - build-linux-java-static
+      - build-macos-java
+      - build-macos-java-static
  build-examples:
    jobs:
      - build-examples
@@ -477,9 +632,20 @@ workflows:
      - build-linux-non-shm:
          start_test: "statistics_test" # make sure unique in src.mk
          end_test: ""
-  build-linux-gcc-4-8:
+  build-linux-compilers-no_test_run:
    jobs:
-      - build-linux-gcc-4-8
+      - build-linux-clang-no_test_run
+      - build-linux-cmake-no_test_run
+      - build-linux-gcc-4_8-no_test_run
+      - build-linux-gcc-8-no_test_run
+      - build-linux-gcc-9-no_test_run
+      - build-linux-gcc-10-cxx20-no_test_run
  build-macos:
    jobs:
      - build-macos
+  build-macos-cmake:
+    jobs:
+      - build-macos-cmake
+  build-cmake-mingw:
+    jobs:
+      - build-cmake-mingw
@@ -10,7 +10,7 @@ $process = Start-Process "${PWD}\vs_installer.exe" -ArgumentList $VS_INSTALL_ARG
 Remove-Item -Path vs_installer.exe -Force
 $exitCode = $process.ExitCode
 if (($exitCode -ne 0) -and ($exitCode -ne 3010)) {
-    echo "VS 2017 installer exited with code $exitCode, which should be one of [0, 3010]."
+    echo "VS 2015 installer exited with code $exitCode, which should be one of [0, 3010]."
    curl.exe --retry 3 -kL $COLLECT_DOWNLOAD_LINK --output Collect.exe
    if ($LASTEXITCODE -ne 0) {
        echo "Download of the VS Collect tool failed."
@@ -90,3 +90,6 @@ buckifier/__pycache__
 compile_commands.json
 clang-format-diff.py
 .py3/
+
+fuzz/proto/gen/
+fuzz/crash-*
@@ -86,25 +86,24 @@ matrix:
    env: JOB_NAME=make-gcc4.8
  - os: linux
    compiler: clang
-  # Exclude all but most unique cmake variants for pull requests, but build all in branches
+  # With migration to CircleCI, exclude Linux/amd64 for pull requests
+  # (but build in branches for now)
  - if: type = pull_request AND commit_message !~ /FULL_CI/
    os : linux
    arch: amd64
-    env: JOB_NAME=cmake
-  - if: type = pull_request AND commit_message !~ /FULL_CI/
-    os : linux
-    arch: amd64
-    env: JOB_NAME=cmake-gcc8
-  - if: type = pull_request AND commit_message !~ /FULL_CI/
-    os : linux
-    arch: amd64
-    env: JOB_NAME=cmake-gcc9
  # Exclude most osx, arm64 and ppc64le tests for pull requests, but build in branches
  # Temporarily disable ppc64le cmake test while snapd is broken
  - if: type = pull_request AND commit_message !~ /FULL_CI/
    os: linux
    arch: ppc64le
    env: JOB_NAME=cmake
+  # Exclude all osx since it should be covered by CircleCI
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
+    os: osx
+    env: TEST_GROUP=platform_dependent
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
+    os: osx
+    env: JOB_NAME=cmake
  # NB: the cmake build is a partial java test
  - if: type = pull_request AND commit_message !~ /FULL_CI/
    os: osx
@@ -814,6 +814,8 @@ set(SOURCES
        utilities/transactions/lock/lock_manager.cc
        utilities/transactions/lock/point/point_lock_tracker.cc
        utilities/transactions/lock/point/point_lock_manager.cc
+        utilities/transactions/lock/range/range_tree/range_tree_lock_manager.cc
+        utilities/transactions/lock/range/range_tree/range_tree_lock_tracker.cc
        utilities/transactions/optimistic_transaction_db_impl.cc
        utilities/transactions/optimistic_transaction.cc
        utilities/transactions/pessimistic_transaction.cc
@@ -831,6 +833,20 @@ set(SOURCES
        utilities/write_batch_with_index/write_batch_with_index_internal.cc
        $<TARGET_OBJECTS:build_version>)

+list(APPEND SOURCES
+  utilities/transactions/lock/range/range_tree/lib/locktree/concurrent_tree.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/keyrange.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/lock_request.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/locktree.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/manager.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/range_buffer.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/treenode.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/txnid_set.cc
+  utilities/transactions/lock/range/range_tree/lib/locktree/wfg.cc
+  utilities/transactions/lock/range/range_tree/lib/standalone_port.cc
+  utilities/transactions/lock/range/range_tree/lib/util/dbt.cc
+  utilities/transactions/lock/range/range_tree/lib/util/memarena.cc)
+
 if(HAVE_SSE42 AND NOT MSVC)
  set_source_files_properties(
    util/crc32c.cc
@@ -1160,6 +1176,7 @@ if(WITH_TESTS)
        table/table_test.cc
        table/block_fetcher_test.cc
        test_util/testutil_test.cc
+        trace_replay/block_cache_tracer_test.cc
        trace_replay/io_tracer_test.cc
        tools/block_cache_analyzer/block_cache_trace_analyzer_test.cc
        tools/io_tracer_parser_test.cc
@@ -1210,6 +1227,7 @@ if(WITH_TESTS)
        utilities/transactions/lock/point/point_lock_manager_test.cc
        utilities/transactions/write_prepared_transaction_test.cc
        utilities/transactions/write_unprepared_transaction_test.cc
+        utilities/transactions/lock/range/range_locking_test.cc
        utilities/ttl/ttl_test.cc
        utilities/write_batch_with_index/write_batch_with_index_test.cc
    )
@@ -1,4 +1,24 @@
 # Rocksdb Change Log
+## 6.16.0 (12/18/2020)
+### Behavior Changes
+* Attempting to write a merge operand without explicitly configuring `merge_operator` now fails immediately, causing the DB to enter read-only mode. Previously, failure was deferred until the `merge_operator` was needed by a user read or a background operation.
+
+### Bug Fixes
+* Truncated WALs ending in incomplete records can no longer produce gaps in the recovered data when `WALRecoveryMode::kPointInTimeRecovery` is used. Gaps are still possible when WALs are truncated exactly on record boundaries; for complete protection, users should enable `track_and_verify_wals_in_manifest`.
+* Fix a bug where compressed blocks read by MultiGet are not inserted into the compressed block cache when use_direct_reads = true.
+* Fixed the issue of full scanning on obsolete files when there are too many outstanding compactions with ConcurrentTaskLimiter enabled.
+* Fixed the logic of populating native data structure for `read_amp_bytes_per_bit` during OPTIONS file parsing on big-endian architecture. Without this fix, original code introduced in PR7659, when running on big-endian machine, can mistakenly store read_amp_bytes_per_bit (an uint32) in little endian format. Future access to `read_amp_bytes_per_bit` will give wrong values. Little endian architecture is not affected.
+* Fixed prefix extractor with timestamp issues.
+* Fixed a bug in atomic flush: in two-phase commit mode, the minimum WAL log number to keep is incorrect.
+
+### New Features
+* User defined timestamp feature supports `CompactRange` and `GetApproximateSizes`.
+* Support getting aggregated table properties (kAggregatedTableProperties and kAggregatedTablePropertiesAtLevel) with DB::GetMapProperty, for easier access to the data in a structured format.
+* Experimental option BlockBasedTableOptions::optimize_filters_for_memory now works with experimental Ribbon filter (as well as Bloom filter).
+
+### Public API Change
+* Deprecated public but rarely-used FilterBitsBuilder::CalculateNumEntry, which is replaced with ApproximateNumEntries taking a size_t parameter and returning size_t.
+
 ## 6.15.0 (11/13/2020)
 ### Bug Fixes
 * Fixed a bug in the following combination of features: indexes with user keys (`format_version >= 3`), indexes are partitioned (`index_type == kTwoLevelIndexSearch`), and some index partitions are pinned in memory (`BlockBasedTableOptions::pin_l0_filter_and_index_blocks_in_cache`). The bug could cause keys to be truncated when read from the index leading to wrong read results or other unexpected behavior.
@@ -13,10 +33,13 @@
 * Fixed MultiGet bugs it doesn't return valid data with user defined timestamp.
 * Fixed a potential bug caused by evaluating `TableBuilder::NeedCompact()` before `TableBuilder::Finish()` in compaction job. For example, the `NeedCompact()` method of `CompactOnDeletionCollector` returned by built-in `CompactOnDeletionCollectorFactory` requires `BlockBasedTable::Finish()` to return the correct result. The bug can cause a compaction-generated file not to be marked for future compaction based on deletion ratio.
 * Fixed a seek issue with prefix extractor and timestamp.
+* Fixed a bug of encoding and parsing BlockBasedTableOptions::read_amp_bytes_per_bit as a 64-bit integer.
+* Fixed a bug of a recovery corner case, details in PR7621.

 ### Public API Change
 * Deprecate `BlockBasedTableOptions::pin_l0_filter_and_index_blocks_in_cache` and `BlockBasedTableOptions::pin_top_level_index_and_filter`. These options still take effect until users migrate to the replacement APIs in `BlockBasedTableOptions::metadata_cache_options`. Migration guidance can be found in the API comments on the deprecated options.
 * Add new API `DB::VerifyFileChecksums` to verify SST file checksum with corresponding entries in the MANIFEST if present. Current implementation requires scanning and recomputing file checksums.
+* Added a new option `track_and_verify_wals_in_manifest`. If `true`, the log numbers and sizes of the synced WALs are tracked in MANIFEST, then during DB recovery, if a synced WAL is missing from disk, or the WAL's size does not match the recorded size in MANIFEST, an error will be reported and the recovery will be aborted. Note that this option does not work with secondary instance.

 ### Behavior Changes
 * The dictionary compression settings specified in `ColumnFamilyOptions::compression_opts` now additionally affect files generated by flush and compaction to non-bottommost level. Previously those settings at most affected files generated by compaction to bottommost level, depending on whether `ColumnFamilyOptions::bottommost_compression_opts` overrode them. Users who relied on dictionary compression settings in `ColumnFamilyOptions::compression_opts` affecting only the bottommost level can keep the behavior by moving their dictionary settings to `ColumnFamilyOptions::bottommost_compression_opts` and setting its `enabled` flag.
@@ -99,14 +99,14 @@ endif

 ifneq ($(findstring rocksdbjava, $(MAKECMDGOALS)),)
 	LIB_MODE=shared
-        ifneq ($(findstring rocksdbjavastatic, $(MAKECMDGOALS)),)
+	ifneq ($(findstring rocksdbjavastatic, $(MAKECMDGOALS)),)
 		OBJ_DIR=jls
-	        ifneq ($(DEBUG_LEVEL),2)
-	            DEBUG_LEVEL=0
-                endif
-                ifeq ($(MAKECMDGOALS),rocksdbjavastaticpublish)
-	            DEBUG_LEVEL=0
-                endif
+		ifneq ($(DEBUG_LEVEL),2)
+			DEBUG_LEVEL=0
+		endif
+		ifeq ($(MAKECMDGOALS),rocksdbjavastaticpublish)
+			DEBUG_LEVEL=0
+		endif
 	else
 		OBJ_DIR=jl
 	endif
@@ -190,6 +190,11 @@ else
 endif

 ifdef ASSERT_STATUS_CHECKED
+# For ASC, turn off constructor elision, preventing the case where a constructor returned
+# by a method may pass the ASC check if the status is checked in the inner method.  Forcing
+# the copy constructor to be invoked disables the optimization and will cause the calling method
+# to check the status in order to prevent an error from being raised.
+PLATFORM_CXXFLAGS += -fno-elide-constructors
 ifeq ($(filter -DROCKSDB_ASSERT_STATUS_CHECKED,$(OPT)),)
 	OPT += -DROCKSDB_ASSERT_STATUS_CHECKED
 endif
@@ -558,6 +563,7 @@ PARALLEL_TEST = \
 	table_test \
 	transaction_test \
 	point_lock_manager_test \
+	range_locking_test \
 	write_prepared_transaction_test \
 	write_unprepared_transaction_test \

@@ -586,21 +592,56 @@ ifdef ASSERT_STATUS_CHECKED
 		blob_file_reader_test \
 		bloom_test \
 		cassandra_format_test \
+		cassandra_functional_test \
 		cassandra_row_merge_test \
 		cassandra_serialize_test \
 		cleanable_test \
+		checkpoint_test \
 		coding_test \
 		crc32c_test \
 		dbformat_test \
 		db_basic_test \
+		compact_files_test \
+		compaction_picker_test \
+		comparator_db_test \
+		db_encryption_test \
+		db_iter_test \
+		db_iter_stress_test \
+		db_log_iter_test \
+		db_bloom_filter_test \
 		db_blob_basic_test \
+		db_blob_index_test \
+		db_block_cache_test \
 		db_flush_test \
+		db_iterator_test \
+		db_logical_block_size_cache_test \
+		db_memtable_test \
+		db_merge_operand_test \
+		db_merge_operator_test \
+		db_wal_test \
 		db_with_timestamp_basic_test \
 		db_with_timestamp_compaction_test \
+		db_write_test \
 		db_options_test \
 		db_properties_test \
+		db_range_del_test \
 		db_secondary_test \
+		deletefile_test \
+		external_sst_file_test \
 		options_file_test \
+		db_statistics_test \
+		db_table_properties_test \
+		db_tailing_iter_test \
+		fault_injection_test \
+		listener_test \
+		log_test \
+		manual_compaction_test \
+		obsolete_files_test \
+		perf_context_test \
+		periodic_work_scheduler_test \
+		perf_context_test \
+		version_set_test \
+		wal_manager_test \
 		defer_test \
 		filename_test \
 		dynamic_bloom_test \
@@ -609,6 +650,7 @@ ifdef ASSERT_STATUS_CHECKED
 		env_logger_test \
 		event_logger_test \
 		error_handler_fs_test \
+		external_sst_file_basic_test \
 		auto_roll_logger_test \
 		file_indexer_test \
 		flush_job_test \
@@ -621,9 +663,11 @@ ifdef ASSERT_STATUS_CHECKED
 		iostats_context_test \
 		ldb_cmd_test \
 		memkind_kmem_allocator_test \
+		merge_test \
 		merger_test \
 		mock_env_test \
 		object_registry_test \
+		optimistic_transaction_test \
 		prefix_test \
 		plain_table_db_test \
 		repair_test \
@@ -631,6 +675,8 @@ ifdef ASSERT_STATUS_CHECKED
 		customizable_test \
 		options_settable_test \
 		options_test \
+		point_lock_manager_test \
+		random_access_file_reader_test \
 		random_test \
 		range_del_aggregator_test \
 		sst_file_reader_test \
@@ -639,11 +685,14 @@ ifdef ASSERT_STATUS_CHECKED
 		ribbon_test \
 		skiplist_test \
 		slice_test \
+		slice_transform_test \
 		sst_dump_test \
 		statistics_test \
 		stats_history_test \
+		stringappend_test \
 		thread_local_test \
 		trace_analyzer_test \
+		transaction_test \
 		env_timed_test \
 		filelock_test \
 		timer_queue_test \
@@ -658,25 +707,42 @@ ifdef ASSERT_STATUS_CHECKED
 		version_builder_test \
 		version_edit_test \
 		work_queue_test \
+		write_buffer_manager_test \
 		write_controller_test \
+		write_prepared_transaction_test \
+		write_unprepared_transaction_test \
 		compaction_iterator_test \
 		compaction_job_test \
 		compaction_job_stats_test \
-	        io_tracer_test \
+		io_tracer_test \
+		io_tracer_parser_test \
+		prefetch_test \
 		merge_helper_test \
 		memtable_list_test \
 		flush_job_test \
 		block_based_filter_block_test \
 		block_fetcher_test \
+		block_test \
+		data_block_hash_index_test \
 		full_filter_block_test \
 		partitioned_filter_block_test \
 		column_family_test \
 		file_reader_writer_test \
+		rate_limiter_test \
 		corruption_test \
+		reduce_levels_test \
+		thread_list_test \
+		compact_on_deletion_collector_test \
 		db_universal_compaction_test \
 		import_column_family_test \
+		option_change_migration_test \
+		cuckoo_table_builder_test \
+		cuckoo_table_db_test \
+		cuckoo_table_reader_test \
 		memory_test \
 		table_test \
+		write_batch_test \
+		write_batch_with_index_test \

 ifeq ($(USE_FOLLY_DISTRIBUTED_MUTEX),1)
 TESTS_PASSING_ASC += folly_synchronization_distributed_mutex_test
@@ -1909,6 +1975,9 @@ blob_db_test: $(OBJ_DIR)/utilities/blob_db/blob_db_test.o $(TEST_LIBRARY) $(LIBR
 repeatable_thread_test: $(OBJ_DIR)/util/repeatable_thread_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)

+range_locking_test: utilities/transactions/lock/range/range_locking_test.o $(TEST_LIBRARY) $(LIBRARY)
+	$(AM_LINK)
+
 range_tombstone_fragmenter_test: $(OBJ_DIR)/db/range_tombstone_fragmenter_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)

@@ -2106,80 +2175,80 @@ ifeq ($(PLATFORM), OS_AIX)
 	SNAPPY_MAKE_TARGET = libsnappy.la
 endif
 ifeq ($(PLATFORM), OS_OPENBSD)
-        JAVA_INCLUDE = -I$(JAVA_HOME)/include -I$(JAVA_HOME)/include/openbsd
+	JAVA_INCLUDE = -I$(JAVA_HOME)/include -I$(JAVA_HOME)/include/openbsd
 	ROCKSDBJNILIB = librocksdbjni-openbsd$(ARCH).so
-        ROCKSDB_JAR = rocksdbjni-$(ROCKSDB_JAVA_VERSION)-openbsd$(ARCH).jar
+	ROCKSDB_JAR = rocksdbjni-$(ROCKSDB_JAVA_VERSION)-openbsd$(ARCH).jar
 endif

-libz.a:
-	-rm -rf zlib-$(ZLIB_VER)
-ifeq (,$(wildcard ./zlib-$(ZLIB_VER).tar.gz))
+zlib-$(ZLIB_VER).tar.gz:
 	curl --fail --output zlib-$(ZLIB_VER).tar.gz --location ${ZLIB_DOWNLOAD_BASE}/zlib-$(ZLIB_VER).tar.gz
-endif
 	ZLIB_SHA256_ACTUAL=`$(SHA256_CMD) zlib-$(ZLIB_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(ZLIB_SHA256)" != "$$ZLIB_SHA256_ACTUAL" ]; then \
 		echo zlib-$(ZLIB_VER).tar.gz checksum mismatch, expected=\"$(ZLIB_SHA256)\" actual=\"$$ZLIB_SHA256_ACTUAL\"; \
 		exit 1; \
 	fi
+
+libz.a: zlib-$(ZLIB_VER).tar.gz
+	-rm -rf zlib-$(ZLIB_VER)
 	tar xvzf zlib-$(ZLIB_VER).tar.gz
-	cd zlib-$(ZLIB_VER) && CFLAGS='-fPIC ${EXTRA_CFLAGS}' LDFLAGS='${EXTRA_LDFLAGS}' ./configure --static && $(MAKE)
+	cd zlib-$(ZLIB_VER) && CFLAGS='-fPIC ${JAVA_STATIC_DEPS_CCFLAGS} ${EXTRA_CFLAGS}' LDFLAGS='${JAVA_STATIC_DEPS_LDFLAGS} ${EXTRA_LDFLAGS}' ./configure --static && $(MAKE)
 	cp zlib-$(ZLIB_VER)/libz.a .

-libbz2.a:
-	-rm -rf bzip2-$(BZIP2_VER)
-ifeq (,$(wildcard ./bzip2-$(BZIP2_VER).tar.gz))
+bzip2-$(BZIP2_VER).tar.gz:
 	curl --fail --output bzip2-$(BZIP2_VER).tar.gz --location ${CURL_SSL_OPTS} ${BZIP2_DOWNLOAD_BASE}/bzip2-$(BZIP2_VER).tar.gz
-endif
 	BZIP2_SHA256_ACTUAL=`$(SHA256_CMD) bzip2-$(BZIP2_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(BZIP2_SHA256)" != "$$BZIP2_SHA256_ACTUAL" ]; then \
 		echo bzip2-$(BZIP2_VER).tar.gz checksum mismatch, expected=\"$(BZIP2_SHA256)\" actual=\"$$BZIP2_SHA256_ACTUAL\"; \
 		exit 1; \
 	fi
+
+libbz2.a: bzip2-$(BZIP2_VER).tar.gz
+	-rm -rf bzip2-$(BZIP2_VER)
 	tar xvzf bzip2-$(BZIP2_VER).tar.gz
-	cd bzip2-$(BZIP2_VER) && $(MAKE) CFLAGS='-fPIC -O2 -g -D_FILE_OFFSET_BITS=64 ${EXTRA_CFLAGS}' AR='ar ${EXTRA_ARFLAGS}'
+	cd bzip2-$(BZIP2_VER) && $(MAKE) CFLAGS='-fPIC -O2 -g -D_FILE_OFFSET_BITS=64 ${JAVA_STATIC_DEPS_CCFLAGS} ${EXTRA_CFLAGS}' LDFLAGS='${JAVA_STATIC_DEPS_LDFLAGS} ${EXTRA_LDFLAGS}' AR='ar ${EXTRA_ARFLAGS}'
 	cp bzip2-$(BZIP2_VER)/libbz2.a .

-libsnappy.a:
-	-rm -rf snappy-$(SNAPPY_VER)
-ifeq (,$(wildcard ./snappy-$(SNAPPY_VER).tar.gz))
+snappy-$(SNAPPY_VER).tar.gz:
 	curl --fail --output snappy-$(SNAPPY_VER).tar.gz --location ${CURL_SSL_OPTS} ${SNAPPY_DOWNLOAD_BASE}/$(SNAPPY_VER).tar.gz
-endif
 	SNAPPY_SHA256_ACTUAL=`$(SHA256_CMD) snappy-$(SNAPPY_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(SNAPPY_SHA256)" != "$$SNAPPY_SHA256_ACTUAL" ]; then \
 		echo snappy-$(SNAPPY_VER).tar.gz checksum mismatch, expected=\"$(SNAPPY_SHA256)\" actual=\"$$SNAPPY_SHA256_ACTUAL\"; \
 		exit 1; \
 	fi
+
+libsnappy.a: snappy-$(SNAPPY_VER).tar.gz
+	-rm -rf snappy-$(SNAPPY_VER)
 	tar xvzf snappy-$(SNAPPY_VER).tar.gz
 	mkdir snappy-$(SNAPPY_VER)/build
-	cd snappy-$(SNAPPY_VER)/build && CFLAGS='${EXTRA_CFLAGS}' CXXFLAGS='${EXTRA_CXXFLAGS}' LDFLAGS='${EXTRA_LDFLAGS}' cmake -DCMAKE_POSITION_INDEPENDENT_CODE=ON .. && $(MAKE) ${SNAPPY_MAKE_TARGET}
+	cd snappy-$(SNAPPY_VER)/build && CFLAGS='${JAVA_STATIC_DEPS_CCFLAGS} ${EXTRA_CFLAGS}' CXXFLAGS='${JAVA_STATIC_DEPS_CXXFLAGS} ${EXTRA_CXXFLAGS}' LDFLAGS='${JAVA_STATIC_DEPS_LDFLAGS} ${EXTRA_LDFLAGS}' cmake -DCMAKE_POSITION_INDEPENDENT_CODE=ON ${PLATFORM_CMAKE_FLAGS} .. && $(MAKE) ${SNAPPY_MAKE_TARGET}
 	cp snappy-$(SNAPPY_VER)/build/libsnappy.a .

-liblz4.a:
-	-rm -rf lz4-$(LZ4_VER)
-ifeq (,$(wildcard ./lz4-$(LZ4_VER).tar.gz))
+lz4-$(LZ4_VER).tar.gz:
 	curl --fail --output lz4-$(LZ4_VER).tar.gz --location ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
-endif
 	LZ4_SHA256_ACTUAL=`$(SHA256_CMD) lz4-$(LZ4_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(LZ4_SHA256)" != "$$LZ4_SHA256_ACTUAL" ]; then \
 		echo lz4-$(LZ4_VER).tar.gz checksum mismatch, expected=\"$(LZ4_SHA256)\" actual=\"$$LZ4_SHA256_ACTUAL\"; \
 		exit 1; \
 	fi
+
+liblz4.a: lz4-$(LZ4_VER).tar.gz
+	-rm -rf lz4-$(LZ4_VER)
 	tar xvzf lz4-$(LZ4_VER).tar.gz
-	cd lz4-$(LZ4_VER)/lib && $(MAKE) CFLAGS='-fPIC -O2 ${EXTRA_CFLAGS}' all
+	cd lz4-$(LZ4_VER)/lib && $(MAKE) CFLAGS='-fPIC -O2 ${JAVA_STATIC_DEPS_CCFLAGS} ${EXTRA_CFLAGS}' LDFLAGS='${JAVA_STATIC_DEPS_LDFLAGS} ${EXTRA_LDFLAGS}' all
 	cp lz4-$(LZ4_VER)/lib/liblz4.a .

-libzstd.a:
-	-rm -rf zstd-$(ZSTD_VER)
-ifeq (,$(wildcard ./zstd-$(ZSTD_VER).tar.gz))
+zstd-$(ZSTD_VER).tar.gz:
 	curl --fail --output zstd-$(ZSTD_VER).tar.gz --location ${CURL_SSL_OPTS} ${ZSTD_DOWNLOAD_BASE}/v$(ZSTD_VER).tar.gz
-endif
 	ZSTD_SHA256_ACTUAL=`$(SHA256_CMD) zstd-$(ZSTD_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(ZSTD_SHA256)" != "$$ZSTD_SHA256_ACTUAL" ]; then \
 		echo zstd-$(ZSTD_VER).tar.gz checksum mismatch, expected=\"$(ZSTD_SHA256)\" actual=\"$$ZSTD_SHA256_ACTUAL\"; \
 		exit 1; \
 	fi
+
+libzstd.a: zstd-$(ZSTD_VER).tar.gz
+	-rm -rf zstd-$(ZSTD_VER)
 	tar xvzf zstd-$(ZSTD_VER).tar.gz
-	cd zstd-$(ZSTD_VER)/lib && DESTDIR=. PREFIX= $(MAKE) CFLAGS='-fPIC -O2 ${EXTRA_CFLAGS}' install
+	cd zstd-$(ZSTD_VER)/lib && DESTDIR=. PREFIX= $(MAKE) CFLAGS='-fPIC -O2 ${JAVA_STATIC_DEPS_CCFLAGS} ${EXTRA_CFLAGS}' LDFLAGS='${JAVA_STATIC_DEPS_LDFLAGS} ${EXTRA_LDFLAGS}' libzstd.a
 	cp zstd-$(ZSTD_VER)/lib/libzstd.a .

 # A version of each $(LIB_OBJECTS) compiled with -fPIC and a fixed set of static compression libraries
@@ -2188,14 +2257,23 @@ JAVA_COMPRESSIONS = libz.a libbz2.a libsnappy.a liblz4.a libzstd.a
 endif

 JAVA_STATIC_FLAGS = -DZLIB -DBZIP2 -DSNAPPY -DLZ4 -DZSTD
-JAVA_STATIC_INCLUDES = -I./zlib-$(ZLIB_VER) -I./bzip2-$(BZIP2_VER) -I./snappy-$(SNAPPY_VER) -I./lz4-$(LZ4_VER)/lib -I./zstd-$(ZSTD_VER)/lib/include
-ifneq ($(findstring rocksdbjavastatic, $(MAKECMDGOALS)),)
+JAVA_STATIC_INCLUDES = -I./zlib-$(ZLIB_VER) -I./bzip2-$(BZIP2_VER) -I./snappy-$(SNAPPY_VER) -I./snappy-$(SNAPPY_VER)/build -I./lz4-$(LZ4_VER)/lib -I./zstd-$(ZSTD_VER)/lib -I./zstd-$(ZSTD_VER)/lib/dictBuilder
+
+ifneq ($(findstring rocksdbjavastatic, $(filter-out rocksdbjavastatic_deps, $(MAKECMDGOALS))),)
 CXXFLAGS += $(JAVA_STATIC_FLAGS) $(JAVA_STATIC_INCLUDES)
-CFLAGS +=  $(JAVA_STATIC_FLAGS) $(JAVA_STATIC_INCLUDES)
+CFLAGS += $(JAVA_STATIC_FLAGS) $(JAVA_STATIC_INCLUDES)
 endif
-rocksdbjavastatic: $(LIB_OBJECTS) $(JAVA_COMPRESSIONS)
-	cd java;$(MAKE) javalib;
-	rm -f ./java/target/$(ROCKSDBJNILIB)
+rocksdbjavastatic:
+ifeq ($(JAVA_HOME),)
+	$(error JAVA_HOME is not set)
+endif
+	$(MAKE) rocksdbjavastatic_deps
+	$(MAKE) rocksdbjavastatic_libobjects
+	$(MAKE) rocksdbjavastatic_javalib
+
+rocksdbjavastatic_javalib:
+	cd java;$(MAKE) javalib
+	rm -f java/target/$(ROCKSDBJNILIB)
 	$(CXX) $(CXXFLAGS) -I./java/. $(JAVA_INCLUDE) -shared -fPIC \
 	  -o ./java/target/$(ROCKSDBJNILIB) $(JNI_NATIVE_SOURCES) \
 	  $(LIB_OBJECTS) $(COVERAGEFLAGS) \
@@ -2212,6 +2290,10 @@ rocksdbjavastatic: $(LIB_OBJECTS) $(JAVA_COMPRESSIONS)
 	openssl sha1 java/target/$(ROCKSDB_JAVADOCS_JAR) | sed 's/.*= \([0-9a-f]*\)/\1/' > java/target/$(ROCKSDB_JAVADOCS_JAR).sha1
 	openssl sha1 java/target/$(ROCKSDB_SOURCES_JAR) | sed 's/.*= \([0-9a-f]*\)/\1/' > java/target/$(ROCKSDB_SOURCES_JAR).sha1

+rocksdbjavastatic_deps: $(JAVA_COMPRESSIONS)
+
+rocksdbjavastatic_libobjects: $(LIB_OBJECTS)
+
 rocksdbjavastaticrelease: rocksdbjavastatic
 	cd java/crossbuild && (vagrant destroy -f || true) && vagrant up linux32 && vagrant halt linux32 && vagrant up linux64 && vagrant halt linux64 && vagrant up linux64-musl && vagrant halt linux64-musl
 	cd java;jar -cf target/$(ROCKSDB_JAR_ALL) HISTORY*.md
@@ -2227,7 +2309,7 @@ rocksdbjavastaticreleasedocker: rocksdbjavastatic rocksdbjavastaticdockerx86 roc

 rocksdbjavastaticdockerx86:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_x86-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos6_x86-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_x86-be --platform linux/386 --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos6_x86-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticdockerx86_64:
 	mkdir -p java/target
@@ -2243,7 +2325,7 @@ rocksdbjavastaticdockerarm64v8:

 rocksdbjavastaticdockerx86musl:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_x86-musl-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_x86-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_x86-musl-be --platform linux/386 --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_x86-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticdockerx86_64musl:
 	mkdir -p java/target
@@ -2281,6 +2363,9 @@ jl/%.o: %.cc
 	$(AM_V_CC)mkdir -p $(@D) && $(CXX) $(CXXFLAGS) -fPIC -c $< -o $@ $(COVERAGEFLAGS)

 rocksdbjava: $(LIB_OBJECTS)
+ifeq ($(JAVA_HOME),)
+	$(error JAVA_HOME is not set)
+endif
 	$(AM_V_GEN)cd java;$(MAKE) javalib;
 	$(AM_V_at)rm -f ./java/target/$(ROCKSDBJNILIB)
 	$(AM_V_at)$(CXX) $(CXXFLAGS) -I./java/. $(JAVA_INCLUDE) -shared -fPIC -o ./java/target/$(ROCKSDBJNILIB) $(JNI_NATIVE_SOURCES) $(LIB_OBJECTS) $(JAVA_LDFLAGS) $(COVERAGEFLAGS)
@@ -2411,6 +2496,8 @@ ifneq ($(MAKECMDGOALS),clean)
 ifneq ($(MAKECMDGOALS),format)
 ifneq ($(MAKECMDGOALS),jclean)
 ifneq ($(MAKECMDGOALS),jtest)
+ifneq ($(MAKECMDGOALS),rocksdbjavastatic)
+ifneq ($(MAKECMDGOALS),rocksdbjavastatic_deps)
 ifneq ($(MAKECMDGOALS),package)
 ifneq ($(MAKECMDGOALS),analyze)
 -include $(DEPFILES)
@@ -2420,3 +2507,5 @@ endif
 endif
 endif
 endif
+endif
+endif
@@ -388,6 +388,20 @@ cpp_library(
        "utilities/transactions/lock/lock_manager.cc",
        "utilities/transactions/lock/point/point_lock_manager.cc",
        "utilities/transactions/lock/point/point_lock_tracker.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/concurrent_tree.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/keyrange.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/lock_request.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/locktree.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/manager.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/range_buffer.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/treenode.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/txnid_set.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/wfg.cc",
+        "utilities/transactions/lock/range/range_tree/lib/standalone_port.cc",
+        "utilities/transactions/lock/range/range_tree/lib/util/dbt.cc",
+        "utilities/transactions/lock/range/range_tree/lib/util/memarena.cc",
+        "utilities/transactions/lock/range/range_tree/range_tree_lock_manager.cc",
+        "utilities/transactions/lock/range/range_tree/range_tree_lock_tracker.cc",
        "utilities/transactions/optimistic_transaction.cc",
        "utilities/transactions/optimistic_transaction_db_impl.cc",
        "utilities/transactions/pessimistic_transaction.cc",
@@ -678,6 +692,20 @@ cpp_library(
        "utilities/transactions/lock/lock_manager.cc",
        "utilities/transactions/lock/point/point_lock_manager.cc",
        "utilities/transactions/lock/point/point_lock_tracker.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/concurrent_tree.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/keyrange.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/lock_request.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/locktree.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/manager.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/range_buffer.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/treenode.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/txnid_set.cc",
+        "utilities/transactions/lock/range/range_tree/lib/locktree/wfg.cc",
+        "utilities/transactions/lock/range/range_tree/lib/standalone_port.cc",
+        "utilities/transactions/lock/range/range_tree/lib/util/dbt.cc",
+        "utilities/transactions/lock/range/range_tree/lib/util/memarena.cc",
+        "utilities/transactions/lock/range/range_tree/range_tree_lock_manager.cc",
+        "utilities/transactions/lock/range/range_tree/range_tree_lock_tracker.cc",
        "utilities/transactions/optimistic_transaction.cc",
        "utilities/transactions/optimistic_transaction_db_impl.cc",
        "utilities/transactions/pessimistic_transaction.cc",
@@ -775,26 +803,24 @@ cpp_library(
    external_deps = ROCKSDB_EXTERNAL_DEPS,
 )

-if not is_opt_mode:
-    cpp_binary(
-        name = "c_test_bin",
-        srcs = ["db/c_test.c"],
-        arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
-        os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
-        compiler_flags = ROCKSDB_COMPILER_FLAGS,
-        preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-        deps = [":rocksdb_test_lib"],
-    )
+cpp_binary(
+    name = "c_test_bin",
+    srcs = ["db/c_test.c"],
+    arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
+    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
+    compiler_flags = ROCKSDB_COMPILER_FLAGS,
+    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    deps = [":rocksdb_test_lib"],
+) if not is_opt_mode else None

-if not is_opt_mode:
-    custom_unittest(
-        "c_test",
-        command = [
-            native.package_name() + "/buckifier/rocks_test_runner.sh",
-            "$(location :{})".format("c_test_bin"),
-        ],
-        type = "simple",
-    )
+custom_unittest(
+    name = "c_test",
+    command = [
+        native.package_name() + "/buckifier/rocks_test_runner.sh",
+        "$(location :{})".format("c_test_bin"),
+    ],
+    type = "simple",
+) if not is_opt_mode else None

 cpp_library(
    name = "env_basic_test_lib",
@@ -1778,6 +1804,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "range_locking_test",
+        "utilities/transactions/lock/range/range_locking_test.cc",
+        "parallel",
+        [],
+        [],
+    ],
    [
        "range_tombstone_fragmenter_test",
        "db/range_tombstone_fragmenter_test.cc",
@@ -44,7 +44,7 @@ Tango is using RocksDB as a graph storage to store all users' connection data an
 Turn is using RocksDB as a storage layer for their key/value store, serving at peak 2.4MM QPS out of different datacenters.
 Check out our RocksDB Protobuf merge operator at: https://github.com/vladb38/rocksdb_protobuf

-## Santanader UK/Cloudera Profession Services
+## Santander UK/Cloudera Profession Services
 Check out their blog post: http://blog.cloudera.com/blog/2015/08/inside-santanders-near-real-time-data-ingest-architecture/

 ## Airbnb
@@ -86,7 +86,7 @@ quasardb uses a heavily tuned RocksDB as its persistence layer.
 [Uber](http://eng.uber.com/cherami/) uses RocksDB as a durable and scalable task queue.

 ## 360 Pika
-[360](http://www.360.cn/) [Pika](https://github.com/Qihoo360/pika) is a nosql compatible with redis. With the huge amount of data stored, redis may suffer for a capacity bottleneck, and pika was born for solving it. It has widely been widely used in many company
+[360](http://www.360.cn/) [Pika](https://github.com/Qihoo360/pika) is a nosql compatible with redis. With the huge amount of data stored, redis may suffer for a capacity bottleneck, and pika was born for solving it. It has widely been used in many companies.

 ## LzLabs
 LzLabs is using RocksDB as a storage engine in their multi-database distributed framework to store application configuration and user data.
@@ -104,8 +104,11 @@ LzLabs is using RocksDB as a storage engine in their multi-database distributed
 [Crux](https://github.com/juxt/crux) is a document database that uses RocksDB for local [EAV](https://en.wikipedia.org/wiki/Entity%E2%80%93attribute%E2%80%93value_model) index storage to enable point-in-time bitemporal Datalog queries. The "unbundled" architecture uses Kafka to provide horizontal scalability.

 ## Nebula Graph
-
 [Nebula Graph](https://github.com/vesoft-inc/nebula) is a distributed, scalable, lightning-fast, open source graph database capable of hosting super large scale graphs with dozens of billions of vertices (nodes) and trillions of edges, with milliseconds of latency.

 ## YugabyteDB
 [YugabyteDB](https://www.yugabyte.com/) is an open source, high performance, distributed SQL database that uses RocksDB as its storage layer. For more information, please see https://github.com/yugabyte/yugabyte-db/.
+
+## ArangoDB
+[ArangoDB](https://www.arangodb.com/) is a native multi-model database with flexible data models for documents, graphs, and key-values, for building high performance applications using a convenient SQL-like query language or JavaScript extensions. It uses RocksDB as its sotrage engine.
+
@@ -79,26 +79,24 @@ class TARGETSBuilder(object):

    def add_c_test(self):
        self.targets_file.write(b"""
-if not is_opt_mode:
-    cpp_binary(
-        name = "c_test_bin",
-        srcs = ["db/c_test.c"],
-        arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
-        os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
-        compiler_flags = ROCKSDB_COMPILER_FLAGS,
-        preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-        deps = [":rocksdb_test_lib"],
-    )
+cpp_binary(
+    name = "c_test_bin",
+    srcs = ["db/c_test.c"],
+    arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
+    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
+    compiler_flags = ROCKSDB_COMPILER_FLAGS,
+    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    deps = [":rocksdb_test_lib"],
+) if not is_opt_mode else None

-if not is_opt_mode:
-    custom_unittest(
-        "c_test",
-        command = [
-            native.package_name() + "/buckifier/rocks_test_runner.sh",
-            "$(location :{})".format("c_test_bin"),
-        ],
-        type = "simple",
-    )
+custom_unittest(
+    name = "c_test",
+    command = [
+        native.package_name() + "/buckifier/rocks_test_runner.sh",
+        "$(location :{})".format("c_test_bin"),
+    ],
+    type = "simple",
+) if not is_opt_mode else None
 """)

    def register_test(self,
@@ -45,8 +45,13 @@ if test -z "$OUTPUT"; then
  exit 1
 fi

-# we depend on C++11
-PLATFORM_CXXFLAGS="-std=c++11"
+# we depend on C++11, but should be compatible with newer standards
+if [ "$ROCKSDB_CXX_STANDARD" ]; then
+  PLATFORM_CXXFLAGS="-std=$ROCKSDB_CXX_STANDARD"
+else
+  PLATFORM_CXXFLAGS="-std=c++11"
+fi
+
 # we currently depend on POSIX platform
 COMMON_FLAGS="-DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX"

@@ -655,6 +660,11 @@ else
    COMMON_FLAGS="$COMMON_FLAGS -mmacosx-version-min=10.12"
    PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -mmacosx-version-min=10.12"
    PLATFORM_SHARED_LDFLAGS="$PLATFORM_SHARED_LDFLAGS -mmacosx-version-min=10.12"
+    PLATFORM_CMAKE_FLAGS="-DCMAKE_OSX_DEPLOYMENT_TARGET=10.12"
+    JAVA_STATIC_DEPS_COMMON_FLAGS="-mmacosx-version-min=10.12"
+    JAVA_STATIC_DEPS_LDFLAGS="$JAVA_STATIC_DEPS_COMMON_FLAGS"
+    JAVA_STATIC_DEPS_CCFLAGS="$JAVA_STATIC_DEPS_COMMON_FLAGS"
+    JAVA_STATIC_DEPS_CXXFLAGS="$JAVA_STATIC_DEPS_COMMON_FLAGS"
  fi
 fi

@@ -812,8 +822,12 @@ echo "CXX=$CXX" >> "$OUTPUT"
 echo "AR=$AR" >> "$OUTPUT"
 echo "PLATFORM=$PLATFORM" >> "$OUTPUT"
 echo "PLATFORM_LDFLAGS=$PLATFORM_LDFLAGS" >> "$OUTPUT"
+echo "PLATFORM_CMAKE_FLAGS=$PLATFORM_CMAKE_FLAGS" >> "$OUTPUT"
 echo "JAVA_LDFLAGS=$JAVA_LDFLAGS" >> "$OUTPUT"
 echo "JAVA_STATIC_LDFLAGS=$JAVA_STATIC_LDFLAGS" >> "$OUTPUT"
+echo "JAVA_STATIC_DEPS_CCFLAGS=$JAVA_STATIC_DEPS_CCFLAGS" >> "$OUTPUT"
+echo "JAVA_STATIC_DEPS_CXXFLAGS=$JAVA_STATIC_DEPS_CXXFLAGS" >> "$OUTPUT"
+echo "JAVA_STATIC_DEPS_LDFLAGS=$JAVA_STATIC_DEPS_LDFLAGS" >> "$OUTPUT"
 echo "JAVAC_ARGS=$JAVAC_ARGS" >> "$OUTPUT"
 echo "VALGRIND_VER=$VALGRIND_VER" >> "$OUTPUT"
 echo "PLATFORM_CCFLAGS=$PLATFORM_CCFLAGS" >> "$OUTPUT"
@@ -67,7 +67,7 @@ struct LRUHandle {
    IS_HIGH_PRI = (1 << 1),
    // Whether this entry is in high-pri pool.
    IN_HIGH_PRI_POOL = (1 << 2),
-    // Wwhether this entry has had any lookups (hits).
+    // Whether this entry has had any lookups (hits).
    HAS_HIT = (1 << 3),
  };

@@ -130,7 +130,7 @@ struct LRUHandle {
    delete[] reinterpret_cast<char*>(this);
  }

-  // Caclculate the memory usage by metadata
+  // Calculate the memory usage by metadata
  inline size_t CalcTotalCharge(
      CacheMetadataChargePolicy metadata_charge_policy) {
    size_t meta_charge = 0;
@@ -30,20 +30,19 @@ Status ArenaWrappedDBIter::GetProperty(std::string prop_name,
  return db_iter_->GetProperty(prop_name, prop);
 }

-void ArenaWrappedDBIter::Init(Env* env, const ReadOptions& read_options,
-                              const ImmutableCFOptions& cf_options,
-                              const MutableCFOptions& mutable_cf_options,
-                              const SequenceNumber& sequence,
-                              uint64_t max_sequential_skip_in_iteration,
-                              uint64_t version_number,
-                              ReadCallback* read_callback, DBImpl* db_impl,
-                              ColumnFamilyData* cfd, bool allow_blob,
-                              bool allow_refresh) {
+void ArenaWrappedDBIter::Init(
+    Env* env, const ReadOptions& read_options,
+    const ImmutableCFOptions& cf_options,
+    const MutableCFOptions& mutable_cf_options, const Version* version,
+    const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iteration,
+    uint64_t version_number, ReadCallback* read_callback, DBImpl* db_impl,
+    ColumnFamilyData* cfd, bool expose_blob_index, bool allow_refresh) {
  auto mem = arena_.AllocateAligned(sizeof(DBIter));
-  db_iter_ = new (mem) DBIter(env, read_options, cf_options, mutable_cf_options,
-                              cf_options.user_comparator, nullptr, sequence,
-                              true, max_sequential_skip_in_iteration,
-                              read_callback, db_impl, cfd, allow_blob);
+  db_iter_ =
+      new (mem) DBIter(env, read_options, cf_options, mutable_cf_options,
+                       cf_options.user_comparator, /* iter */ nullptr, version,
+                       sequence, true, max_sequential_skip_in_iteration,
+                       read_callback, db_impl, cfd, expose_blob_index);
  sv_number_ = version_number;
  read_options_ = read_options;
  allow_refresh_ = allow_refresh;
@@ -72,8 +71,9 @@ Status ArenaWrappedDBIter::Refresh() {
      read_callback_->Refresh(latest_seq);
    }
    Init(env, read_options_, *(cfd_->ioptions()), sv->mutable_cf_options,
-         latest_seq, sv->mutable_cf_options.max_sequential_skip_in_iterations,
-         cur_sv_number, read_callback_, db_impl_, cfd_, allow_blob_,
+         sv->current, latest_seq,
+         sv->mutable_cf_options.max_sequential_skip_in_iterations,
+         cur_sv_number, read_callback_, db_impl_, cfd_, expose_blob_index_,
         allow_refresh_);

    InternalIterator* internal_iter = db_impl_->NewInternalIterator(
@@ -90,16 +90,16 @@ Status ArenaWrappedDBIter::Refresh() {
 ArenaWrappedDBIter* NewArenaWrappedDbIterator(
    Env* env, const ReadOptions& read_options,
    const ImmutableCFOptions& cf_options,
-    const MutableCFOptions& mutable_cf_options, const SequenceNumber& sequence,
-    uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
-    ReadCallback* read_callback, DBImpl* db_impl, ColumnFamilyData* cfd,
-    bool allow_blob, bool allow_refresh) {
+    const MutableCFOptions& mutable_cf_options, const Version* version,
+    const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iterations,
+    uint64_t version_number, ReadCallback* read_callback, DBImpl* db_impl,
+    ColumnFamilyData* cfd, bool expose_blob_index, bool allow_refresh) {
  ArenaWrappedDBIter* iter = new ArenaWrappedDBIter();
-  iter->Init(env, read_options, cf_options, mutable_cf_options, sequence,
-             max_sequential_skip_in_iterations, version_number, read_callback,
-             db_impl, cfd, allow_blob, allow_refresh);
+  iter->Init(env, read_options, cf_options, mutable_cf_options, version,
+             sequence, max_sequential_skip_in_iterations, version_number,
+             read_callback, db_impl, cfd, expose_blob_index, allow_refresh);
  if (db_impl != nullptr && cfd != nullptr && allow_refresh) {
-    iter->StoreRefreshInfo(db_impl, cfd, read_callback, allow_blob);
+    iter->StoreRefreshInfo(db_impl, cfd, read_callback, expose_blob_index);
  }

  return iter;
@@ -23,6 +23,7 @@
 namespace ROCKSDB_NAMESPACE {

 class Arena;
+class Version;

 // A wrapper iterator which wraps DB Iterator and the arena, with which the DB
 // iterator is supposed to be allocated. This class is used as an entry point of
@@ -72,20 +73,20 @@ class ArenaWrappedDBIter : public Iterator {

  void Init(Env* env, const ReadOptions& read_options,
            const ImmutableCFOptions& cf_options,
-            const MutableCFOptions& mutable_cf_options,
+            const MutableCFOptions& mutable_cf_options, const Version* version,
            const SequenceNumber& sequence,
            uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
            ReadCallback* read_callback, DBImpl* db_impl, ColumnFamilyData* cfd,
-            bool allow_blob, bool allow_refresh);
+            bool expose_blob_index, bool allow_refresh);

  // Store some parameters so we can refresh the iterator at a later point
  // with these same params
  void StoreRefreshInfo(DBImpl* db_impl, ColumnFamilyData* cfd,
-                        ReadCallback* read_callback, bool allow_blob) {
+                        ReadCallback* read_callback, bool expose_blob_index) {
    db_impl_ = db_impl;
    cfd_ = cfd;
    read_callback_ = read_callback;
-    allow_blob_ = allow_blob;
+    expose_blob_index_ = expose_blob_index;
  }

 private:
@@ -96,7 +97,7 @@ class ArenaWrappedDBIter : public Iterator {
  DBImpl* db_impl_ = nullptr;
  ReadOptions read_options_;
  ReadCallback* read_callback_;
-  bool allow_blob_ = false;
+  bool expose_blob_index_ = false;
  bool allow_refresh_ = true;
 };

@@ -106,9 +107,9 @@ class ArenaWrappedDBIter : public Iterator {
 extern ArenaWrappedDBIter* NewArenaWrappedDbIterator(
    Env* env, const ReadOptions& read_options,
    const ImmutableCFOptions& cf_options,
-    const MutableCFOptions& mutable_cf_options, const SequenceNumber& sequence,
-    uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
-    ReadCallback* read_callback, DBImpl* db_impl = nullptr,
-    ColumnFamilyData* cfd = nullptr, bool allow_blob = false,
-    bool allow_refresh = true);
+    const MutableCFOptions& mutable_cf_options, const Version* version,
+    const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iterations,
+    uint64_t version_number, ReadCallback* read_callback,
+    DBImpl* db_impl = nullptr, ColumnFamilyData* cfd = nullptr,
+    bool expose_blob_index = false, bool allow_refresh = true);
 }  // namespace ROCKSDB_NAMESPACE
@@ -3,6 +3,8 @@
 //  COPYING file in the root directory) and Apache 2.0 License
 //  (found in the LICENSE.Apache file in the root directory).

+#include <array>
+
 #include "db/blob/blob_index.h"
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
@@ -44,6 +46,85 @@ TEST_F(DBBlobBasicTest, GetBlob) {
                  .IsIncomplete());
 }

+TEST_F(DBBlobBasicTest, MultiGetBlobs) {
+  constexpr size_t min_blob_size = 6;
+
+  Options options = GetDefaultOptions();
+  options.enable_blob_files = true;
+  options.min_blob_size = min_blob_size;
+
+  Reopen(options);
+
+  // Put then retrieve three key-values. The first value is below the size limit
+  // and is thus stored inline; the other two are stored separately as blobs.
+  constexpr size_t num_keys = 3;
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "short";
+  static_assert(sizeof(first_value) - 1 < min_blob_size,
+                "first_value too long to be inlined");
+
+  ASSERT_OK(Put(first_key, first_value));
+
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "long_value";
+  static_assert(sizeof(second_value) - 1 >= min_blob_size,
+                "second_value too short to be stored as blob");
+
+  ASSERT_OK(Put(second_key, second_value));
+
+  constexpr char third_key[] = "third_key";
+  constexpr char third_value[] = "other_long_value";
+  static_assert(sizeof(third_value) - 1 >= min_blob_size,
+                "third_value too short to be stored as blob");
+
+  ASSERT_OK(Put(third_key, third_value));
+
+  ASSERT_OK(Flush());
+
+  ReadOptions read_options;
+
+  std::array<Slice, num_keys> keys{{first_key, second_key, third_key}};
+
+  {
+    std::array<PinnableSlice, num_keys> values;
+    std::array<Status, num_keys> statuses;
+
+    db_->MultiGet(read_options, db_->DefaultColumnFamily(), num_keys, &keys[0],
+                  &values[0], &statuses[0]);
+
+    ASSERT_OK(statuses[0]);
+    ASSERT_EQ(values[0], first_value);
+
+    ASSERT_OK(statuses[1]);
+    ASSERT_EQ(values[1], second_value);
+
+    ASSERT_OK(statuses[2]);
+    ASSERT_EQ(values[2], third_value);
+  }
+
+  // Try again with no I/O allowed. The table and the necessary blocks should
+  // already be in their respective caches. The first (inlined) value should be
+  // successfully read; however, the two blob values could only be read from the
+  // blob file, so for those the read should return Incomplete.
+  read_options.read_tier = kBlockCacheTier;
+
+  {
+    std::array<PinnableSlice, num_keys> values;
+    std::array<Status, num_keys> statuses;
+
+    db_->MultiGet(read_options, db_->DefaultColumnFamily(), num_keys, &keys[0],
+                  &values[0], &statuses[0]);
+
+    ASSERT_OK(statuses[0]);
+    ASSERT_EQ(values[0], first_value);
+
+    ASSERT_TRUE(statuses[1].IsIncomplete());
+
+    ASSERT_TRUE(statuses[2].IsIncomplete());
+  }
+}
+
 TEST_F(DBBlobBasicTest, GetBlob_CorruptIndex) {
  Options options = GetDefaultOptions();
  options.enable_blob_files = true;
@@ -175,6 +256,48 @@ TEST_P(DBBlobBasicIOErrorTest, GetBlob_IOError) {
  SyncPoint::GetInstance()->ClearAllCallBacks();
 }

+TEST_P(DBBlobBasicIOErrorTest, MultiGetBlobs_IOError) {
+  Options options = GetDefaultOptions();
+  options.env = fault_injection_env_.get();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+
+  Reopen(options);
+
+  constexpr size_t num_keys = 2;
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "first_value";
+
+  ASSERT_OK(Put(first_key, first_value));
+
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "second_value";
+
+  ASSERT_OK(Put(second_key, second_value));
+
+  ASSERT_OK(Flush());
+
+  std::array<Slice, num_keys> keys{{first_key, second_key}};
+  std::array<PinnableSlice, num_keys> values;
+  std::array<Status, num_keys> statuses;
+
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
+    fault_injection_env_->SetFilesystemActive(false,
+                                              Status::IOError(sync_point_));
+  });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  db_->MultiGet(ReadOptions(), db_->DefaultColumnFamily(), num_keys, &keys[0],
+                &values[0], &statuses[0]);
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+
+  ASSERT_TRUE(statuses[0].IsIOError());
+  ASSERT_TRUE(statuses[1].IsIOError());
+}
+
 }  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
@@ -98,7 +98,7 @@ class DBBlobIndexTest : public DBTestBase {
  ArenaWrappedDBIter* GetBlobIterator() {
    return dbfull()->NewIteratorImpl(
        ReadOptions(), cfd(), dbfull()->GetLatestSequenceNumber(),
-        nullptr /*read_callback*/, true /*allow_blob*/);
+        nullptr /*read_callback*/, true /*expose_blob_index*/);
  }

  Options GetTestOptions() {
@@ -238,8 +238,11 @@ TEST_F(DBBlobIndexTest, Updated) {
  }
 }

-// Iterator should get blob value if allow_blob flag is set,
-// otherwise return Status::NotSupported status.
+// Note: the following test case pertains to the StackableDB-based BlobDB
+// implementation. When a blob iterator is used, it should set the
+// expose_blob_index flag for the underlying DBIter, and retrieve/return the
+// corresponding blob value. If a regular DBIter is created (i.e.
+// expose_blob_index is not set), it should return Status::Corruption.
 TEST_F(DBBlobIndexTest, Iterate) {
  const std::vector<std::vector<ValueType>> data = {
      /*00*/ {kTypeValue},
@@ -302,6 +305,7 @@ TEST_F(DBBlobIndexTest, Iterate) {
                    std::function<void(Iterator*)> extra_check = nullptr) {
    // Seek
    auto* iterator = create_iterator();
+    ASSERT_OK(iterator->status());
    ASSERT_OK(iterator->Refresh());
    iterator->Seek(get_key(index));
    check_iterator(iterator, expected_status, forward_value);
@@ -315,6 +319,7 @@ TEST_F(DBBlobIndexTest, Iterate) {
    ASSERT_OK(iterator->Refresh());
    iterator->Seek(get_key(index - 1));
    ASSERT_TRUE(iterator->Valid());
+    ASSERT_OK(iterator->status());
    iterator->Next();
    check_iterator(iterator, expected_status, forward_value);
    if (extra_check) {
@@ -324,6 +329,7 @@ TEST_F(DBBlobIndexTest, Iterate) {

    // SeekForPrev
    iterator = create_iterator();
+    ASSERT_OK(iterator->status());
    ASSERT_OK(iterator->Refresh());
    iterator->SeekForPrev(get_key(index));
    check_iterator(iterator, expected_status, backward_value);
@@ -336,6 +342,7 @@ TEST_F(DBBlobIndexTest, Iterate) {
    iterator = create_iterator();
    iterator->Seek(get_key(index + 1));
    ASSERT_TRUE(iterator->Valid());
+    ASSERT_OK(iterator->status());
    iterator->Prev();
    check_iterator(iterator, expected_status, backward_value);
    if (extra_check) {
@@ -373,7 +380,7 @@ TEST_F(DBBlobIndexTest, Iterate) {
            ASSERT_OK(Write(&batch));
            break;
          default:
-            assert(false);
+            FAIL();
        };
      }
      snapshots.push_back(dbfull()->GetSnapshot());
@@ -384,8 +391,8 @@ TEST_F(DBBlobIndexTest, Iterate) {
    MoveDataTo(tier);

    // Normal iterator
-    verify(1, Status::kNotSupported, "", "", create_normal_iterator);
-    verify(3, Status::kNotSupported, "", "", create_normal_iterator);
+    verify(1, Status::kCorruption, "", "", create_normal_iterator);
+    verify(3, Status::kCorruption, "", "", create_normal_iterator);
    verify(5, Status::kOk, get_value(5, 0), get_value(5, 0),
           create_normal_iterator);
    verify(7, Status::kOk, get_value(8, 0), get_value(6, 0),
@@ -335,7 +335,9 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
  // was not used)
  auto sfm = static_cast<SstFileManagerImpl*>(db_options.sst_file_manager.get());
  for (size_t i = 0; i < result.cf_paths.size(); i++) {
-    DeleteScheduler::CleanupDirectory(db_options.env, sfm, result.cf_paths[i].path);
+    DeleteScheduler::CleanupDirectory(db_options.env, sfm,
+                                      result.cf_paths[i].path)
+        .PermitUncheckedError();
  }
 #endif

@@ -449,9 +451,7 @@ void SuperVersion::Cleanup() {
    to_delete.push_back(m);
  }
  current->Unref();
-  if (cfd->Unref()) {
-    delete cfd;
-  }
+  cfd->UnrefAndTryDelete(this);
 }

 void SuperVersion::Init(ColumnFamilyData* new_cfd, MemTable* new_mem,
@@ -660,7 +660,7 @@ ColumnFamilyData::~ColumnFamilyData() {
  }
 }

-bool ColumnFamilyData::UnrefAndTryDelete() {
+bool ColumnFamilyData::UnrefAndTryDelete(SuperVersion* sv_under_cleanup) {
  int old_refs = refs_.fetch_sub(1);
  assert(old_refs > 0);

@@ -670,7 +670,11 @@ bool ColumnFamilyData::UnrefAndTryDelete() {
    return true;
  }

-  if (old_refs == 2 && super_version_ != nullptr) {
+  // If called under SuperVersion::Cleanup, we should not re-enter Cleanup on
+  // the same SuperVersion. (But while installing a new SuperVersion, this
+  // cfd could be referenced only by two SuperVersions.)
+  if (old_refs == 2 && super_version_ != nullptr &&
+      super_version_ != sv_under_cleanup) {
    // Only the super_version_ holds me
    SuperVersion* sv = super_version_;
    super_version_ = nullptr;
@@ -708,9 +712,7 @@ uint64_t ColumnFamilyData::OldestLogToKeep() {
  auto current_log = GetLogNumber();

  if (allow_2pc_) {
-    autovector<MemTable*> empty_list;
-    auto imm_prep_log =
-        imm()->PrecomputeMinLogContainingPrepSection(empty_list);
+    auto imm_prep_log = imm()->PrecomputeMinLogContainingPrepSection();
    auto mem_prep_log = mem()->GetMinLogContainingPrepSection();

    if (imm_prep_log > 0 && imm_prep_log < current_log) {
@@ -1339,6 +1341,15 @@ Status ColumnFamilyData::ValidateOptions(
          "Block-Based Table format. ");
    }
  }
+
+  if (cf_options.enable_blob_garbage_collection &&
+      (cf_options.blob_garbage_collection_age_cutoff < 0.0 ||
+       cf_options.blob_garbage_collection_age_cutoff > 1.0)) {
+    return Status::InvalidArgument(
+        "The age cutoff for blob garbage collection should be in the range "
+        "[0.0, 1.0].");
+  }
+
  return s;
 }

@@ -279,21 +279,11 @@ class ColumnFamilyData {
  // holding a DB mutex, or as the leader in a write batch group).
  void Ref() { refs_.fetch_add(1); }

-  // Unref decreases the reference count, but does not handle deletion
-  // when the count goes to 0.  If this method returns true then the
-  // caller should delete the instance immediately, or later, by calling
-  // FreeDeadColumnFamilies().  Unref() can only be called while holding
-  // a DB mutex, or during single-threaded recovery.
-  bool Unref() {
-    int old_refs = refs_.fetch_sub(1);
-    assert(old_refs > 0);
-    return old_refs == 1;
-  }
-
  // UnrefAndTryDelete() decreases the reference count and do free if needed,
  // return true if this is freed else false, UnrefAndTryDelete() can only
  // be called while holding a DB mutex, or during single-threaded recovery.
-  bool UnrefAndTryDelete();
+  // sv_under_cleanup is only provided when called from SuperVersion::Cleanup.
+  bool UnrefAndTryDelete(SuperVersion* sv_under_cleanup = nullptr);

  // SetDropped() can only be called under following conditions:
  // 1) Holding a DB mutex,
@@ -360,6 +350,11 @@ class ColumnFamilyData {

  MemTableList* imm() { return &imm_; }
  MemTable* mem() { return mem_; }
+
+  bool IsEmpty() {
+    return mem()->GetFirstSequenceNumber() == 0 && imm()->NumNotFlushed() == 0;
+  }
+
  Version* current() { return current_; }
  Version* dummy_versions() { return dummy_versions_; }
  void SetCurrent(Version* _current);
@@ -509,6 +504,21 @@ class ColumnFamilyData {

  FSDirectory* GetDataDir(size_t path_id) const;

+  // full_history_ts_low_ can only increase.
+  void SetFullHistoryTsLow(std::string ts_low) {
+    assert(!ts_low.empty());
+    const Comparator* ucmp = user_comparator();
+    assert(ucmp);
+    if (full_history_ts_low_.empty() ||
+        ucmp->CompareTimestamp(ts_low, full_history_ts_low_) > 0) {
+      full_history_ts_low_ = std::move(ts_low);
+    }
+  }
+
+  const std::string& GetFullHistoryTsLow() const {
+    return full_history_ts_low_;
+  }
+
  ThreadLocalPtr* TEST_GetLocalSV() { return local_sv_.get(); }

 private:
@@ -605,6 +615,8 @@ class ColumnFamilyData {
  std::vector<std::shared_ptr<FSDirectory>> data_dirs_;

  bool db_paths_registered_;
+
+  std::string full_history_ts_low_;
 };

 // ColumnFamilySet has interesting thread-safety requirements
@@ -73,7 +73,7 @@ class ColumnFamilyTestBase : public testing::Test {
    db_options_.create_if_missing = true;
    db_options_.fail_if_options_file_error = true;
    db_options_.env = env_;
-    DestroyDB(dbname_, Options(db_options_, column_family_options_));
+    EXPECT_OK(DestroyDB(dbname_, Options(db_options_, column_family_options_)));
  }

  ~ColumnFamilyTestBase() override {
@@ -653,8 +653,8 @@ TEST_P(FlushEmptyCFTestWithParam, FlushEmptyCFTest) {
  // after flushing file B is deleted. At the same time, the min log number of
  // default CF is not written to manifest. Log file A still remains.
  // Flushed to SST file Y.
-  Flush(1);
-  Flush(0);
+  ASSERT_OK(Flush(1));
+  ASSERT_OK(Flush(0));
  ASSERT_OK(Put(1, "bar", "v3"));  // seqID 4
  ASSERT_OK(Put(1, "foo", "v4"));  // seqID 5
  ASSERT_OK(db_->FlushWAL(/*sync=*/false));
@@ -708,15 +708,15 @@ TEST_P(FlushEmptyCFTestWithParam, FlushEmptyCFTest2) {
  // and is set to current. Both CFs' min log number is set to file C so after
  // flushing file B is deleted. Log file A still remains.
  // Flushed to SST file Y.
-  Flush(1);
+  ASSERT_OK(Flush(1));
  ASSERT_OK(Put(0, "bar", "v2"));  // seqID 4
  ASSERT_OK(Put(2, "bar", "v2"));  // seqID 5
  ASSERT_OK(Put(1, "bar", "v3"));  // seqID 6
  // Flushing all column families. This forces all CFs' min log to current. This
  // is written to the manifest file. Log file C is cleared.
-  Flush(0);
-  Flush(1);
-  Flush(2);
+  ASSERT_OK(Flush(0));
+  ASSERT_OK(Flush(1));
+  ASSERT_OK(Flush(2));
  // Write to log file D
  ASSERT_OK(Put(1, "bar", "v4"));  // seqID 7
  ASSERT_OK(Put(1, "bar", "v5"));  // seqID 8
@@ -985,7 +985,7 @@ TEST_P(ColumnFamilyTest, FlushTest) {
    for (int i = 0; i < 3; ++i) {
      uint64_t max_total_in_memory_state =
          MaxTotalInMemoryState();
-      Flush(i);
+      ASSERT_OK(Flush(i));
      AssertMaxTotalInMemoryState(max_total_in_memory_state);
    }
    ASSERT_OK(Put(1, "foofoo", "bar"));
@@ -1093,7 +1093,7 @@ TEST_P(ColumnFamilyTest, CrashAfterFlush) {
  ASSERT_OK(batch.Put(handles_[0], Slice("foo"), Slice("bar")));
  ASSERT_OK(batch.Put(handles_[1], Slice("foo"), Slice("bar")));
  ASSERT_OK(db_->Write(WriteOptions(), &batch));
-  Flush(0);
+  ASSERT_OK(Flush(0));
  fault_env->SetFilesystemActive(false);

  std::vector<std::string> names;
@@ -1103,7 +1103,7 @@ TEST_P(ColumnFamilyTest, CrashAfterFlush) {
    }
  }
  Close();
-  fault_env->DropUnsyncedFileData();
+  ASSERT_OK(fault_env->DropUnsyncedFileData());
  fault_env->ResetState();
  Open(names, {});

@@ -2236,7 +2236,7 @@ TEST_P(ColumnFamilyTest, FlushStaleColumnFamilies) {
  // files for column family [one], because it's empty
  AssertCountLiveFiles(4);

-  Flush(0);
+  ASSERT_OK(Flush(0));
  ASSERT_EQ(0, dbfull()->TEST_total_log_size());
  Close();
 }
@@ -3040,7 +3040,7 @@ TEST_P(ColumnFamilyTest, IteratorCloseWALFile1) {
  Iterator* it = db_->NewIterator(ReadOptions(), handles_[1]);
  ASSERT_OK(it->status());
  // A flush will make `it` hold the last reference of its super version.
-  Flush(1);
+  ASSERT_OK(Flush(1));

  ASSERT_OK(Put(1, "fodor", "mirko"));
  ASSERT_OK(Put(0, "fodor", "mirko"));
@@ -3093,7 +3093,7 @@ TEST_P(ColumnFamilyTest, IteratorCloseWALFile2) {
  Iterator* it = db_->NewIterator(ro, handles_[1]);
  ASSERT_OK(it->status());
  // A flush will make `it` hold the last reference of its super version.
-  Flush(1);
+  ASSERT_OK(Flush(1));

  ASSERT_OK(Put(1, "fodor", "mirko"));
  ASSERT_OK(Put(0, "fodor", "mirko"));
@@ -3147,7 +3147,7 @@ TEST_P(ColumnFamilyTest, ForwardIteratorCloseWALFile) {
  CreateColumnFamilies({"one"});
  ASSERT_OK(Put(1, "fodor", "mirko"));
  ASSERT_OK(Put(1, "fodar2", "mirko"));
-  Flush(1);
+  ASSERT_OK(Flush(1));

  // Create an iterator holding the current super version, as well as
  // the SST file just flushed.
@@ -3159,7 +3159,7 @@ TEST_P(ColumnFamilyTest, ForwardIteratorCloseWALFile) {

  ASSERT_OK(Put(1, "fodor", "mirko"));
  ASSERT_OK(Put(1, "fodar2", "mirko"));
-  Flush(1);
+  ASSERT_OK(Flush(1));

  WaitForCompaction();

@@ -3232,9 +3232,9 @@ TEST_P(ColumnFamilyTest, LogSyncConflictFlush) {
  ROCKSDB_NAMESPACE::port::Thread thread([&] { ASSERT_OK(db_->SyncWAL()); });

  TEST_SYNC_POINT("ColumnFamilyTest::LogSyncConflictFlush:1");
-  Flush(1);
+  ASSERT_OK(Flush(1));
  ASSERT_OK(Put(1, "foo", "bar"));
-  Flush(1);
+  ASSERT_OK(Flush(1));

  TEST_SYNC_POINT("ColumnFamilyTest::LogSyncConflictFlush:2");

@@ -3256,7 +3256,7 @@ TEST_P(ColumnFamilyTest, DISABLED_LogTruncationTest) {
  Build(0, 100);

  // Flush the 0th column family to force a roll of the wal log
-  Flush(0);
+  ASSERT_OK(Flush(0));

  // Add some more entries
  Build(100, 100);
@@ -3332,14 +3332,14 @@ TEST_P(ColumnFamilyTest, DefaultCfPathsTest) {

  // Fill Column family 1.
  PutRandomData(1, 100, 100);
-  Flush(1);
+  ASSERT_OK(Flush(1));

  ASSERT_EQ(1, GetSstFileCount(cf_opt1.cf_paths[0].path));
  ASSERT_EQ(0, GetSstFileCount(dbname_));

  // Fill column family 2
  PutRandomData(2, 100, 100);
-  Flush(2);
+  ASSERT_OK(Flush(2));

  // SST from Column family 2 should be generated in
  // db_paths which is dbname_ in this case.
@@ -3358,14 +3358,14 @@ TEST_P(ColumnFamilyTest, MultipleCFPathsTest) {
  Reopen({ColumnFamilyOptions(), cf_opt1, cf_opt2});

  PutRandomData(1, 100, 100, true /* save */);
-  Flush(1);
+  ASSERT_OK(Flush(1));

  // Check that files are generated in appropriate paths.
  ASSERT_EQ(1, GetSstFileCount(cf_opt1.cf_paths[0].path));
  ASSERT_EQ(0, GetSstFileCount(dbname_));

  PutRandomData(2, 100, 100, true /* save */);
-  Flush(2);
+  ASSERT_OK(Flush(2));

  ASSERT_EQ(1, GetSstFileCount(cf_opt2.cf_paths[0].path));
  ASSERT_EQ(0, GetSstFileCount(dbname_));
@@ -3391,6 +3391,30 @@ TEST_P(ColumnFamilyTest, MultipleCFPathsTest) {
  }
 }

+TEST(ColumnFamilyTest, ValidateBlobGCCutoff) {
+  DBOptions db_options;
+
+  ColumnFamilyOptions cf_options;
+  cf_options.enable_blob_garbage_collection = true;
+
+  cf_options.blob_garbage_collection_age_cutoff = -0.5;
+  ASSERT_TRUE(ColumnFamilyData::ValidateOptions(db_options, cf_options)
+                  .IsInvalidArgument());
+
+  cf_options.blob_garbage_collection_age_cutoff = 0.0;
+  ASSERT_OK(ColumnFamilyData::ValidateOptions(db_options, cf_options));
+
+  cf_options.blob_garbage_collection_age_cutoff = 0.5;
+  ASSERT_OK(ColumnFamilyData::ValidateOptions(db_options, cf_options));
+
+  cf_options.blob_garbage_collection_age_cutoff = 1.0;
+  ASSERT_OK(ColumnFamilyData::ValidateOptions(db_options, cf_options));
+
+  cf_options.blob_garbage_collection_age_cutoff = 1.5;
+  ASSERT_TRUE(ColumnFamilyData::ValidateOptions(db_options, cf_options)
+                  .IsInvalidArgument());
+}
+
 }  // namespace ROCKSDB_NAMESPACE

 #ifdef ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS
@@ -91,9 +91,9 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
  // create couple files
  // Background compaction starts and waits in BackgroundCallCompaction:0
  for (int i = 0; i < kLevel0Trigger * 4; ++i) {
-    db->Put(WriteOptions(), ToString(i), "");
-    db->Put(WriteOptions(), ToString(100 - i), "");
-    db->Flush(FlushOptions());
+    ASSERT_OK(db->Put(WriteOptions(), ToString(i), ""));
+    ASSERT_OK(db->Put(WriteOptions(), ToString(100 - i), ""));
+    ASSERT_OK(db->Flush(FlushOptions()));
  }

  ROCKSDB_NAMESPACE::ColumnFamilyMetaData meta;
@@ -138,18 +138,18 @@ TEST_F(CompactFilesTest, ObsoleteFiles) {
  DB* db = nullptr;
  DestroyDB(db_name_, options);
  Status s = DB::Open(options, db_name_, &db);
-  assert(s.ok());
-  assert(db);
+  ASSERT_OK(s);
+  ASSERT_NE(db, nullptr);

  // create couple files
  for (int i = 1000; i < 2000; ++i) {
-    db->Put(WriteOptions(), ToString(i),
-            std::string(kWriteBufferSize / 10, 'a' + (i % 26)));
+    ASSERT_OK(db->Put(WriteOptions(), ToString(i),
+                      std::string(kWriteBufferSize / 10, 'a' + (i % 26))));
  }

  auto l0_files = collector->GetFlushedFiles();
  ASSERT_OK(db->CompactFiles(CompactionOptions(), l0_files, 1));
-  static_cast_with_check<DBImpl>(db)->TEST_WaitForCompact();
+  ASSERT_OK(static_cast_with_check<DBImpl>(db)->TEST_WaitForCompact());

  // verify all compaction input files are deleted
  for (auto fname : l0_files) {
@@ -182,15 +182,17 @@ TEST_F(CompactFilesTest, NotCutOutputOnLevel0) {

  // create couple files
  for (int i = 0; i < 500; ++i) {
-    db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
+    ASSERT_OK(db->Put(WriteOptions(), ToString(i),
+                      std::string(1000, 'a' + (i % 26))));
  }
-  static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable();
+  ASSERT_OK(static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable());
  auto l0_files_1 = collector->GetFlushedFiles();
  collector->ClearFlushedFiles();
  for (int i = 0; i < 500; ++i) {
-    db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
+    ASSERT_OK(db->Put(WriteOptions(), ToString(i),
+                      std::string(1000, 'a' + (i % 26))));
  }
-  static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable();
+  ASSERT_OK(static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable());
  auto l0_files_2 = collector->GetFlushedFiles();
  ASSERT_OK(db->CompactFiles(CompactionOptions(), l0_files_1, 0));
  ASSERT_OK(db->CompactFiles(CompactionOptions(), l0_files_2, 0));
@@ -213,13 +215,13 @@ TEST_F(CompactFilesTest, CapturingPendingFiles) {
  DB* db = nullptr;
  DestroyDB(db_name_, options);
  Status s = DB::Open(options, db_name_, &db);
-  assert(s.ok());
+  ASSERT_OK(s);
  assert(db);

  // Create 5 files.
  for (int i = 0; i < 5; ++i) {
-    db->Put(WriteOptions(), "key" + ToString(i), "value");
-    db->Flush(FlushOptions());
+    ASSERT_OK(db->Put(WriteOptions(), "key" + ToString(i), "value"));
+    ASSERT_OK(db->Flush(FlushOptions()));
  }

  auto l0_files = collector->GetFlushedFiles();
@@ -237,8 +239,8 @@ TEST_F(CompactFilesTest, CapturingPendingFiles) {

  // In the meantime flush another file.
  TEST_SYNC_POINT("CompactFilesTest.CapturingPendingFiles:0");
-  db->Put(WriteOptions(), "key5", "value");
-  db->Flush(FlushOptions());
+  ASSERT_OK(db->Put(WriteOptions(), "key5", "value"));
+  ASSERT_OK(db->Flush(FlushOptions()));
  TEST_SYNC_POINT("CompactFilesTest.CapturingPendingFiles:1");

  compaction_thread.join();
@@ -249,7 +251,7 @@ TEST_F(CompactFilesTest, CapturingPendingFiles) {

  // Make sure we can reopen the DB.
  s = DB::Open(options, db_name_, &db);
-  ASSERT_TRUE(s.ok());
+  ASSERT_OK(s);
  assert(db);
  delete db;
 }
@@ -293,8 +295,8 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
  cf->SetDB(db);

  // Write one L0 file
-  db->Put(WriteOptions(), "K1", "V1");
-  db->Flush(FlushOptions());
+  ASSERT_OK(db->Put(WriteOptions(), "K1", "V1"));
+  ASSERT_OK(db->Flush(FlushOptions()));

  // Compact all L0 files using CompactFiles
  ROCKSDB_NAMESPACE::ColumnFamilyMetaData meta;
@@ -337,8 +339,8 @@ TEST_F(CompactFilesTest, SentinelCompressionType) {
    DB* db = nullptr;
    ASSERT_OK(DB::Open(options, db_name_, &db));

-    db->Put(WriteOptions(), "key", "val");
-    db->Flush(FlushOptions());
+    ASSERT_OK(db->Put(WriteOptions(), "key", "val"));
+    ASSERT_OK(db->Flush(FlushOptions()));

    auto l0_files = collector->GetFlushedFiles();
    ASSERT_EQ(1, l0_files.size());
@@ -377,14 +379,15 @@ TEST_F(CompactFilesTest, GetCompactionJobInfo) {
  DB* db = nullptr;
  DestroyDB(db_name_, options);
  Status s = DB::Open(options, db_name_, &db);
-  assert(s.ok());
+  ASSERT_OK(s);
  assert(db);

  // create couple files
  for (int i = 0; i < 500; ++i) {
-    db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
+    ASSERT_OK(db->Put(WriteOptions(), ToString(i),
+                      std::string(1000, 'a' + (i % 26))));
  }
-  static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable();
+  ASSERT_OK(static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable());
  auto l0_files_1 = collector->GetFlushedFiles();
  CompactionOptions co;
  co.compression = CompressionType::kLZ4Compression;
@@ -82,6 +82,11 @@ class CompactedDBImpl : public DBImpl {
                       ColumnFamilyHandle* /*column_family*/) override {
    return Status::NotSupported("Not supported in compacted db mode.");
  }
+
+  virtual Status SyncWAL() override {
+    return Status::NotSupported("Not supported in compacted db mode.");
+  }
+
  using DB::IngestExternalFile;
  virtual Status IngestExternalFile(
      ColumnFamilyHandle* /*column_family*/,
@@ -5,9 +5,11 @@

 #include "db/compaction/compaction_iterator.h"

-#include <cinttypes>
+#include <iterator>
+#include <limits>

 #include "db/blob/blob_file_builder.h"
+#include "db/blob/blob_index.h"
 #include "db/snapshot_checker.h"
 #include "port/likely.h"
 #include "rocksdb/listener.h"
@@ -94,6 +96,8 @@ CompactionIterator::CompactionIterator(
      current_user_key_sequence_(0),
      current_user_key_snapshot_(0),
      merge_out_iter_(merge_helper_),
+      blob_garbage_collection_cutoff_file_number_(
+          ComputeBlobGarbageCollectionCutoffFileNumber(compaction_.get())),
      current_key_committed_(false),
      cmp_with_history_ts_low_(0) {
  assert(compaction_filter_ == nullptr || compaction_ != nullptr);
@@ -733,40 +737,136 @@ void CompactionIterator::NextFromInput() {
  }
 }

+bool CompactionIterator::ExtractLargeValueIfNeededImpl() {
+  if (!blob_file_builder_) {
+    return false;
+  }
+
+  blob_index_.clear();
+  const Status s = blob_file_builder_->Add(user_key(), value_, &blob_index_);
+
+  if (!s.ok()) {
+    status_ = s;
+    valid_ = false;
+
+    return false;
+  }
+
+  if (blob_index_.empty()) {
+    return false;
+  }
+
+  value_ = blob_index_;
+
+  return true;
+}
+
+void CompactionIterator::ExtractLargeValueIfNeeded() {
+  assert(ikey_.type == kTypeValue);
+
+  if (!ExtractLargeValueIfNeededImpl()) {
+    return;
+  }
+
+  ikey_.type = kTypeBlobIndex;
+  current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
+}
+
+void CompactionIterator::GarbageCollectBlobIfNeeded() {
+  assert(ikey_.type == kTypeBlobIndex);
+
+  if (!compaction_) {
+    return;
+  }
+
+  // GC for integrated BlobDB
+  if (compaction_->enable_blob_garbage_collection()) {
+    BlobIndex blob_index;
+
+    {
+      const Status s = blob_index.DecodeFrom(value_);
+
+      if (!s.ok()) {
+        status_ = s;
+        valid_ = false;
+
+        return;
+      }
+    }
+
+    if (blob_index.IsInlined() || blob_index.HasTTL()) {
+      status_ = Status::Corruption("Unexpected TTL/inlined blob index");
+      valid_ = false;
+
+      return;
+    }
+
+    if (blob_index.file_number() >=
+        blob_garbage_collection_cutoff_file_number_) {
+      return;
+    }
+
+    const Version* const version = compaction_->input_version();
+    assert(version);
+
+    {
+      const Status s =
+          version->GetBlob(ReadOptions(), user_key(), blob_index, &blob_value_);
+
+      if (!s.ok()) {
+        status_ = s;
+        valid_ = false;
+
+        return;
+      }
+    }
+
+    value_ = blob_value_;
+
+    if (ExtractLargeValueIfNeededImpl()) {
+      return;
+    }
+
+    ikey_.type = kTypeValue;
+    current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
+
+    return;
+  }
+
+  // GC for stacked BlobDB
+  if (compaction_filter_) {
+    const auto blob_decision = compaction_filter_->PrepareBlobOutput(
+        user_key(), value_, &compaction_filter_value_);
+
+    if (blob_decision == CompactionFilter::BlobDecision::kCorruption) {
+      status_ =
+          Status::Corruption("Corrupted blob reference encountered during GC");
+      valid_ = false;
+
+      return;
+    }
+
+    if (blob_decision == CompactionFilter::BlobDecision::kIOError) {
+      status_ = Status::IOError("Could not relocate blob during GC");
+      valid_ = false;
+
+      return;
+    }
+
+    if (blob_decision == CompactionFilter::BlobDecision::kChangeValue) {
+      value_ = compaction_filter_value_;
+
+      return;
+    }
+  }
+}
+
 void CompactionIterator::PrepareOutput() {
  if (valid_) {
    if (ikey_.type == kTypeValue) {
-      if (blob_file_builder_) {
-        blob_index_.clear();
-        const Status s =
-            blob_file_builder_->Add(user_key(), value_, &blob_index_);
-
-        if (!s.ok()) {
-          status_ = s;
-          valid_ = false;
-        } else if (!blob_index_.empty()) {
-          value_ = blob_index_;
-          ikey_.type = kTypeBlobIndex;
-          current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-        }
-      }
+      ExtractLargeValueIfNeeded();
    } else if (ikey_.type == kTypeBlobIndex) {
-      if (compaction_filter_) {
-        const auto blob_decision = compaction_filter_->PrepareBlobOutput(
-            user_key(), value_, &compaction_filter_value_);
-
-        if (blob_decision == CompactionFilter::BlobDecision::kCorruption) {
-          status_ = Status::Corruption(
-              "Corrupted blob reference encountered during GC");
-          valid_ = false;
-        } else if (blob_decision == CompactionFilter::BlobDecision::kIOError) {
-          status_ = Status::IOError("Could not relocate blob during GC");
-          valid_ = false;
-        } else if (blob_decision ==
-                   CompactionFilter::BlobDecision::kChangeValue) {
-          value_ = compaction_filter_value_;
-        }
-      }
+      GarbageCollectBlobIfNeeded();
    }

    // Zeroing out the sequence number leads to better compression.
@@ -894,4 +994,30 @@ bool CompactionIterator::IsInEarliestSnapshot(SequenceNumber sequence) {
  return in_snapshot == SnapshotCheckerResult::kInSnapshot;
 }

+uint64_t CompactionIterator::ComputeBlobGarbageCollectionCutoffFileNumber(
+    const CompactionProxy* compaction) {
+  if (!compaction) {
+    return 0;
+  }
+
+  if (!compaction->enable_blob_garbage_collection()) {
+    return 0;
+  }
+
+  Version* const version = compaction->input_version();
+  assert(version);
+
+  const VersionStorageInfo* const storage_info = version->storage_info();
+  assert(storage_info);
+
+  const auto& blob_files = storage_info->GetBlobFiles();
+
+  auto it = blob_files.begin();
+  std::advance(
+      it, compaction->blob_garbage_collection_age_cutoff() * blob_files.size());
+
+  return it != blob_files.end() ? it->first
+                                : std::numeric_limits<uint64_t>::max();
+}
+
 }  // namespace ROCKSDB_NAMESPACE
@@ -5,6 +5,7 @@
 #pragma once

 #include <algorithm>
+#include <cinttypes>
 #include <deque>
 #include <string>
 #include <unordered_set>
@@ -45,6 +46,12 @@ class CompactionIterator {
    virtual bool allow_ingest_behind() const = 0;

    virtual bool preserve_deletes() const = 0;
+
+    virtual bool enable_blob_garbage_collection() const = 0;
+
+    virtual double blob_garbage_collection_age_cutoff() const = 0;
+
+    virtual Version* input_version() const = 0;
  };

  class RealCompaction : public CompactionProxy {
@@ -53,6 +60,7 @@ class CompactionIterator {
        : compaction_(compaction) {
      assert(compaction_);
      assert(compaction_->immutable_cf_options());
+      assert(compaction_->mutable_cf_options());
    }

    int level() const override { return compaction_->level(); }
@@ -80,6 +88,19 @@ class CompactionIterator {
      return compaction_->immutable_cf_options()->preserve_deletes;
    }

+    bool enable_blob_garbage_collection() const override {
+      return compaction_->mutable_cf_options()->enable_blob_garbage_collection;
+    }
+
+    double blob_garbage_collection_age_cutoff() const override {
+      return compaction_->mutable_cf_options()
+          ->blob_garbage_collection_age_cutoff;
+    }
+
+    Version* input_version() const override {
+      return compaction_->input_version();
+    }
+
   private:
    const Compaction* compaction_;
  };
@@ -146,11 +167,30 @@ class CompactionIterator {
  // Processes the input stream to find the next output
  void NextFromInput();

-  // Do last preparations before presenting the output to the callee. At this
-  // point this only zeroes out the sequence number if possible for better
-  // compression.
+  // Do final preparations before presenting the output to the callee.
  void PrepareOutput();

+  // Passes the output value to the blob file builder (if any), and replaces it
+  // with the corresponding blob reference if it has been actually written to a
+  // blob file (i.e. if it passed the value size check). Returns true if the
+  // value got extracted to a blob file, false otherwise.
+  bool ExtractLargeValueIfNeededImpl();
+
+  // Extracts large values as described above, and updates the internal key's
+  // type to kTypeBlobIndex if the value got extracted. Should only be called
+  // for regular values (kTypeValue).
+  void ExtractLargeValueIfNeeded();
+
+  // Relocates valid blobs residing in the oldest blob files if garbage
+  // collection is enabled. Relocated blobs are written to new blob files or
+  // inlined in the LSM tree depending on the current settings (i.e.
+  // enable_blob_files and min_blob_size). Should only be called for blob
+  // references (kTypeBlobIndex).
+  //
+  // Note: the stacked BlobDB implementation's compaction filter based GC
+  // algorithm is also called from here.
+  void GarbageCollectBlobIfNeeded();
+
  // Invoke compaction filter if needed.
  // Return true on success, false on failures (e.g.: kIOError).
  bool InvokeFilterIfNeeded(bool* need_skip, Slice* skip_until);
@@ -191,6 +231,9 @@ class CompactionIterator {
    }
  }

+  static uint64_t ComputeBlobGarbageCollectionCutoffFileNumber(
+      const CompactionProxy* compaction);
+
  InternalIterator* input_;
  const Comparator* cmp_;
  MergeHelper* merge_helper_;
@@ -273,7 +316,11 @@ class CompactionIterator {
  // PinnedIteratorsManager used to pin input_ Iterator blocks while reading
  // merge operands and then releasing them after consuming them.
  PinnedIteratorsManager pinned_iters_mgr_;
+
+  uint64_t blob_garbage_collection_cutoff_file_number_;
+
  std::string blob_index_;
+  PinnableSlice blob_value_;
  std::string compaction_filter_value_;
  InternalKey compaction_filter_skip_until_;
  // "level_ptrs" holds indices that remember which file of an associated
@@ -176,6 +176,12 @@ class FakeCompaction : public CompactionIterator::CompactionProxy {

  bool preserve_deletes() const override { return false; }

+  bool enable_blob_garbage_collection() const override { return false; }
+
+  double blob_garbage_collection_age_cutoff() const override { return 0.0; }
+
+  Version* input_version() const override { return nullptr; }
+
  bool key_not_exists_beyond_output_level = false;

  bool is_bottommost_level = false;
@@ -1449,9 +1449,8 @@ Status CompactionJob::FinishCompactionOutputFile(
  } else {
    sub_compact->builder->Abandon();
  }
-  IOStatus io_s = sub_compact->builder->io_status();
  if (s.ok()) {
-    s = io_s;
+    s = sub_compact->builder->io_status();
  }
  const uint64_t current_bytes = sub_compact->builder->FileSize();
  if (s.ok()) {
@@ -1462,6 +1461,7 @@ Status CompactionJob::FinishCompactionOutputFile(
  sub_compact->total_bytes += current_bytes;

  // Finish and check for file errors
+  IOStatus io_s;
  if (s.ok()) {
    StopWatch sw(env_, stats_, COMPACTION_OUTFILE_SYNC_MICROS);
    io_s = sub_compact->outfile->Sync(db_options_.use_fsync);
@@ -1500,7 +1500,18 @@ Status CompactionJob::FinishCompactionOutputFile(
    std::string fname =
        TableFileName(sub_compact->compaction->immutable_cf_options()->cf_paths,
                      meta->fd.GetNumber(), meta->fd.GetPathId());
-    env_->DeleteFile(fname);
+
+    // TODO(AR) it is not clear if there are any larger implications if
+    // DeleteFile fails here
+    Status ds = env_->DeleteFile(fname);
+    if (!ds.ok()) {
+      ROCKS_LOG_WARN(
+          db_options_.info_log,
+          "[%s] [JOB %d] Unable to remove SST file for table #%" PRIu64
+          " at bottom level%s",
+          cfd->GetName().c_str(), job_id_, output_number,
+          meta->marked_for_compaction ? " (need compaction)" : "");
+    }

    // Also need to remove the file from outputs, or it will be added to the
    // VersionEdit.
@@ -1555,9 +1566,7 @@ Status CompactionJob::FinishCompactionOutputFile(
          "CompactionJob::FinishCompactionOutputFile:"
          "MaxAllowedSpaceReached");
      InstrumentedMutexLock l(db_mutex_);
-      // Should handle return error?
-      db_error_handler_->SetBGError(s, BackgroundErrorReason::kCompaction)
-          .PermitUncheckedError();
+      db_error_handler_->SetBGError(s, BackgroundErrorReason::kCompaction);
    }
  }
 #endif
@@ -269,7 +269,7 @@ class CompactionJobTestBase : public testing::Test {
                       &write_buffer_manager_, &write_controller_,
                       /*block_cache_tracer=*/nullptr, /*io_tracer=*/nullptr));
    compaction_job_stats_.Reset();
-    SetIdentityFile(env_, dbname_);
+    ASSERT_OK(SetIdentityFile(env_, dbname_));

    VersionEdit new_db;
    new_db.SetLogNumber(0);
@@ -141,7 +141,7 @@ class CompactionPickerTest : public testing::Test {
    if (temp_vstorage_) {
      VersionBuilder builder(FileOptions(), &ioptions_, nullptr,
                             vstorage_.get(), nullptr);
-      builder.SaveTo(temp_vstorage_.get());
+      ASSERT_OK(builder.SaveTo(temp_vstorage_.get()));
      vstorage_ = std::move(temp_vstorage_);
    }
    vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
@@ -21,6 +21,7 @@
 #include "db/version_set.h"
 #include "env/composite_env_wrapper.h"
 #include "file/filename.h"
+#include "port/stack_trace.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/convenience.h"
 #include "rocksdb/db.h"
@@ -42,7 +43,8 @@ static constexpr int kValueSize = 1000;

 class CorruptionTest : public testing::Test {
 public:
-  test::ErrorEnv env_;
+  std::shared_ptr<Env> env_guard_;
+  test::ErrorEnv* env_;
  std::string dbname_;
  std::shared_ptr<Cache> tiny_cache_;
  Options options_;
@@ -53,9 +55,19 @@ class CorruptionTest : public testing::Test {
    // set it to 0), test SequenceNumberRecovery will fail, likely because of a
    // bug in recovery code. Keep it 4 for now to make the test passes.
    tiny_cache_ = NewLRUCache(100, 4);
+    Env* base_env = Env::Default();
+#ifndef ROCKSDB_LITE
+    const char* test_env_uri = getenv("TEST_ENV_URI");
+    if (test_env_uri) {
+      Status s = Env::LoadEnv(test_env_uri, &base_env, &env_guard_);
+      EXPECT_OK(s);
+      EXPECT_NE(Env::Default(), base_env);
+    }
+#endif  //! ROCKSDB_LITE
+    env_ = new test::ErrorEnv(base_env);
    options_.wal_recovery_mode = WALRecoveryMode::kTolerateCorruptedTailRecords;
-    options_.env = &env_;
-    dbname_ = test::PerThreadDBPath("corruption_test");
+    options_.env = env_;
+    dbname_ = test::PerThreadDBPath(env_, "corruption_test");
    Status s = DestroyDB(dbname_, options_);
    EXPECT_OK(s);

@@ -77,8 +89,11 @@ class CorruptionTest : public testing::Test {
    if (getenv("KEEP_DB")) {
      fprintf(stdout, "db is still at %s\n", dbname_.c_str());
    } else {
-      EXPECT_OK(DestroyDB(dbname_, Options()));
+      Options opts;
+      opts.env = env_->target();
+      EXPECT_OK(DestroyDB(dbname_, opts));
    }
+    delete env_;
  }

  void CloseDb() {
@@ -93,7 +108,7 @@ class CorruptionTest : public testing::Test {
    if (opt.env == Options().env) {
      // If env is not overridden, replace it with ErrorEnv.
      // Otherwise, the test already uses a non-default Env.
-      opt.env = &env_;
+      opt.env = env_;
    }
    opt.arena_block_size = 4096;
    BlockBasedTableOptions table_options;
@@ -176,7 +191,7 @@ class CorruptionTest : public testing::Test {
  void Corrupt(FileType filetype, int offset, int bytes_to_corrupt) {
    // Pick file to corrupt
    std::vector<std::string> filenames;
-    ASSERT_OK(env_.GetChildren(dbname_, &filenames));
+    ASSERT_OK(env_->GetChildren(dbname_, &filenames));
    uint64_t number;
    FileType type;
    std::string fname;
@@ -191,7 +206,7 @@ class CorruptionTest : public testing::Test {
    }
    ASSERT_TRUE(!fname.empty()) << filetype;

-    ASSERT_OK(test::CorruptFile(&env_, fname, offset, bytes_to_corrupt));
+    ASSERT_OK(test::CorruptFile(env_, fname, offset, bytes_to_corrupt));
  }

  // corrupts exactly one file at level `level`. if no file found at level,
@@ -201,7 +216,7 @@ class CorruptionTest : public testing::Test {
    db_->GetLiveFilesMetaData(&metadata);
    for (const auto& m : metadata) {
      if (m.level == level) {
-        ASSERT_OK(test::CorruptFile(&env_, dbname_ + "/" + m.name, offset,
+        ASSERT_OK(test::CorruptFile(env_, dbname_ + "/" + m.name, offset,
                                    bytes_to_corrupt));
        return;
      }
@@ -268,14 +283,14 @@ TEST_F(CorruptionTest, Recovery) {
 }

 TEST_F(CorruptionTest, RecoverWriteError) {
-  env_.writable_file_error_ = true;
+  env_->writable_file_error_ = true;
  Status s = TryReopen();
  ASSERT_TRUE(!s.ok());
 }

 TEST_F(CorruptionTest, NewFileErrorDuringWrite) {
  // Do enough writing to force minor compaction
-  env_.writable_file_error_ = true;
+  env_->writable_file_error_ = true;
  const int num =
      static_cast<int>(3 + (Options().write_buffer_size / kValueSize));
  std::string value_storage;
@@ -291,8 +306,8 @@ TEST_F(CorruptionTest, NewFileErrorDuringWrite) {
    ASSERT_TRUE(!failed || !s.ok());
  }
  ASSERT_TRUE(!s.ok());
-  ASSERT_GE(env_.num_writable_file_errors_, 1);
-  env_.writable_file_error_ = false;
+  ASSERT_GE(env_->num_writable_file_errors_, 1);
+  env_->writable_file_error_ = false;
  Reopen();
 }

@@ -310,7 +325,7 @@ TEST_F(CorruptionTest, TableFile) {

 TEST_F(CorruptionTest, VerifyChecksumReadahead) {
  Options options;
-  SpecialEnv senv(Env::Default());
+  SpecialEnv senv(env_->target());
  options.env = &senv;
  // Disable block cache as we are going to check checksum for
  // the same file twice and measure number of reads.
@@ -432,6 +447,7 @@ TEST_F(CorruptionTest, CorruptedDescriptor) {

 TEST_F(CorruptionTest, CompactionInputError) {
  Options options;
+  options.env = env_;
  Reopen(&options);
  Build(10);
  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
@@ -452,6 +468,7 @@ TEST_F(CorruptionTest, CompactionInputError) {

 TEST_F(CorruptionTest, CompactionInputErrorParanoid) {
  Options options;
+  options.env = env_;
  options.paranoid_checks = true;
  options.write_buffer_size = 131072;
  options.max_write_buffer_number = 2;
@@ -537,7 +554,7 @@ TEST_F(CorruptionTest, RangeDeletionCorrupted) {
      ImmutableCFOptions(options_), kRangeDelBlock, &range_del_handle));

  ASSERT_OK(TryReopen());
-  ASSERT_OK(test::CorruptFile(&env_, filename,
+  ASSERT_OK(test::CorruptFile(env_, filename,
                              static_cast<int>(range_del_handle.offset()), 1));
  ASSERT_TRUE(TryReopen().IsCorruption());
 }
@@ -545,6 +562,7 @@ TEST_F(CorruptionTest, RangeDeletionCorrupted) {
 TEST_F(CorruptionTest, FileSystemStateCorrupted) {
  for (int iter = 0; iter < 2; ++iter) {
    Options options;
+    options.env = env_;
    options.paranoid_checks = true;
    options.create_if_missing = true;
    Reopen(&options);
@@ -561,13 +579,13 @@ TEST_F(CorruptionTest, FileSystemStateCorrupted) {

    if (iter == 0) {  // corrupt file size
      std::unique_ptr<WritableFile> file;
-      env_.NewWritableFile(filename, &file, EnvOptions());
+      ASSERT_OK(env_->NewWritableFile(filename, &file, EnvOptions()));
      ASSERT_OK(file->Append(Slice("corrupted sst")));
      file.reset();
      Status x = TryReopen(&options);
      ASSERT_TRUE(x.IsCorruption());
    } else {  // delete the file
-      ASSERT_OK(env_.DeleteFile(filename));
+      ASSERT_OK(env_->DeleteFile(filename));
      Status x = TryReopen(&options);
      ASSERT_TRUE(x.IsCorruption());
    }
@@ -583,6 +601,7 @@ static const auto& corruption_modes = {

 TEST_F(CorruptionTest, ParanoidFileChecksOnFlush) {
  Options options;
+  options.env = env_;
  options.check_flush_compaction_key_order = false;
  options.paranoid_file_checks = true;
  options.create_if_missing = true;
@@ -597,7 +616,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksOnFlush) {
    options.table_factory = mock;
    mock->SetCorruptionMode(mode);
    ASSERT_OK(DB::Open(options, dbname_, &db_));
-    assert(db_ != nullptr);
+    assert(db_ != nullptr);  // suppress false clang-analyze report
    Build(10);
    s = db_->Flush(FlushOptions());
    if (mode == mock::MockTableFactory::kCorruptNone) {
@@ -610,6 +629,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksOnFlush) {

 TEST_F(CorruptionTest, ParanoidFileChecksOnCompact) {
  Options options;
+  options.env = env_;
  options.paranoid_file_checks = true;
  options.create_if_missing = true;
  options.check_flush_compaction_key_order = false;
@@ -622,7 +642,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksOnCompact) {
        std::make_shared<mock::MockTableFactory>();
    options.table_factory = mock;
    ASSERT_OK(DB::Open(options, dbname_, &db_));
-    assert(db_ != nullptr);
+    assert(db_ != nullptr);  // suppress false clang-analyze report
    Build(100, 2);
    // ASSERT_OK(db_->Flush(FlushOptions()));
    DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
@@ -639,6 +659,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksOnCompact) {

 TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeFirst) {
  Options options;
+  options.env = env_;
  options.check_flush_compaction_key_order = false;
  options.paranoid_file_checks = true;
  options.create_if_missing = true;
@@ -648,7 +669,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeFirst) {
    ASSERT_OK(DestroyDB(dbname_, options));
    ASSERT_OK(DB::Open(options, dbname_, &db_));
    std::string start, end;
-    assert(db_ != nullptr);
+    assert(db_ != nullptr);  // suppress false clang-analyze report
    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
                               Key(3, &start), Key(7, &end)));
    auto snap = db_->GetSnapshot();
@@ -671,6 +692,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeFirst) {

 TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRange) {
  Options options;
+  options.env = env_;
  options.check_flush_compaction_key_order = false;
  options.paranoid_file_checks = true;
  options.create_if_missing = true;
@@ -679,7 +701,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRange) {
    db_ = nullptr;
    ASSERT_OK(DestroyDB(dbname_, options));
    ASSERT_OK(DB::Open(options, dbname_, &db_));
-    assert(db_ != nullptr);
+    assert(db_ != nullptr);  // suppress false clang-analyze report
    Build(10, 0, 0);
    std::string start, end;
    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
@@ -706,6 +728,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRange) {

 TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeLast) {
  Options options;
+  options.env = env_;
  options.check_flush_compaction_key_order = false;
  options.paranoid_file_checks = true;
  options.create_if_missing = true;
@@ -714,7 +737,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeLast) {
    db_ = nullptr;
    ASSERT_OK(DestroyDB(dbname_, options));
    ASSERT_OK(DB::Open(options, dbname_, &db_));
-    assert(db_ != nullptr);
+    assert(db_ != nullptr);  // suppress false clang-analyze report
    std::string start, end;
    Build(10);
    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
@@ -738,6 +761,7 @@ TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeLast) {

 TEST_F(CorruptionTest, LogCorruptionErrorsInCompactionIterator) {
  Options options;
+  options.env = env_;
  options.create_if_missing = true;
  options.allow_data_in_errors = true;
  auto mode = mock::MockTableFactory::kCorruptKey;
@@ -751,7 +775,7 @@ TEST_F(CorruptionTest, LogCorruptionErrorsInCompactionIterator) {
  options.table_factory = mock;

  ASSERT_OK(DB::Open(options, dbname_, &db_));
-  assert(db_ != nullptr);
+  assert(db_ != nullptr);  // suppress false clang-analyze report
  Build(100, 2);

  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
@@ -763,6 +787,7 @@ TEST_F(CorruptionTest, LogCorruptionErrorsInCompactionIterator) {

 TEST_F(CorruptionTest, CompactionKeyOrderCheck) {
  Options options;
+  options.env = env_;
  options.paranoid_file_checks = false;
  options.create_if_missing = true;
  options.check_flush_compaction_key_order = false;
@@ -773,7 +798,7 @@ TEST_F(CorruptionTest, CompactionKeyOrderCheck) {
      std::make_shared<mock::MockTableFactory>();
  options.table_factory = mock;
  ASSERT_OK(DB::Open(options, dbname_, &db_));
-  assert(db_ != nullptr);
+  assert(db_ != nullptr);  // suppress false clang-analyze report
  mock->SetCorruptionMode(mock::MockTableFactory::kCorruptReorderKey);
  Build(100, 2);
  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
@@ -786,6 +811,7 @@ TEST_F(CorruptionTest, CompactionKeyOrderCheck) {

 TEST_F(CorruptionTest, FlushKeyOrderCheck) {
  Options options;
+  options.env = env_;
  options.paranoid_file_checks = false;
  options.create_if_missing = true;
  ASSERT_OK(db_->SetOptions({{"check_flush_compaction_key_order", "true"}}));
@@ -814,7 +840,6 @@ TEST_F(CorruptionTest, FlushKeyOrderCheck) {
 }

 TEST_F(CorruptionTest, DisableKeyOrderCheck) {
-  Options options;
  ASSERT_OK(db_->SetOptions({{"check_flush_compaction_key_order", "false"}}));
  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);

@@ -836,7 +861,7 @@ TEST_F(CorruptionTest, DisableKeyOrderCheck) {
 TEST_F(CorruptionTest, VerifyWholeTableChecksum) {
  CloseDb();
  Options options;
-  options.env = &env_;
+  options.env = env_;
  ASSERT_OK(DestroyDB(dbname_, options));
  options.create_if_missing = true;
  options.file_checksum_gen_factory =
@@ -859,7 +884,7 @@ TEST_F(CorruptionTest, VerifyWholeTableChecksum) {
  SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::VerifySstFileChecksum:mismatch", [&](void* arg) {
        auto* s = reinterpret_cast<Status*>(arg);
-        assert(s);
+        ASSERT_NE(s, nullptr);
        ++count;
        ASSERT_NOK(*s);
      });
@@ -870,8 +895,18 @@ TEST_F(CorruptionTest, VerifyWholeTableChecksum) {

 }  // namespace ROCKSDB_NAMESPACE

+#ifdef ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS
+extern "C" {
+void RegisterCustomObjects(int argc, char** argv);
+}
+#else
+void RegisterCustomObjects(int /*argc*/, char** /*argv*/) {}
+#endif  // !ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS
+
 int main(int argc, char** argv) {
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
+  RegisterCustomObjects(argc, argv);
  return RUN_ALL_TESTS();
 }

@@ -129,10 +129,10 @@ TEST_F(CuckooTableDBTest, Flush) {
  ASSERT_OK(Put("key1", "v1"));
  ASSERT_OK(Put("key2", "v2"));
  ASSERT_OK(Put("key3", "v3"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

  TablePropertiesCollection ptc;
-  reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc);
+  ASSERT_OK(reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc));
  ASSERT_EQ(1U, ptc.size());
  ASSERT_EQ(3U, ptc.begin()->second->num_entries);
  ASSERT_EQ("1", FilesPerLevel());
@@ -146,9 +146,9 @@ TEST_F(CuckooTableDBTest, Flush) {
  ASSERT_OK(Put("key4", "v4"));
  ASSERT_OK(Put("key5", "v5"));
  ASSERT_OK(Put("key6", "v6"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

-  reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc);
+  ASSERT_OK(reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc));
  ASSERT_EQ(2U, ptc.size());
  auto row = ptc.begin();
  ASSERT_EQ(3U, row->second->num_entries);
@@ -164,8 +164,8 @@ TEST_F(CuckooTableDBTest, Flush) {
  ASSERT_OK(Delete("key6"));
  ASSERT_OK(Delete("key5"));
  ASSERT_OK(Delete("key4"));
-  dbfull()->TEST_FlushMemTable();
-  reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc);
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());
+  ASSERT_OK(reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc));
  ASSERT_EQ(3U, ptc.size());
  row = ptc.begin();
  ASSERT_EQ(3U, row->second->num_entries);
@@ -186,10 +186,10 @@ TEST_F(CuckooTableDBTest, FlushWithDuplicateKeys) {
  ASSERT_OK(Put("key1", "v1"));
  ASSERT_OK(Put("key2", "v2"));
  ASSERT_OK(Put("key1", "v3"));  // Duplicate
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

  TablePropertiesCollection ptc;
-  reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc);
+  ASSERT_OK(reinterpret_cast<DB*>(dbfull())->GetPropertiesOfAllTables(&ptc));
  ASSERT_EQ(1U, ptc.size());
  ASSERT_EQ(2U, ptc.begin()->second->num_entries);
  ASSERT_EQ("1", FilesPerLevel());
@@ -219,7 +219,7 @@ TEST_F(CuckooTableDBTest, Uint64Comparator) {
  ASSERT_OK(Put(Uint64Key(1), "v1"));
  ASSERT_OK(Put(Uint64Key(2), "v2"));
  ASSERT_OK(Put(Uint64Key(3), "v3"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

  ASSERT_EQ("v1", Get(Uint64Key(1)));
  ASSERT_EQ("v2", Get(Uint64Key(2)));
@@ -228,10 +228,10 @@ TEST_F(CuckooTableDBTest, Uint64Comparator) {

  // Add more keys.
  ASSERT_OK(Delete(Uint64Key(2)));  // Delete.
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());
  ASSERT_OK(Put(Uint64Key(3), "v0"));  // Update.
  ASSERT_OK(Put(Uint64Key(4), "v4"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());
  ASSERT_EQ("v1", Get(Uint64Key(1)));
  ASSERT_EQ("NOT_FOUND", Get(Uint64Key(2)));
  ASSERT_EQ("v0", Get(Uint64Key(3)));
@@ -251,11 +251,11 @@ TEST_F(CuckooTableDBTest, CompactionIntoMultipleFiles) {
  for (int idx = 0; idx < 28; ++idx) {
    ASSERT_OK(Put(Key(idx), std::string(10000, 'a' + char(idx))));
  }
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  ASSERT_EQ("1", FilesPerLevel());

-  dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                              true /* disallow trivial move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                        true /* disallow trivial move */));
  ASSERT_EQ("0,2", FilesPerLevel());
  for (int idx = 0; idx < 28; ++idx) {
    ASSERT_EQ(std::string(10000, 'a' + char(idx)), Get(Key(idx)));
@@ -274,15 +274,15 @@ TEST_F(CuckooTableDBTest, SameKeyInsertedInTwoDifferentFilesAndCompacted) {
  for (int idx = 0; idx < 11; ++idx) {
    ASSERT_OK(Put(Key(idx), std::string(10000, 'a')));
  }
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  ASSERT_EQ("1", FilesPerLevel());

  // Generate one more file in level-0, and should trigger level-0 compaction
  for (int idx = 0; idx < 11; ++idx) {
    ASSERT_OK(Put(Key(idx), std::string(10000, 'a' + char(idx))));
  }
-  dbfull()->TEST_WaitForFlushMemTable();
-  dbfull()->TEST_CompactRange(0, nullptr, nullptr);
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr));

  ASSERT_EQ("0,1", FilesPerLevel());
  for (int idx = 0; idx < 11; ++idx) {
@@ -303,7 +303,7 @@ TEST_F(CuckooTableDBTest, AdaptiveTable) {
  ASSERT_OK(Put("key1", "v1"));
  ASSERT_OK(Put("key2", "v2"));
  ASSERT_OK(Put("key3", "v3"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

  // Write some keys using plain table.
  std::shared_ptr<TableFactory> block_based_factory(
@@ -319,7 +319,7 @@ TEST_F(CuckooTableDBTest, AdaptiveTable) {
  Reopen(&options);
  ASSERT_OK(Put("key4", "v4"));
  ASSERT_OK(Put("key1", "v5"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

  // Write some keys using block based table.
  options.table_factory.reset(NewAdaptiveTableFactory(
@@ -328,7 +328,7 @@ TEST_F(CuckooTableDBTest, AdaptiveTable) {
  Reopen(&options);
  ASSERT_OK(Put("key5", "v6"));
  ASSERT_OK(Put("key2", "v7"));
-  dbfull()->TEST_FlushMemTable();
+  ASSERT_OK(dbfull()->TEST_FlushMemTable());

  ASSERT_EQ("v5", Get("key1"));
  ASSERT_EQ("v7", Get("key2"));
@@ -38,7 +38,10 @@ TEST_F(DBBasicTest, OpenWhenOpen) {
  options.env = env_;
  DB* db2 = nullptr;
  Status s = DB::Open(options, dbname_, &db2);
-  ASSERT_NOK(s);
+  ASSERT_NOK(s) << [db2]() {
+    delete db2;
+    return "db2 open: ok";
+  }();
  ASSERT_EQ(Status::Code::kIOError, s.code());
  ASSERT_EQ(Status::SubCode::kNone, s.subcode());
  ASSERT_TRUE(strstr(s.getState(), "lock ") != nullptr);
@@ -143,7 +146,7 @@ TEST_F(DBBasicTest, ReadOnlyDB) {

  // Reopen and flush memtable.
  Reopen(options);
-  Flush();
+  ASSERT_OK(Flush());
  Close();
  // Now check keys in read only mode.
  ASSERT_OK(ReadOnlyReopen(options));
@@ -179,7 +182,7 @@ TEST_F(DBBasicTest, ReadOnlyDBWithWriteDBIdToManifestSet) {

  // Reopen and flush memtable.
  Reopen(options);
-  Flush();
+  ASSERT_OK(Flush());
  Close();
  // Now check keys in read only mode.
  ASSERT_OK(ReadOnlyReopen(options));
@@ -202,7 +205,7 @@ TEST_F(DBBasicTest, CompactedDB) {
  Reopen(options);
  // 1 L0 file, use CompactedDB if max_open_files = -1
  ASSERT_OK(Put("aaa", DummyString(kFileSize / 2, '1')));
-  Flush();
+  ASSERT_OK(Flush());
  Close();
  ASSERT_OK(ReadOnlyReopen(options));
  Status s = Put("new", "value");
@@ -220,12 +223,12 @@ TEST_F(DBBasicTest, CompactedDB) {
  Reopen(options);
  // Add more L0 files
  ASSERT_OK(Put("bbb", DummyString(kFileSize / 2, '2')));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(Put("aaa", DummyString(kFileSize / 2, 'a')));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(Put("bbb", DummyString(kFileSize / 2, 'b')));
  ASSERT_OK(Put("eee", DummyString(kFileSize / 2, 'e')));
-  Flush();
+  ASSERT_OK(Flush());
  Close();

  ASSERT_OK(ReadOnlyReopen(options));
@@ -402,13 +405,16 @@ TEST_F(DBBasicTest, GetSnapshot) {

 TEST_F(DBBasicTest, CheckLock) {
  do {
-    DB* localdb;
+    DB* localdb = nullptr;
    Options options = CurrentOptions();
    ASSERT_OK(TryReopen(options));

    // second open should fail
    Status s = DB::Open(options, dbname_, &localdb);
-    ASSERT_NOK(s);
+    ASSERT_NOK(s) << [localdb]() {
+      delete localdb;
+      return "localdb open: ok";
+    }();
 #ifdef OS_LINUX
    ASSERT_TRUE(s.ToString().find("lock ") != std::string::npos);
 #endif  // OS_LINUX
@@ -1407,10 +1413,10 @@ TEST_F(DBBasicTest, MultiGetBatchedSortedMultiFile) {
    // mix with memtable
    ASSERT_OK(Put(1, "k1", "v1"));
    ASSERT_OK(Put(1, "k2", "v2"));
-    Flush(1);
+    ASSERT_OK(Flush(1));
    ASSERT_OK(Put(1, "k3", "v3"));
    ASSERT_OK(Put(1, "k4", "v4"));
-    Flush(1);
+    ASSERT_OK(Flush(1));
    ASSERT_OK(Delete(1, "k4"));
    ASSERT_OK(Put(1, "k5", "v5"));
    ASSERT_OK(Delete(1, "no_key"));
@@ -1453,19 +1459,19 @@ TEST_F(DBBasicTest, MultiGetBatchedDuplicateKeys) {
  // mix with memtable
  ASSERT_OK(Merge(1, "k1", "v1"));
  ASSERT_OK(Merge(1, "k2", "v2"));
-  Flush(1);
+  ASSERT_OK(Flush(1));
  MoveFilesToLevel(2, 1);
  ASSERT_OK(Merge(1, "k3", "v3"));
  ASSERT_OK(Merge(1, "k4", "v4"));
-  Flush(1);
+  ASSERT_OK(Flush(1));
  MoveFilesToLevel(2, 1);
  ASSERT_OK(Merge(1, "k4", "v4_2"));
  ASSERT_OK(Merge(1, "k6", "v6"));
-  Flush(1);
+  ASSERT_OK(Flush(1));
  MoveFilesToLevel(2, 1);
  ASSERT_OK(Merge(1, "k7", "v7"));
  ASSERT_OK(Merge(1, "k8", "v8"));
-  Flush(1);
+  ASSERT_OK(Flush(1));
  MoveFilesToLevel(2, 1);

  get_perf_context()->Reset();
@@ -1505,12 +1511,12 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevel) {
    ASSERT_OK(Put("key_" + std::to_string(i), "val_l2_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  MoveFilesToLevel(2);
@@ -1519,12 +1525,12 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevel) {
    ASSERT_OK(Put("key_" + std::to_string(i), "val_l1_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  MoveFilesToLevel(1);
@@ -1533,12 +1539,12 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevel) {
    ASSERT_OK(Put("key_" + std::to_string(i), "val_l0_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  ASSERT_EQ(0, num_keys);
@@ -1584,12 +1590,12 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevelMerge) {
    ASSERT_OK(Put("key_" + std::to_string(i), "val_l2_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  MoveFilesToLevel(2);
@@ -1598,12 +1604,12 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevelMerge) {
    ASSERT_OK(Merge("key_" + std::to_string(i), "val_l1_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  MoveFilesToLevel(1);
@@ -1612,12 +1618,12 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevelMerge) {
    ASSERT_OK(Merge("key_" + std::to_string(i), "val_l0_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  ASSERT_EQ(0, num_keys);
@@ -1699,7 +1705,7 @@ TEST_F(DBBasicTest, MultiGetBatchedValueSize) {
    ASSERT_OK(Put(1, "k7", "v7_"));
    ASSERT_OK(Put(1, "k3", "v3_"));
    ASSERT_OK(Put(1, "k4", "v4"));
-    Flush(1);
+    ASSERT_OK(Flush(1));
    ASSERT_OK(Delete(1, "k4"));
    ASSERT_OK(Put(1, "k11", "v11"));
    ASSERT_OK(Delete(1, "no_key"));
@@ -1709,7 +1715,7 @@ TEST_F(DBBasicTest, MultiGetBatchedValueSize) {
    ASSERT_OK(Put(1, "k15", "v15"));
    ASSERT_OK(Put(1, "k16", "v16"));
    ASSERT_OK(Put(1, "k17", "v17"));
-    Flush(1);
+    ASSERT_OK(Flush(1));

    ASSERT_OK(Put(1, "k1", "v1_"));
    ASSERT_OK(Put(1, "k2", "v2_"));
@@ -1779,12 +1785,12 @@ TEST_F(DBBasicTest, MultiGetBatchedValueSizeMultiLevelMerge) {
    ASSERT_OK(Put("key_" + std::to_string(i), "val_l2_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  MoveFilesToLevel(2);
@@ -1793,12 +1799,12 @@ TEST_F(DBBasicTest, MultiGetBatchedValueSizeMultiLevelMerge) {
    ASSERT_OK(Merge("key_" + std::to_string(i), "val_l1_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  MoveFilesToLevel(1);
@@ -1807,12 +1813,12 @@ TEST_F(DBBasicTest, MultiGetBatchedValueSizeMultiLevelMerge) {
    ASSERT_OK(Merge("key_" + std::to_string(i), "val_l0_" + std::to_string(i)));
    num_keys++;
    if (num_keys == 8) {
-      Flush();
+      ASSERT_OK(Flush());
      num_keys = 0;
    }
  }
  if (num_keys > 0) {
-    Flush();
+    ASSERT_OK(Flush());
    num_keys = 0;
  }
  ASSERT_EQ(0, num_keys);
@@ -1902,10 +1908,10 @@ TEST_F(DBBasicTest, MultiGetStats) {
    keys[i] = Slice(keys_str[i]);
    ASSERT_OK(Put(1, "k" + std::to_string(i), rnd.RandomString(1000)));
    if (i % 100 == 0) {
-      Flush(1);
+      ASSERT_OK(Flush(1));
    }
  }
-  Flush(1);
+  ASSERT_OK(Flush(1));
  MoveFilesToLevel(2, 1);

  for (int i = 501; i < 1000; ++i) {
@@ -1913,11 +1919,11 @@ TEST_F(DBBasicTest, MultiGetStats) {
    keys[i] = Slice(keys_str[i]);
    ASSERT_OK(Put(1, "k" + std::to_string(i), rnd.RandomString(1000)));
    if (i % 100 == 0) {
-      Flush(1);
+      ASSERT_OK(Flush(1));
    }
  }

-  Flush(1);
+  ASSERT_OK(Flush(1));
  MoveFilesToLevel(2, 1);

  for (int i = 1001; i < total_keys; ++i) {
@@ -1925,10 +1931,10 @@ TEST_F(DBBasicTest, MultiGetStats) {
    keys[i] = Slice(keys_str[i]);
    ASSERT_OK(Put(1, "k" + std::to_string(i), rnd.RandomString(1000)));
    if (i % 100 == 0) {
-      Flush(1);
+      ASSERT_OK(Flush(1));
    }
  }
-  Flush(1);
+  ASSERT_OK(Flush(1));
  Close();

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
@@ -1955,7 +1961,7 @@ TEST_F(DBBasicTest, MultiGetStats) {
  ASSERT_GT(hist_sst.max, 0);

  // Minimun number of blocks read in a level.
-  ASSERT_EQ(hist_data_blocks.min, 0);
+  ASSERT_EQ(hist_data_blocks.min, 3);
  ASSERT_GT(hist_index_and_filter_blocks.min, 0);
  // Minimun number of sst files read in a level.
  ASSERT_GT(hist_sst.max, 0);
@@ -2036,11 +2042,11 @@ TEST_P(DBMultiGetRowCacheTest, MultiGetBatched) {
    ASSERT_OK(Put(1, "k2", "v2"));
    ASSERT_OK(Put(1, "k3", "v3"));
    ASSERT_OK(Put(1, "k4", "v4"));
-    Flush(1);
+    ASSERT_OK(Flush(1));
    ASSERT_OK(Put(1, "k5", "v5"));
    const Snapshot* snap1 = dbfull()->GetSnapshot();
    ASSERT_OK(Delete(1, "k4"));
-    Flush(1);
+    ASSERT_OK(Flush(1));
    const Snapshot* snap2 = dbfull()->GetSnapshot();

    get_perf_context()->Reset();
@@ -2179,7 +2185,7 @@ TEST_F(DBBasicTest, MultiGetIOBufferOverrun) {
    std::string value(rnd.RandomString(128) + zero_str);
    assert(Put(Key(i), value) == Status::OK());
  }
-  Flush();
+  ASSERT_OK(Flush());

  std::vector<std::string> key_data(10);
  std::vector<Slice> keys;
@@ -2499,6 +2505,42 @@ TEST_F(DBBasicTest, SkipWALIfMissingTableFiles) {
  ASSERT_FALSE(iter->Valid());
  ASSERT_OK(iter->status());
 }
+
+TEST_F(DBBasicTest, DisableTrackWal) {
+  // If WAL tracking was enabled, and then disabled during reopen,
+  // the previously tracked WALs should be removed from MANIFEST.
+
+  Options options = CurrentOptions();
+  options.track_and_verify_wals_in_manifest = true;
+  // extremely small write buffer size,
+  // so that new WALs are created more frequently.
+  options.write_buffer_size = 100;
+  options.env = env_;
+  DestroyAndReopen(options);
+  for (int i = 0; i < 100; i++) {
+    ASSERT_OK(Put("foo" + std::to_string(i), "value" + std::to_string(i)));
+  }
+  ASSERT_OK(dbfull()->TEST_SwitchMemtable());
+  ASSERT_OK(db_->SyncWAL());
+  // Some WALs are tracked.
+  ASSERT_FALSE(dbfull()->TEST_GetVersionSet()->GetWalSet().GetWals().empty());
+  Close();
+
+  // Disable WAL tracking.
+  options.track_and_verify_wals_in_manifest = false;
+  options.create_if_missing = false;
+  ASSERT_OK(TryReopen(options));
+  // Previously tracked WALs are cleared.
+  ASSERT_TRUE(dbfull()->TEST_GetVersionSet()->GetWalSet().GetWals().empty());
+  Close();
+
+  // Re-enable WAL tracking again.
+  options.track_and_verify_wals_in_manifest = true;
+  options.create_if_missing = false;
+  ASSERT_OK(TryReopen(options));
+  ASSERT_TRUE(dbfull()->TEST_GetVersionSet()->GetWalSet().GetWals().empty());
+  Close();
+}
 #endif  // !ROCKSDB_LITE

 TEST_F(DBBasicTest, ManifestChecksumMismatch) {
@@ -2528,6 +2570,64 @@ TEST_F(DBBasicTest, ManifestChecksumMismatch) {
  ASSERT_TRUE(s.IsCorruption());
 }

+#ifndef ROCKSDB_LITE
+class DBBasicTestTrackWal : public DBTestBase,
+                            public testing::WithParamInterface<bool> {
+ public:
+  DBBasicTestTrackWal()
+      : DBTestBase("/db_basic_test_track_wal", /*env_do_fsync=*/false) {}
+
+  int CountWalFiles() {
+    VectorLogPtr log_files;
+    EXPECT_OK(dbfull()->GetSortedWalFiles(log_files));
+    return static_cast<int>(log_files.size());
+  };
+};
+
+TEST_P(DBBasicTestTrackWal, DoNotTrackObsoleteWal) {
+  // If a WAL becomes obsolete after flushing, but is not deleted from disk yet,
+  // then if SyncWAL is called afterwards, the obsolete WAL should not be
+  // tracked in MANIFEST.
+
+  Options options = CurrentOptions();
+  options.create_if_missing = true;
+  options.track_and_verify_wals_in_manifest = true;
+  options.atomic_flush = GetParam();
+
+  DestroyAndReopen(options);
+  CreateAndReopenWithCF({"cf"}, options);
+  ASSERT_EQ(handles_.size(), 2);  // default, cf
+  // Do not delete WALs.
+  ASSERT_OK(db_->DisableFileDeletions());
+  constexpr int n = 10;
+  std::vector<std::unique_ptr<LogFile>> wals(n);
+  for (size_t i = 0; i < n; i++) {
+    // Generate a new WAL for each key-value.
+    const int cf = i % 2;
+    ASSERT_OK(db_->GetCurrentWalFile(&wals[i]));
+    ASSERT_OK(Put(cf, "k" + std::to_string(i), "v" + std::to_string(i)));
+    ASSERT_OK(Flush({0, 1}));
+  }
+  ASSERT_EQ(CountWalFiles(), n);
+  // Since all WALs are obsolete, no WAL should be tracked in MANIFEST.
+  ASSERT_OK(db_->SyncWAL());
+
+  // Manually delete all WALs.
+  Close();
+  for (const auto& wal : wals) {
+    ASSERT_OK(env_->DeleteFile(LogFileName(dbname_, wal->LogNumber())));
+  }
+
+  // If SyncWAL tracks the obsolete WALs in MANIFEST,
+  // reopen will fail because the WALs are missing from disk.
+  ASSERT_OK(TryReopenWithColumnFamilies({"default", "cf"}, options));
+  Destroy(options);
+}
+
+INSTANTIATE_TEST_CASE_P(DBBasicTestTrackWal, DBBasicTestTrackWal,
+                        testing::Bool());
+#endif  // ROCKSDB_LITE
+
 class DBBasicTestMultiGet : public DBTestBase {
 public:
  DBBasicTestMultiGet(std::string test_dir, int num_cfs, bool compressed_cache,
@@ -2593,6 +2693,7 @@ class DBBasicTestMultiGet : public DBTestBase {
    } else {
      options.compression_opts.parallel_threads = compression_parallel_threads;
    }
+    options_ = options;
    Reopen(options);

    if (num_cfs > 1) {
@@ -2613,9 +2714,9 @@ class DBBasicTestMultiGet : public DBTestBase {
                               : Put(cf, Key(i), values_[i])) == Status::OK());
      }
      if (num_cfs == 1) {
-        Flush();
+        EXPECT_OK(Flush());
      } else {
-        dbfull()->Flush(FlushOptions(), handles_[cf]);
+        EXPECT_OK(dbfull()->Flush(FlushOptions(), handles_[cf]));
      }

      for (int i = 0; i < 100; ++i) {
@@ -2627,9 +2728,9 @@ class DBBasicTestMultiGet : public DBTestBase {
               Status::OK());
      }
      if (num_cfs == 1) {
-        Flush();
+        EXPECT_OK(Flush());
      } else {
-        dbfull()->Flush(FlushOptions(), handles_[cf]);
+        EXPECT_OK(dbfull()->Flush(FlushOptions(), handles_[cf]));
      }
    }
  }
@@ -2662,6 +2763,7 @@ class DBBasicTestMultiGet : public DBTestBase {
  bool compression_enabled() { return compression_enabled_; }
  bool has_compressed_cache() { return compressed_cache_ != nullptr; }
  bool has_uncompressed_cache() { return uncompressed_cache_ != nullptr; }
+  Options get_options() { return options_; }

  static void SetUpTestCase() {}
  static void TearDownTestCase() {}
@@ -2747,6 +2849,7 @@ class DBBasicTestMultiGet : public DBTestBase {

  std::shared_ptr<MyBlockCache> compressed_cache_;
  std::shared_ptr<MyBlockCache> uncompressed_cache_;
+  Options options_;
  bool compression_enabled_;
  std::vector<std::string> values_;
  std::vector<std::string> uncompressable_values_;
@@ -2889,6 +2992,123 @@ TEST_P(DBBasicTestWithParallelIO, MultiGet) {
  }
 }

+#ifndef ROCKSDB_LITE
+TEST_P(DBBasicTestWithParallelIO, MultiGetDirectIO) {
+  class FakeDirectIOEnv : public EnvWrapper {
+    class FakeDirectIOSequentialFile;
+    class FakeDirectIORandomAccessFile;
+
+   public:
+    FakeDirectIOEnv(Env* env) : EnvWrapper(env) {}
+
+    Status NewRandomAccessFile(const std::string& fname,
+                               std::unique_ptr<RandomAccessFile>* result,
+                               const EnvOptions& options) override {
+      std::unique_ptr<RandomAccessFile> file;
+      assert(options.use_direct_reads);
+      EnvOptions opts = options;
+      opts.use_direct_reads = false;
+      Status s = target()->NewRandomAccessFile(fname, &file, opts);
+      if (!s.ok()) {
+        return s;
+      }
+      result->reset(new FakeDirectIORandomAccessFile(std::move(file)));
+      return s;
+    }
+
+   private:
+    class FakeDirectIOSequentialFile : public SequentialFileWrapper {
+     public:
+      FakeDirectIOSequentialFile(std::unique_ptr<SequentialFile>&& file)
+          : SequentialFileWrapper(file.get()), file_(std::move(file)) {}
+      ~FakeDirectIOSequentialFile() {}
+
+      bool use_direct_io() const override { return true; }
+      size_t GetRequiredBufferAlignment() const override { return 1; }
+
+     private:
+      std::unique_ptr<SequentialFile> file_;
+    };
+
+    class FakeDirectIORandomAccessFile : public RandomAccessFileWrapper {
+     public:
+      FakeDirectIORandomAccessFile(std::unique_ptr<RandomAccessFile>&& file)
+          : RandomAccessFileWrapper(file.get()), file_(std::move(file)) {}
+      ~FakeDirectIORandomAccessFile() {}
+
+      bool use_direct_io() const override { return true; }
+      size_t GetRequiredBufferAlignment() const override { return 1; }
+
+     private:
+      std::unique_ptr<RandomAccessFile> file_;
+    };
+  };
+
+  std::unique_ptr<FakeDirectIOEnv> env(new FakeDirectIOEnv(env_));
+  Options opts = get_options();
+  opts.env = env.get();
+  opts.use_direct_reads = true;
+  Reopen(opts);
+
+  std::vector<std::string> key_data(10);
+  std::vector<Slice> keys;
+  // We cannot resize a PinnableSlice vector, so just set initial size to
+  // largest we think we will need
+  std::vector<PinnableSlice> values(10);
+  std::vector<Status> statuses;
+  ReadOptions ro;
+  ro.fill_cache = fill_cache();
+
+  // Warm up the cache first
+  key_data.emplace_back(Key(0));
+  keys.emplace_back(Slice(key_data.back()));
+  key_data.emplace_back(Key(50));
+  keys.emplace_back(Slice(key_data.back()));
+  statuses.resize(keys.size());
+
+  dbfull()->MultiGet(ro, dbfull()->DefaultColumnFamily(), keys.size(),
+                     keys.data(), values.data(), statuses.data(), true);
+  ASSERT_TRUE(CheckValue(0, values[0].ToString()));
+  ASSERT_TRUE(CheckValue(50, values[1].ToString()));
+
+  int random_reads = env_->random_read_counter_.Read();
+  key_data[0] = Key(1);
+  key_data[1] = Key(51);
+  keys[0] = Slice(key_data[0]);
+  keys[1] = Slice(key_data[1]);
+  values[0].Reset();
+  values[1].Reset();
+  if (uncompressed_cache_) {
+    uncompressed_cache_->SetCapacity(0);
+    uncompressed_cache_->SetCapacity(1048576);
+  }
+  dbfull()->MultiGet(ro, dbfull()->DefaultColumnFamily(), keys.size(),
+                     keys.data(), values.data(), statuses.data(), true);
+  ASSERT_TRUE(CheckValue(1, values[0].ToString()));
+  ASSERT_TRUE(CheckValue(51, values[1].ToString()));
+
+  bool read_from_cache = false;
+  if (fill_cache()) {
+    if (has_uncompressed_cache()) {
+      read_from_cache = true;
+    } else if (has_compressed_cache() && compression_enabled()) {
+      read_from_cache = true;
+    }
+  }
+
+  int expected_reads = random_reads;
+  if (!compression_enabled() || !has_compressed_cache()) {
+    expected_reads += 2;
+  } else {
+    expected_reads += (read_from_cache ? 0 : 2);
+  }
+  if (env_->random_read_counter_.Read() != expected_reads) {
+    ASSERT_EQ(env_->random_read_counter_.Read(), expected_reads);
+  }
+  Close();
+}
+#endif  // ROCKSDB_LITE
+
 TEST_P(DBBasicTestWithParallelIO, MultiGetWithChecksumMismatch) {
  std::vector<std::string> key_data(10);
  std::vector<Slice> keys;
@@ -3423,7 +3643,7 @@ TEST_P(DBBasicTestDeadline, PointLookupDeadline) {
      std::string key = "k" + ToString(i);
      ASSERT_OK(Put(key, rnd.RandomString(100)));
    }
-    Flush();
+    ASSERT_OK(Flush());

    bool timedout = true;
    // A timeout will be forced when the IO counter reaches this value
@@ -677,7 +677,7 @@ TEST_F(DBBlockCacheTest, ParanoidFileChecks) {
  // Create a new SST file. This will further trigger a compaction
  // and generate another file.
  ASSERT_OK(Flush(1));
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(3, /* Totally 3 files created up to now */
            TestGetTickerCount(options, BLOCK_CACHE_ADD));

@@ -692,7 +692,7 @@ TEST_F(DBBlockCacheTest, ParanoidFileChecks) {
  ASSERT_OK(Put(1, "1_key4", "val4"));
  ASSERT_OK(Put(1, "9_key4", "val4"));
  ASSERT_OK(Flush(1));
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(3, /* Totally 3 files created up to now */
            TestGetTickerCount(options, BLOCK_CACHE_ADD));
 }
@@ -860,7 +860,7 @@ TEST_F(DBBlockCacheTest, CacheCompressionDict) {
      }
      ASSERT_OK(Flush());
    }
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
    ASSERT_EQ(0, NumTableFilesAtLevel(0));
    ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(1));

@@ -128,8 +128,8 @@ TEST_P(DBBloomFilterTestDefFormatVersion, KeyMayExist) {
    ASSERT_EQ(cache_added, TestGetTickerCount(options, BLOCK_CACHE_ADD));

    ASSERT_OK(Flush(1));
-    dbfull()->TEST_CompactRange(0, nullptr, nullptr, handles_[1],
-                                true /* disallow trivial move */);
+    ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, handles_[1],
+                                          true /* disallow trivial move */));

    numopen = TestGetTickerCount(options, NO_FILE_OPENS);
    cache_added = TestGetTickerCount(options, BLOCK_CACHE_ADD);
@@ -178,7 +178,7 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloomCustomPrefixExtractor) {
    ASSERT_OK(dbfull()->Put(wo, "barbarbar2", "foo2"));
    ASSERT_OK(dbfull()->Put(wo, "foofoofoo", "bar"));

-    dbfull()->Flush(fo);
+    ASSERT_OK(dbfull()->Flush(fo));

    ASSERT_EQ("foo", Get("barbarbar"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
@@ -244,7 +244,7 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloom) {
    ASSERT_OK(dbfull()->Put(wo, "barbarbar2", "foo2"));
    ASSERT_OK(dbfull()->Put(wo, "foofoofoo", "bar"));

-    dbfull()->Flush(fo);
+    ASSERT_OK(dbfull()->Flush(fo));

    ASSERT_EQ("foo", Get("barbarbar"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
@@ -297,7 +297,7 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
    // ranges.
    ASSERT_OK(dbfull()->Put(wo, "aaa", ""));
    ASSERT_OK(dbfull()->Put(wo, "zzz", ""));
-    dbfull()->Flush(fo);
+    ASSERT_OK(dbfull()->Flush(fo));

    Reopen(options);
    ASSERT_EQ("NOT_FOUND", Get("foo"));
@@ -328,7 +328,7 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
    // ranges.
    ASSERT_OK(dbfull()->Put(wo, "aaa", ""));
    ASSERT_OK(dbfull()->Put(wo, "zzz", ""));
-    db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+    ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

    // Reopen with both of whole key off and prefix extractor enabled.
    // Still no bloom filter should be used.
@@ -351,7 +351,7 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
    // ranges.
    ASSERT_OK(dbfull()->Put(wo, "aaa", ""));
    ASSERT_OK(dbfull()->Put(wo, "zzz", ""));
-    db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+    ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

    options.prefix_extractor.reset();
    bbto.whole_key_filtering = true;
@@ -364,7 +364,7 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
    // not filtered out by key ranges.
    ASSERT_OK(dbfull()->Put(wo, "aaa", ""));
    ASSERT_OK(dbfull()->Put(wo, "zzz", ""));
-    Flush();
+    ASSERT_OK(Flush());

    // Now we have two files:
    // File 1: An older file with prefix bloom.
@@ -467,7 +467,7 @@ TEST_P(DBBloomFilterTestWithParam, BloomFilter) {
    for (int i = 0; i < N; i += 100) {
      ASSERT_OK(Put(1, Key(i), Key(i)));
    }
-    Flush(1);
+    ASSERT_OK(Flush(1));

    // Prevent auto compactions triggered by seeks
    env_->delay_sstable_sync_.store(true, std::memory_order_release);
@@ -880,7 +880,7 @@ TEST_F(DBBloomFilterTest, ContextCustomFilterPolicy) {

    // Destroy
    ASSERT_OK(dbfull()->DropColumnFamily(handles_[1]));
-    dbfull()->DestroyColumnFamilyHandle(handles_[1]);
+    ASSERT_OK(dbfull()->DestroyColumnFamilyHandle(handles_[1]));
    handles_[1] = nullptr;
  }
 }
@@ -1444,9 +1444,9 @@ void PrefixScanInit(DBBloomFilterTest* dbtest) {
  snprintf(buf, sizeof(buf), "%02d______:end", 10);
  keystr = std::string(buf);
  ASSERT_OK(dbtest->Put(keystr, keystr));
-  dbtest->Flush();
-  dbtest->dbfull()->CompactRange(CompactRangeOptions(), nullptr,
-                                 nullptr);  // move to level 1
+  ASSERT_OK(dbtest->Flush());
+  ASSERT_OK(dbtest->dbfull()->CompactRange(CompactRangeOptions(), nullptr,
+                                           nullptr));  // move to level 1

  // GROUP 1
  for (int i = 1; i <= small_range_sstfiles; i++) {
@@ -1563,21 +1563,21 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
  for (int key : keys) {
    ASSERT_OK(Put(1, Key(key), "val"));
    if (++num_inserted % 1000 == 0) {
-      dbfull()->TEST_WaitForFlushMemTable();
-      dbfull()->TEST_WaitForCompact();
+      ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+      ASSERT_OK(dbfull()->TEST_WaitForCompact());
    }
  }
  ASSERT_OK(Put(1, Key(0), "val"));
  ASSERT_OK(Put(1, Key(numkeys), "val"));
  ASSERT_OK(Flush(1));
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  if (NumTableFilesAtLevel(0, 1) == 0) {
    // No Level 0 file. Create one.
    ASSERT_OK(Put(1, Key(0), "val"));
    ASSERT_OK(Put(1, Key(numkeys), "val"));
    ASSERT_OK(Flush(1));
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
  }

  for (int i = 1; i < numkeys; i += 2) {
@@ -1682,7 +1682,8 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
      BottommostLevelCompaction::kSkip;
  compact_options.change_level = true;
  compact_options.target_level = 7;
-  db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);
+  ASSERT_TRUE(db_->CompactRange(compact_options, handles_[1], nullptr, nullptr)
+                  .IsNotSupported());

  ASSERT_EQ(trivial_move, 1);
  ASSERT_EQ(non_trivial_move, 0);
@@ -1714,10 +1715,10 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {

 int CountIter(std::unique_ptr<Iterator>& iter, const Slice& key) {
  int count = 0;
-  for (iter->Seek(key); iter->Valid() && iter->status() == Status::OK();
-       iter->Next()) {
+  for (iter->Seek(key); iter->Valid(); iter->Next()) {
    count++;
  }
+  EXPECT_OK(iter->status());
  return count;
 }

@@ -1747,7 +1748,7 @@ TEST_F(DBBloomFilterTest, DynamicBloomFilterUpperBound) {
    ASSERT_OK(Put("abcdxxx1", "val2"));
    ASSERT_OK(Put("abcdxxx2", "val3"));
    ASSERT_OK(Put("abcdxxx3", "val4"));
-    dbfull()->Flush(FlushOptions());
+    ASSERT_OK(dbfull()->Flush(FlushOptions()));
    {
      // prefix_extractor has not changed, BF will always be read
      Slice upper_bound("abce");
@@ -1905,7 +1906,7 @@ TEST_F(DBBloomFilterTest, DynamicBloomFilterMultipleSST) {
    ASSERT_OK(Put("foo4", "bar4"));
    ASSERT_OK(Put("foq5", "bar5"));
    ASSERT_OK(Put("fpb", "1"));
-    dbfull()->Flush(FlushOptions());
+    ASSERT_OK(dbfull()->Flush(FlushOptions()));
    {
      // BF is cappped:3 now
      std::unique_ptr<Iterator> iter_tmp(db_->NewIterator(read_options));
@@ -1929,7 +1930,7 @@ TEST_F(DBBloomFilterTest, DynamicBloomFilterMultipleSST) {
    ASSERT_OK(Put("foo7", "bar7"));
    ASSERT_OK(Put("foq8", "bar8"));
    ASSERT_OK(Put("fpc", "2"));
-    dbfull()->Flush(FlushOptions());
+    ASSERT_OK(dbfull()->Flush(FlushOptions()));
    {
      // BF is fixed:2 now
      std::unique_ptr<Iterator> iter_tmp(db_->NewIterator(read_options));
@@ -2040,10 +2041,10 @@ TEST_F(DBBloomFilterTest, DynamicBloomFilterNewColumnFamily) {
      ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_PREFIX_USEFUL), 0);
    }
    ASSERT_OK(dbfull()->DropColumnFamily(handles_[2]));
-    dbfull()->DestroyColumnFamilyHandle(handles_[2]);
+    ASSERT_OK(dbfull()->DestroyColumnFamilyHandle(handles_[2]));
    handles_[2] = nullptr;
    ASSERT_OK(dbfull()->DropColumnFamily(handles_[1]));
-    dbfull()->DestroyColumnFamilyHandle(handles_[1]);
+    ASSERT_OK(dbfull()->DestroyColumnFamilyHandle(handles_[1]));
    handles_[1] = nullptr;
    iteration++;
  }
@@ -7,6 +7,9 @@
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file. See the AUTHORS file for names of contributors.

+#include <tuple>
+
+#include "db/blob/blob_index.h"
 #include "db/db_test_util.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
@@ -28,6 +31,31 @@ class DBCompactionTest : public DBTestBase {
 public:
  DBCompactionTest()
      : DBTestBase("/db_compaction_test", /*env_do_fsync=*/true) {}
+
+  std::vector<uint64_t> GetBlobFileNumbers() {
+    VersionSet* const versions = dbfull()->TEST_GetVersionSet();
+    assert(versions);
+
+    ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
+    assert(cfd);
+
+    Version* const current = cfd->current();
+    assert(current);
+
+    const VersionStorageInfo* const storage_info = current->storage_info();
+    assert(storage_info);
+
+    const auto& blob_files = storage_info->GetBlobFiles();
+
+    std::vector<uint64_t> result;
+    result.reserve(blob_files.size());
+
+    for (const auto& blob_file : blob_files) {
+      result.emplace_back(blob_file.first);
+    }
+
+    return result;
+  }
 };

 class DBCompactionTestWithParam
@@ -6020,6 +6048,233 @@ TEST_P(DBCompactionTestBlobError, CompactionError) {
  }
 }

+class DBCompactionTestBlobGC
+    : public DBCompactionTest,
+      public testing::WithParamInterface<std::tuple<double, bool>> {
+ public:
+  DBCompactionTestBlobGC()
+      : blob_gc_age_cutoff_(std::get<0>(GetParam())),
+        updated_enable_blob_files_(std::get<1>(GetParam())) {}
+
+  double blob_gc_age_cutoff_;
+  bool updated_enable_blob_files_;
+};
+
+INSTANTIATE_TEST_CASE_P(DBCompactionTestBlobGC, DBCompactionTestBlobGC,
+                        ::testing::Combine(::testing::Values(0.0, 0.5, 1.0),
+                                           ::testing::Bool()));
+
+TEST_P(DBCompactionTestBlobGC, CompactionWithBlobGC) {
+  Options options;
+  options.env = env_;
+  options.disable_auto_compactions = true;
+  options.enable_blob_files = true;
+  options.blob_file_size = 32;  // one blob per file
+  options.enable_blob_garbage_collection = true;
+  options.blob_garbage_collection_age_cutoff = blob_gc_age_cutoff_;
+
+  Reopen(options);
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "first_value";
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "second_value";
+
+  ASSERT_OK(Put(first_key, first_value));
+  ASSERT_OK(Put(second_key, second_value));
+  ASSERT_OK(Flush());
+
+  constexpr char third_key[] = "third_key";
+  constexpr char third_value[] = "third_value";
+  constexpr char fourth_key[] = "fourth_key";
+  constexpr char fourth_value[] = "fourth_value";
+
+  ASSERT_OK(Put(third_key, third_value));
+  ASSERT_OK(Put(fourth_key, fourth_value));
+  ASSERT_OK(Flush());
+
+  const std::vector<uint64_t> original_blob_files = GetBlobFileNumbers();
+
+  ASSERT_EQ(original_blob_files.size(), 4);
+
+  const size_t cutoff_index = static_cast<size_t>(
+      options.blob_garbage_collection_age_cutoff * original_blob_files.size());
+
+  // Note: turning off enable_blob_files before the compaction results in
+  // garbage collected values getting inlined.
+  size_t expected_number_of_files = original_blob_files.size();
+
+  if (!updated_enable_blob_files_) {
+    ASSERT_OK(db_->SetOptions({{"enable_blob_files", "false"}}));
+
+    expected_number_of_files -= cutoff_index;
+  }
+
+  constexpr Slice* begin = nullptr;
+  constexpr Slice* end = nullptr;
+
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), begin, end));
+
+  ASSERT_EQ(Get(first_key), first_value);
+  ASSERT_EQ(Get(second_key), second_value);
+  ASSERT_EQ(Get(third_key), third_value);
+  ASSERT_EQ(Get(fourth_key), fourth_value);
+
+  const std::vector<uint64_t> new_blob_files = GetBlobFileNumbers();
+
+  ASSERT_EQ(new_blob_files.size(), expected_number_of_files);
+
+  // Original blob files below the cutoff should be gone, original blob files at
+  // or above the cutoff should be still there
+  for (size_t i = cutoff_index; i < original_blob_files.size(); ++i) {
+    ASSERT_EQ(new_blob_files[i - cutoff_index], original_blob_files[i]);
+  }
+}
+
+TEST_F(DBCompactionTest, CompactionWithBlobGCError_CorruptIndex) {
+  Options options;
+  options.env = env_;
+  options.disable_auto_compactions = true;
+  options.enable_blob_files = true;
+  options.enable_blob_garbage_collection = true;
+  options.blob_garbage_collection_age_cutoff = 1.0;
+
+  Reopen(options);
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "first_value";
+  ASSERT_OK(Put(first_key, first_value));
+
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "second_value";
+  ASSERT_OK(Put(second_key, second_value));
+
+  ASSERT_OK(Flush());
+
+  constexpr char third_key[] = "third_key";
+  constexpr char third_value[] = "third_value";
+  ASSERT_OK(Put(third_key, third_value));
+
+  constexpr char fourth_key[] = "fourth_key";
+  constexpr char corrupt_blob_index[] = "foobar";
+
+  WriteBatch batch;
+  ASSERT_OK(WriteBatchInternal::PutBlobIndex(&batch, 0, fourth_key,
+                                             corrupt_blob_index));
+  ASSERT_OK(db_->Write(WriteOptions(), &batch));
+
+  ASSERT_OK(Flush());
+
+  constexpr Slice* begin = nullptr;
+  constexpr Slice* end = nullptr;
+
+  ASSERT_TRUE(
+      db_->CompactRange(CompactRangeOptions(), begin, end).IsCorruption());
+}
+
+TEST_F(DBCompactionTest, CompactionWithBlobGCError_InlinedTTLIndex) {
+  constexpr uint64_t min_blob_size = 10;
+
+  Options options;
+  options.env = env_;
+  options.disable_auto_compactions = true;
+  options.enable_blob_files = true;
+  options.min_blob_size = min_blob_size;
+  options.enable_blob_garbage_collection = true;
+  options.blob_garbage_collection_age_cutoff = 1.0;
+
+  Reopen(options);
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "first_value";
+  ASSERT_OK(Put(first_key, first_value));
+
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "second_value";
+  ASSERT_OK(Put(second_key, second_value));
+
+  ASSERT_OK(Flush());
+
+  constexpr char third_key[] = "third_key";
+  constexpr char third_value[] = "third_value";
+  ASSERT_OK(Put(third_key, third_value));
+
+  constexpr char fourth_key[] = "fourth_key";
+  constexpr char blob[] = "short";
+  static_assert(sizeof(short) - 1 < min_blob_size,
+                "Blob too long to be inlined");
+
+  // Fake an inlined TTL blob index.
+  std::string blob_index;
+
+  constexpr uint64_t expiration = 1234567890;
+
+  BlobIndex::EncodeInlinedTTL(&blob_index, expiration, blob);
+
+  WriteBatch batch;
+  ASSERT_OK(
+      WriteBatchInternal::PutBlobIndex(&batch, 0, fourth_key, blob_index));
+  ASSERT_OK(db_->Write(WriteOptions(), &batch));
+
+  ASSERT_OK(Flush());
+
+  constexpr Slice* begin = nullptr;
+  constexpr Slice* end = nullptr;
+
+  ASSERT_TRUE(
+      db_->CompactRange(CompactRangeOptions(), begin, end).IsCorruption());
+}
+
+TEST_F(DBCompactionTest, CompactionWithBlobGCError_IndexWithInvalidFileNumber) {
+  Options options;
+  options.env = env_;
+  options.disable_auto_compactions = true;
+  options.enable_blob_files = true;
+  options.enable_blob_garbage_collection = true;
+  options.blob_garbage_collection_age_cutoff = 1.0;
+
+  Reopen(options);
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "first_value";
+  ASSERT_OK(Put(first_key, first_value));
+
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "second_value";
+  ASSERT_OK(Put(second_key, second_value));
+
+  ASSERT_OK(Flush());
+
+  constexpr char third_key[] = "third_key";
+  constexpr char third_value[] = "third_value";
+  ASSERT_OK(Put(third_key, third_value));
+
+  constexpr char fourth_key[] = "fourth_key";
+
+  // Fake a blob index referencing a non-existent blob file.
+  std::string blob_index;
+
+  constexpr uint64_t blob_file_number = 1000;
+  constexpr uint64_t offset = 1234;
+  constexpr uint64_t size = 5678;
+
+  BlobIndex::EncodeBlob(&blob_index, blob_file_number, offset, size,
+                        kNoCompression);
+
+  WriteBatch batch;
+  ASSERT_OK(
+      WriteBatchInternal::PutBlobIndex(&batch, 0, fourth_key, blob_index));
+  ASSERT_OK(db_->Write(WriteOptions(), &batch));
+
+  ASSERT_OK(Flush());
+
+  constexpr Slice* begin = nullptr;
+  constexpr Slice* end = nullptr;
+
+  ASSERT_TRUE(
+      db_->CompactRange(CompactRangeOptions(), begin, end).IsCorruption());
+}
+
 #endif  // !defined(ROCKSDB_LITE)

 }  // namespace ROCKSDB_NAMESPACE
@@ -14,6 +14,7 @@
 #include "file/filename.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
+#include "rocksdb/utilities/transaction_db.h"
 #include "test_util/sync_point.h"
 #include "util/cast_util.h"
 #include "util/mutexlock.h"
@@ -620,6 +621,96 @@ TEST_P(DBFlushTestBlobError, FlushError) {
 #endif  // ROCKSDB_LITE
 }

+#ifndef ROCKSDB_LITE
+TEST_P(DBAtomicFlushTest, ManualFlushUnder2PC) {
+  Options options = CurrentOptions();
+  options.create_if_missing = true;
+  options.allow_2pc = true;
+  options.atomic_flush = GetParam();
+  // 64MB so that memtable flush won't be trigger by the small writes.
+  options.write_buffer_size = (static_cast<size_t>(64) << 20);
+
+  // Destroy the DB to recreate as a TransactionDB.
+  Close();
+  Destroy(options, true);
+
+  // Create a TransactionDB.
+  TransactionDB* txn_db = nullptr;
+  TransactionDBOptions txn_db_opts;
+  txn_db_opts.write_policy = TxnDBWritePolicy::WRITE_COMMITTED;
+  ASSERT_OK(TransactionDB::Open(options, txn_db_opts, dbname_, &txn_db));
+  ASSERT_NE(txn_db, nullptr);
+  db_ = txn_db;
+
+  // Create two more columns other than default CF.
+  std::vector<std::string> cfs = {"puppy", "kitty"};
+  CreateColumnFamilies(cfs, options);
+  ASSERT_EQ(handles_.size(), 2);
+  ASSERT_EQ(handles_[0]->GetName(), cfs[0]);
+  ASSERT_EQ(handles_[1]->GetName(), cfs[1]);
+  const size_t kNumCfToFlush = options.atomic_flush ? 2 : 1;
+
+  WriteOptions wopts;
+  TransactionOptions txn_opts;
+  // txn1 only prepare, but does not commit.
+  // The WAL containing the prepared but uncommitted data must be kept.
+  Transaction* txn1 = txn_db->BeginTransaction(wopts, txn_opts, nullptr);
+  // txn2 not only prepare, but also commit.
+  Transaction* txn2 = txn_db->BeginTransaction(wopts, txn_opts, nullptr);
+  ASSERT_NE(txn1, nullptr);
+  ASSERT_NE(txn2, nullptr);
+  for (size_t i = 0; i < kNumCfToFlush; i++) {
+    ASSERT_OK(txn1->Put(handles_[i], "k1", "v1"));
+    ASSERT_OK(txn2->Put(handles_[i], "k2", "v2"));
+  }
+  // A txn must be named before prepare.
+  ASSERT_OK(txn1->SetName("txn1"));
+  ASSERT_OK(txn2->SetName("txn2"));
+  // Prepare writes to WAL, but not to memtable. (WriteCommitted)
+  ASSERT_OK(txn1->Prepare());
+  ASSERT_OK(txn2->Prepare());
+  // Commit writes to memtable.
+  ASSERT_OK(txn2->Commit());
+  delete txn1;
+  delete txn2;
+
+  // There are still data in memtable not flushed.
+  // But since data is small enough to reside in the active memtable,
+  // there are no immutable memtable.
+  for (size_t i = 0; i < kNumCfToFlush; i++) {
+    auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
+    ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
+    ASSERT_FALSE(cfh->cfd()->mem()->IsEmpty());
+  }
+
+  // Atomic flush memtables,
+  // the min log with prepared data should be written to MANIFEST.
+  std::vector<ColumnFamilyHandle*> cfs_to_flush(kNumCfToFlush);
+  for (size_t i = 0; i < kNumCfToFlush; i++) {
+    cfs_to_flush[i] = handles_[i];
+  }
+  ASSERT_OK(txn_db->Flush(FlushOptions(), cfs_to_flush));
+
+  // There are no remaining data in memtable after flush.
+  for (size_t i = 0; i < kNumCfToFlush; i++) {
+    auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
+    ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
+    ASSERT_TRUE(cfh->cfd()->mem()->IsEmpty());
+    ASSERT_EQ(cfh->cfd()->GetFlushReason(), FlushReason::kManualFlush);
+  }
+
+  // The recovered min log number with prepared data should be non-zero.
+  // In 2pc mode, MinLogNumberToKeep returns the
+  // VersionSet::min_log_number_to_keep_2pc recovered from MANIFEST, if it's 0,
+  // it means atomic flush didn't write the min_log_number_to_keep to MANIFEST.
+  cfs.push_back(kDefaultColumnFamilyName);
+  ASSERT_OK(TryReopenWithColumnFamilies(cfs, options));
+  DBImpl* db_impl = reinterpret_cast<DBImpl*>(db_);
+  ASSERT_TRUE(db_impl->allow_2pc());
+  ASSERT_NE(db_impl->MinLogNumberToKeep(), 0);
+}
+#endif  // ROCKSDB_LITE
+
 TEST_P(DBAtomicFlushTest, ManualAtomicFlush) {
  Options options = CurrentOptions();
  options.create_if_missing = true;
@@ -634,18 +725,84 @@ TEST_P(DBAtomicFlushTest, ManualAtomicFlush) {
  for (size_t i = 0; i != num_cfs; ++i) {
    ASSERT_OK(Put(static_cast<int>(i) /*cf*/, "key", "value", wopts));
  }
+
+  for (size_t i = 0; i != num_cfs; ++i) {
+    auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
+    ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
+    ASSERT_FALSE(cfh->cfd()->mem()->IsEmpty());
+  }
+
  std::vector<int> cf_ids;
  for (size_t i = 0; i != num_cfs; ++i) {
    cf_ids.emplace_back(static_cast<int>(i));
  }
  ASSERT_OK(Flush(cf_ids));
+
  for (size_t i = 0; i != num_cfs; ++i) {
    auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
+    ASSERT_EQ(cfh->cfd()->GetFlushReason(), FlushReason::kManualFlush);
    ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
    ASSERT_TRUE(cfh->cfd()->mem()->IsEmpty());
  }
 }

+TEST_P(DBAtomicFlushTest, PrecomputeMinLogNumberToKeepNon2PC) {
+  Options options = CurrentOptions();
+  options.create_if_missing = true;
+  options.atomic_flush = GetParam();
+  options.write_buffer_size = (static_cast<size_t>(64) << 20);
+  CreateAndReopenWithCF({"pikachu"}, options);
+
+  const size_t num_cfs = handles_.size();
+  ASSERT_EQ(num_cfs, 2);
+  WriteOptions wopts;
+  for (size_t i = 0; i != num_cfs; ++i) {
+    ASSERT_OK(Put(static_cast<int>(i) /*cf*/, "key", "value", wopts));
+  }
+
+  {
+    // Flush the default CF only.
+    std::vector<int> cf_ids{0};
+    ASSERT_OK(Flush(cf_ids));
+
+    autovector<ColumnFamilyData*> flushed_cfds;
+    autovector<autovector<VersionEdit*>> flush_edits;
+    auto flushed_cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[0]);
+    flushed_cfds.push_back(flushed_cfh->cfd());
+    flush_edits.push_back({});
+    auto unflushed_cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[1]);
+
+    ASSERT_EQ(PrecomputeMinLogNumberToKeepNon2PC(dbfull()->TEST_GetVersionSet(),
+                                                 flushed_cfds, flush_edits),
+              unflushed_cfh->cfd()->GetLogNumber());
+  }
+
+  {
+    // Flush all CFs.
+    std::vector<int> cf_ids;
+    for (size_t i = 0; i != num_cfs; ++i) {
+      cf_ids.emplace_back(static_cast<int>(i));
+    }
+    ASSERT_OK(Flush(cf_ids));
+    uint64_t log_num_after_flush = dbfull()->TEST_GetCurrentLogNumber();
+
+    uint64_t min_log_number_to_keep = port::kMaxUint64;
+    autovector<ColumnFamilyData*> flushed_cfds;
+    autovector<autovector<VersionEdit*>> flush_edits;
+    for (size_t i = 0; i != num_cfs; ++i) {
+      auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
+      flushed_cfds.push_back(cfh->cfd());
+      flush_edits.push_back({});
+      min_log_number_to_keep =
+          std::min(min_log_number_to_keep, cfh->cfd()->GetLogNumber());
+    }
+    ASSERT_EQ(min_log_number_to_keep, log_num_after_flush);
+    ASSERT_EQ(PrecomputeMinLogNumberToKeepNon2PC(dbfull()->TEST_GetVersionSet(),
+                                                 flushed_cfds, flush_edits),
+              min_log_number_to_keep);
+  }
+}
+
 TEST_P(DBAtomicFlushTest, AtomicFlushTriggeredByMemTableFull) {
  Options options = CurrentOptions();
  options.create_if_missing = true;
@@ -1323,7 +1323,8 @@ Status DBImpl::MarkLogsSynced(uint64_t up_to, bool synced_dir) {
    auto& wal = *it;
    assert(wal.getting_synced);
    if (logs_.size() > 1) {
-      if (immutable_db_options_.track_and_verify_wals_in_manifest) {
+      if (immutable_db_options_.track_and_verify_wals_in_manifest &&
+          wal.writer->file()->GetFileSize() > 0) {
        synced_wals.AddWal(wal.number,
                           WalMetadata(wal.writer->file()->GetFileSize()));
      }
@@ -2204,7 +2205,7 @@ void DBImpl::MultiGet(const ReadOptions& read_options, const size_t num_keys,
  for (; cf_iter != multiget_cf_data.end(); ++cf_iter) {
    s = MultiGetImpl(read_options, cf_iter->start, cf_iter->num_keys,
                     &sorted_keys, cf_iter->super_version, consistent_seqnum,
-                     read_callback, nullptr);
+                     read_callback);
    if (!s.ok()) {
      break;
    }
@@ -2376,7 +2377,7 @@ void DBImpl::MultiGetWithCallback(

  Status s = MultiGetImpl(read_options, 0, num_keys, sorted_keys,
                          multiget_cf_data[0].super_version, consistent_seqnum,
-                          read_callback, nullptr);
+                          read_callback);
  assert(s.ok() || s.IsTimedOut() || s.IsAborted());
  ReturnAndCleanupSuperVersion(multiget_cf_data[0].cfd,
                               multiget_cf_data[0].super_version);
@@ -2395,7 +2396,7 @@ Status DBImpl::MultiGetImpl(
    const ReadOptions& read_options, size_t start_key, size_t num_keys,
    autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE>* sorted_keys,
    SuperVersion* super_version, SequenceNumber snapshot,
-    ReadCallback* callback, bool* is_blob_index) {
+    ReadCallback* callback) {
  PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
  StopWatch sw(env_, stats_, DB_MULTIGET);

@@ -2434,11 +2435,9 @@ Status DBImpl::MultiGetImpl(
        (read_options.read_tier == kPersistedTier &&
         has_unpersisted_data_.load(std::memory_order_relaxed));
    if (!skip_memtable) {
-      super_version->mem->MultiGet(read_options, &range, callback,
-                                   is_blob_index);
+      super_version->mem->MultiGet(read_options, &range, callback);
      if (!range.empty()) {
-        super_version->imm->MultiGet(read_options, &range, callback,
-                                     is_blob_index);
+        super_version->imm->MultiGet(read_options, &range, callback);
      }
      if (!range.empty()) {
        lookup_current = true;
@@ -2448,8 +2447,7 @@ Status DBImpl::MultiGetImpl(
    }
    if (lookup_current) {
      PERF_TIMER_GUARD(get_from_output_files_time);
-      super_version->current->MultiGet(read_options, &range, callback,
-                                       is_blob_index);
+      super_version->current->MultiGet(read_options, &range, callback);
    }
    curr_value_size = range.GetValueSize();
    if (curr_value_size > read_options.value_size_soft_limit) {
@@ -2803,7 +2801,7 @@ Iterator* DBImpl::NewIterator(const ReadOptions& read_options,
                                    /* allow_unprepared_value */ true);
    result = NewDBIterator(
        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
-        cfd->user_comparator(), iter, kMaxSequenceNumber,
+        cfd->user_comparator(), iter, sv->current, kMaxSequenceNumber,
        sv->mutable_cf_options.max_sequential_skip_in_iterations, read_callback,
        this, cfd);
 #endif
@@ -2824,7 +2822,7 @@ ArenaWrappedDBIter* DBImpl::NewIteratorImpl(const ReadOptions& read_options,
                                            ColumnFamilyData* cfd,
                                            SequenceNumber snapshot,
                                            ReadCallback* read_callback,
-                                            bool allow_blob,
+                                            bool expose_blob_index,
                                            bool allow_refresh) {
  SuperVersion* sv = cfd->GetReferencedSuperVersion(this);

@@ -2889,9 +2887,9 @@ ArenaWrappedDBIter* DBImpl::NewIteratorImpl(const ReadOptions& read_options,
  // likely that any iterator pointer is close to the iterator it points to so
  // that they are likely to be in the same cache line and/or page.
  ArenaWrappedDBIter* db_iter = NewArenaWrappedDbIterator(
-      env_, read_options, *cfd->ioptions(), sv->mutable_cf_options, snapshot,
-      sv->mutable_cf_options.max_sequential_skip_in_iterations,
-      sv->version_number, read_callback, this, cfd, allow_blob,
+      env_, read_options, *cfd->ioptions(), sv->mutable_cf_options, sv->current,
+      snapshot, sv->mutable_cf_options.max_sequential_skip_in_iterations,
+      sv->version_number, read_callback, this, cfd, expose_blob_index,
      read_options.snapshot != nullptr ? false : allow_refresh);

  InternalIterator* internal_iter = NewInternalIterator(
@@ -2929,7 +2927,7 @@ Status DBImpl::NewIterators(
                                      /* allow_unprepared_value */ true);
      iterators->push_back(NewDBIterator(
          env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
-          cfd->user_comparator(), iter, kMaxSequenceNumber,
+          cfd->user_comparator(), iter, sv->current, kMaxSequenceNumber,
          sv->mutable_cf_options.max_sequential_skip_in_iterations,
          read_callback, this, cfd));
    }
@@ -3400,6 +3398,10 @@ Status DBImpl::GetApproximateSizes(const SizeApproximationOptions& options,
    return Status::InvalidArgument("Invalid options");
  }

+  const Comparator* const ucmp = column_family->GetComparator();
+  assert(ucmp);
+  size_t ts_sz = ucmp->timestamp_size();
+
  Version* v;
  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
@@ -3407,9 +3409,23 @@ Status DBImpl::GetApproximateSizes(const SizeApproximationOptions& options,
  v = sv->current;

  for (int i = 0; i < n; i++) {
+    Slice start = range[i].start;
+    Slice limit = range[i].limit;
+
+    // Add timestamp if needed
+    std::string start_with_ts, limit_with_ts;
+    if (ts_sz > 0) {
+      // Maximum timestamp means including all key with any timestamp
+      AppendKeyWithMaxTimestamp(&start_with_ts, start, ts_sz);
+      // Append a maximum timestamp as the range limit is exclusive:
+      // [start, limit)
+      AppendKeyWithMaxTimestamp(&limit_with_ts, limit, ts_sz);
+      start = start_with_ts;
+      limit = limit_with_ts;
+    }
    // Convert user_key into a corresponding internal key.
-    InternalKey k1(range[i].start, kMaxSequenceNumber, kValueTypeForSeek);
-    InternalKey k2(range[i].limit, kMaxSequenceNumber, kValueTypeForSeek);
+    InternalKey k1(start, kMaxSequenceNumber, kValueTypeForSeek);
+    InternalKey k2(limit, kMaxSequenceNumber, kValueTypeForSeek);
    sizes[i] = 0;
    if (options.include_files) {
      sizes[i] += versions_->ApproximateSize(
@@ -3467,7 +3483,6 @@ Status DBImpl::DeleteFile(std::string name) {
    return Status::InvalidArgument("Invalid file name");
  }

-  Status status;
  if (type == kWalFile) {
    // Only allow deleting archived log files
    if (log_type != kArchivedLogFile) {
@@ -3476,7 +3491,7 @@ Status DBImpl::DeleteFile(std::string name) {
                      name.c_str());
      return Status::NotSupported("Delete only supported for archived logs");
    }
-    status = wal_manager_.DeleteFile(name, number);
+    Status status = wal_manager_.DeleteFile(name, number);
    if (!status.ok()) {
      ROCKS_LOG_ERROR(immutable_db_options_.info_log,
                      "DeleteFile %s failed -- %s.\n", name.c_str(),
@@ -3485,6 +3500,7 @@ Status DBImpl::DeleteFile(std::string name) {
    return status;
  }

+  Status status;
  int level;
  FileMetaData* metadata;
  ColumnFamilyData* cfd;
@@ -4338,7 +4354,7 @@ Status DBImpl::IngestExternalFiles(
    }
  }
  // Ingest multiple external SST files atomically.
-  size_t num_cfs = args.size();
+  const size_t num_cfs = args.size();
  for (size_t i = 0; i != num_cfs; ++i) {
    if (args[i].external_files.empty()) {
      char err_msg[128] = {0};
@@ -4379,10 +4395,7 @@ Status DBImpl::IngestExternalFiles(
        env_, versions_.get(), cfd, immutable_db_options_, file_options_,
        &snapshots_, arg.options, &directories_, &event_logger_, io_tracer_);
  }
-  std::vector<std::pair<bool, Status>> exec_results;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    exec_results.emplace_back(false, Status::OK());
-  }
+
  // TODO(yanqin) maybe make jobs run in parallel
  uint64_t start_file_number = next_file_number;
  for (size_t i = 1; i != num_cfs; ++i) {
@@ -4390,10 +4403,13 @@ Status DBImpl::IngestExternalFiles(
    auto* cfd =
        static_cast<ColumnFamilyHandleImpl*>(args[i].column_family)->cfd();
    SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
-    exec_results[i].second = ingestion_jobs[i].Prepare(
+    Status es = ingestion_jobs[i].Prepare(
        args[i].external_files, args[i].files_checksums,
        args[i].files_checksum_func_names, start_file_number, super_version);
-    exec_results[i].first = true;
+    // capture first error only
+    if (!es.ok() && status.ok()) {
+      status = es;
+    }
    CleanupSuperVersion(super_version);
  }
  TEST_SYNC_POINT("DBImpl::IngestExternalFiles:BeforeLastJobPrepare:0");
@@ -4402,23 +4418,17 @@ Status DBImpl::IngestExternalFiles(
    auto* cfd =
        static_cast<ColumnFamilyHandleImpl*>(args[0].column_family)->cfd();
    SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
-    exec_results[0].second = ingestion_jobs[0].Prepare(
+    Status es = ingestion_jobs[0].Prepare(
        args[0].external_files, args[0].files_checksums,
        args[0].files_checksum_func_names, next_file_number, super_version);
-    exec_results[0].first = true;
-    CleanupSuperVersion(super_version);
-  }
-  for (const auto& exec_result : exec_results) {
-    if (!exec_result.second.ok()) {
-      status = exec_result.second;
-      break;
+    if (!es.ok()) {
+      status = es;
    }
+    CleanupSuperVersion(super_version);
  }
  if (!status.ok()) {
    for (size_t i = 0; i != num_cfs; ++i) {
-      if (exec_results[i].first) {
-        ingestion_jobs[i].Cleanup(status);
-      }
+      ingestion_jobs[i].Cleanup(status);
    }
    InstrumentedMutexLock l(&mutex_);
    ReleaseFileNumberFromPendingOutputs(pending_output_elem);
@@ -4588,8 +4598,7 @@ Status DBImpl::IngestExternalFiles(
      // TODO: distinguish between MANIFEST write and CURRENT renaming
      const IOStatus& io_s = versions_->io_status();
      // Should handle return error?
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite);
    }

    // Resume writes to the DB
@@ -4863,7 +4872,7 @@ Status DBImpl::VerifySstFileChecksum(const FileMetaData& fmeta,
      fs_.get(), fname, immutable_db_options_.file_checksum_gen_factory.get(),
      fmeta.file_checksum_func_name, &file_checksum, &func_name,
      read_options.readahead_size, immutable_db_options_.allow_mmap_reads,
-      io_tracer_);
+      io_tracer_, immutable_db_options_.rate_limiter.get());
  if (s.ok()) {
    assert(fmeta.file_checksum_func_name == func_name);
    if (file_checksum != fmeta.file_checksum) {
@@ -522,7 +522,7 @@ class DBImpl : public DB {
                                      ColumnFamilyData* cfd,
                                      SequenceNumber snapshot,
                                      ReadCallback* read_callback,
-                                      bool allow_blob = false,
+                                      bool expose_blob_index = false,
                                      bool allow_refresh = true);

  virtual SequenceNumber GetLastPublishedSequence() const {
@@ -922,7 +922,7 @@ class DBImpl : public DB {
                           ColumnFamilyHandle* column_family = nullptr,
                           bool disallow_trivial_move = false);

-  void TEST_SwitchWAL();
+  Status TEST_SwitchWAL();

  bool TEST_UnableToReleaseOldestLog() { return unable_to_release_oldest_log_; }

@@ -1017,6 +1017,12 @@ class DBImpl : public DB {

  VersionSet* TEST_GetVersionSet() const { return versions_.get(); }

+  uint64_t TEST_GetCurrentLogNumber() const {
+    InstrumentedMutexLock l(mutex());
+    assert(!logs_.empty());
+    return logs_.back().number;
+  }
+
  const std::unordered_set<uint64_t>& TEST_GetFilesGrabbedForPurge() const {
    return files_grabbed_for_purge_;
  }
@@ -1104,6 +1110,15 @@ class DBImpl : public DB {
  // If need_enter_write_thread = false, the method will enter write thread.
  Status WriteOptionsFile(bool need_mutex_lock, bool need_enter_write_thread);

+  Status CompactRangeInternal(const CompactRangeOptions& options,
+                              ColumnFamilyHandle* column_family,
+                              const Slice* begin, const Slice* end);
+
+  Status GetApproximateSizesInternal(const SizeApproximationOptions& options,
+                                     ColumnFamilyHandle* column_family,
+                                     const Range* range, int n,
+                                     uint64_t* sizes);
+
  // The following two functions can only be called when:
  // 1. WriteThread::Writer::EnterUnbatched() is used.
  // 2. db_mutex is NOT held
@@ -1851,8 +1866,7 @@ class DBImpl : public DB {
  Status MultiGetImpl(
      const ReadOptions& read_options, size_t start_key, size_t num_keys,
      autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE>* sorted_keys,
-      SuperVersion* sv, SequenceNumber snap_seqnum, ReadCallback* callback,
-      bool* is_blob_index);
+      SuperVersion* sv, SequenceNumber snap_seqnum, ReadCallback* callback);

  Status DisableFileDeletionsWithLock();

@@ -2219,12 +2233,22 @@ extern uint64_t PrecomputeMinLogNumberToKeep2PC(
    const autovector<VersionEdit*>& edit_list,
    const autovector<MemTable*>& memtables_to_flush,
    LogsWithPrepTracker* prep_tracker);
+// For atomic flush.
+extern uint64_t PrecomputeMinLogNumberToKeep2PC(
+    VersionSet* vset, const autovector<ColumnFamilyData*>& cfds_to_flush,
+    const autovector<autovector<VersionEdit*>>& edit_lists,
+    const autovector<const autovector<MemTable*>*>& memtables_to_flush,
+    LogsWithPrepTracker* prep_tracker);

 // In non-2PC mode, WALs with log number < the returned number can be
 // deleted after the cfd_to_flush column family is flushed successfully.
 extern uint64_t PrecomputeMinLogNumberToKeepNon2PC(
    VersionSet* vset, const ColumnFamilyData& cfd_to_flush,
    const autovector<VersionEdit*>& edit_list);
+// For atomic flush.
+extern uint64_t PrecomputeMinLogNumberToKeepNon2PC(
+    VersionSet* vset, const autovector<ColumnFamilyData*>& cfds_to_flush,
+    const autovector<autovector<VersionEdit*>>& edit_lists);

 // `cfd_to_flush` is the column family whose memtable will be flushed and thus
 // will not depend on any WAL file. nullptr means no memtable is being flushed.
@@ -2232,6 +2256,10 @@ extern uint64_t PrecomputeMinLogNumberToKeepNon2PC(
 extern uint64_t FindMinPrepLogReferencedByMemTable(
    VersionSet* vset, const ColumnFamilyData* cfd_to_flush,
    const autovector<MemTable*>& memtables_to_flush);
+// For atomic flush.
+extern uint64_t FindMinPrepLogReferencedByMemTable(
+    VersionSet* vset, const autovector<ColumnFamilyData*>& cfds_to_flush,
+    const autovector<const autovector<MemTable*>*>& memtables_to_flush);

 // Fix user-supplied options to be reasonable
 template <class T, class V>
@@ -130,12 +130,10 @@ IOStatus DBImpl::SyncClosedLogs(JobContext* job_context) {
    }
    if (!io_s.ok()) {
      if (total_log_size_ > 0) {
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
      } else {
        // If the WAL is empty, we use different error reason
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
      }
      TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Failed");
      return io_s;
@@ -159,14 +157,14 @@ Status DBImpl::FlushMemTableToOutputFile(

  FlushJob flush_job(
      dbname_, cfd, immutable_db_options_, mutable_cf_options,
-      nullptr /* memtable_id */, file_options_for_compaction_, versions_.get(),
-      &mutex_, &shutting_down_, snapshot_seqs, earliest_write_conflict_snapshot,
-      snapshot_checker, job_context, log_buffer, directories_.GetDbDir(),
-      GetDataDir(cfd, 0U),
+      port::kMaxUint64 /* memtable_id */, file_options_for_compaction_,
+      versions_.get(), &mutex_, &shutting_down_, snapshot_seqs,
+      earliest_write_conflict_snapshot, snapshot_checker, job_context,
+      log_buffer, directories_.GetDbDir(), GetDataDir(cfd, 0U),
      GetCompressionFlush(*cfd->ioptions(), mutable_cf_options), stats_,
      &event_logger_, mutable_cf_options.report_bg_io_stats,
      true /* sync_output_directory */, true /* write_manifest */, thread_pri,
-      io_tracer_, db_id_, db_session_id_);
+      io_tracer_, db_id_, db_session_id_, cfd->GetFullHistoryTsLow());
  FileMetaData file_meta;

  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:BeforePickMemtables");
@@ -192,8 +190,7 @@ Status DBImpl::FlushMemTableToOutputFile(
    s = io_s;
    if (!io_s.ok() && !io_s.IsShutdownInProgress() &&
        !io_s.IsColumnFamilyDropped()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
    }
  } else {
    TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Skip");
@@ -253,24 +250,23 @@ Status DBImpl::FlushMemTableToOutputFile(
      // be pessimistic and try write to a new MANIFEST.
      // TODO: distinguish between MANIFEST write and CURRENT renaming
      if (!versions_->io_status().ok()) {
-        // Should handle return error?
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite)
-            .PermitUncheckedError();
+        if (total_log_size_ > 0) {
+          // If the WAL is empty, we use different error reason
+          error_handler_.SetBGError(io_s,
+                                    BackgroundErrorReason::kManifestWrite);
+        } else {
+          error_handler_.SetBGError(io_s,
+                                    BackgroundErrorReason::kManifestWriteNoWAL);
+        }
      } else if (total_log_size_ > 0) {
-        // Should handle return error?
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
      } else {
        // If the WAL is empty, we use different error reason
-        // Should handle return error?
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
      }
    } else {
      Status new_bg_error = s;
-      // Should handle return error?
-      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
    }
  } else {
    // If we got here, then we decided not to care about the i_os status (either
@@ -288,16 +284,14 @@ Status DBImpl::FlushMemTableToOutputFile(
      // Notify sst_file_manager that a new file was added
      std::string file_path = MakeTableFileName(
          cfd->ioptions()->cf_paths[0].path, file_meta.fd.GetNumber());
-      sfm->OnAddFile(file_path);
+      s = sfm->OnAddFile(file_path);
      if (sfm->IsMaxAllowedSpaceReached()) {
        Status new_bg_error =
            Status::SpaceLimit("Max allowed space was reached");
        TEST_SYNC_POINT_CALLBACK(
            "DBImpl::FlushMemTableToOutputFile:MaxAllowedSpaceReached",
            &new_bg_error);
-        // Should handle this error?
-        error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
      }
    }
 #endif  // ROCKSDB_LITE
@@ -313,30 +307,22 @@ Status DBImpl::FlushMemTablesToOutputFiles(
    return AtomicFlushMemTablesToOutputFiles(
        bg_flush_args, made_progress, job_context, log_buffer, thread_pri);
  }
+  assert(bg_flush_args.size() == 1);
  std::vector<SequenceNumber> snapshot_seqs;
  SequenceNumber earliest_write_conflict_snapshot;
  SnapshotChecker* snapshot_checker;
  GetSnapshotContext(job_context, &snapshot_seqs,
                     &earliest_write_conflict_snapshot, &snapshot_checker);
-  Status status;
-  for (auto& arg : bg_flush_args) {
-    ColumnFamilyData* cfd = arg.cfd_;
-    MutableCFOptions mutable_cf_options = *cfd->GetLatestMutableCFOptions();
-    SuperVersionContext* superversion_context = arg.superversion_context_;
-    Status s = FlushMemTableToOutputFile(
-        cfd, mutable_cf_options, made_progress, job_context,
-        superversion_context, snapshot_seqs, earliest_write_conflict_snapshot,
-        snapshot_checker, log_buffer, thread_pri);
-    if (!s.ok()) {
-      status = s;
-      if (!s.IsShutdownInProgress() && !s.IsColumnFamilyDropped()) {
-        // At this point, DB is not shutting down, nor is cfd dropped.
-        // Something is wrong, thus we break out of the loop.
-        break;
-      }
-    }
-  }
-  return status;
+  const auto& bg_flush_arg = bg_flush_args[0];
+  ColumnFamilyData* cfd = bg_flush_arg.cfd_;
+  MutableCFOptions mutable_cf_options = *cfd->GetLatestMutableCFOptions();
+  SuperVersionContext* superversion_context =
+      bg_flush_arg.superversion_context_;
+  Status s = FlushMemTableToOutputFile(
+      cfd, mutable_cf_options, made_progress, job_context, superversion_context,
+      snapshot_seqs, earliest_write_conflict_snapshot, snapshot_checker,
+      log_buffer, thread_pri);
+  return s;
 }

 /*
@@ -399,7 +385,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(

    all_mutable_cf_options.emplace_back(*cfd->GetLatestMutableCFOptions());
    const MutableCFOptions& mutable_cf_options = all_mutable_cf_options.back();
-    const uint64_t* max_memtable_id = &(bg_flush_args[i].max_memtable_id_);
+    uint64_t max_memtable_id = bg_flush_args[i].max_memtable_id_;
    jobs.emplace_back(new FlushJob(
        dbname_, cfd, immutable_db_options_, mutable_cf_options,
        max_memtable_id, file_options_for_compaction_, versions_.get(), &mutex_,
@@ -408,7 +394,8 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
        data_dir, GetCompressionFlush(*cfd->ioptions(), mutable_cf_options),
        stats_, &event_logger_, mutable_cf_options.report_bg_io_stats,
        false /* sync_output_directory */, false /* write_manifest */,
-        thread_pri, io_tracer_, db_id_, db_session_id_));
+        thread_pri, io_tracer_, db_id_, db_session_id_,
+        cfd->GetFullHistoryTsLow()));
    jobs.back()->PickMemTable();
  }

@@ -585,7 +572,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(

    s = InstallMemtableAtomicFlushResults(
        nullptr /* imm_lists */, tmp_cfds, mutable_cf_options_list, mems_list,
-        versions_.get(), &mutex_, tmp_file_meta,
+        versions_.get(), &logs_with_prep_tracker_, &mutex_, tmp_file_meta,
        &job_context->memtables_to_free, directories_.GetDbDir(), log_buffer);
  }

@@ -631,7 +618,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
    auto sfm = static_cast<SstFileManagerImpl*>(
        immutable_db_options_.sst_file_manager.get());
    assert(all_mutable_cf_options.size() == static_cast<size_t>(num_cfs));
-    for (int i = 0; i != num_cfs; ++i) {
+    for (int i = 0; s.ok() && i != num_cfs; ++i) {
      if (cfds[i]->IsDropped()) {
        continue;
      }
@@ -640,14 +627,13 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
      if (sfm) {
        std::string file_path = MakeTableFileName(
            cfds[i]->ioptions()->cf_paths[0].path, file_meta[i].fd.GetNumber());
-        sfm->OnAddFile(file_path);
+        s = sfm->OnAddFile(file_path);
        if (sfm->IsMaxAllowedSpaceReached() &&
            error_handler_.GetBGError().ok()) {
          Status new_bg_error =
              Status::SpaceLimit("Max allowed space was reached");
-          // Should Handle this error?
-          error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush)
-              .PermitUncheckedError();
+          error_handler_.SetBGError(new_bg_error,
+                                    BackgroundErrorReason::kFlush);
        }
      }
    }
@@ -664,24 +650,23 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
      // be pessimistic and try write to a new MANIFEST.
      // TODO: distinguish between MANIFEST write and CURRENT renaming
      if (!versions_->io_status().ok()) {
-        // Should Handle this error?
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite)
-            .PermitUncheckedError();
+        if (total_log_size_ > 0) {
+          // If the WAL is empty, we use different error reason
+          error_handler_.SetBGError(io_s,
+                                    BackgroundErrorReason::kManifestWrite);
+        } else {
+          error_handler_.SetBGError(io_s,
+                                    BackgroundErrorReason::kManifestWriteNoWAL);
+        }
      } else if (total_log_size_ > 0) {
-        // Should Handle this error?
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
      } else {
        // If the WAL is empty, we use different error reason
-        // Should Handle this error?
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL)
-            .PermitUncheckedError();
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
      }
    } else {
      Status new_bg_error = s;
-      // Should Handle this error?
-      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
    }
  }

@@ -781,7 +766,41 @@ void DBImpl::NotifyOnFlushCompleted(

 Status DBImpl::CompactRange(const CompactRangeOptions& options,
                            ColumnFamilyHandle* column_family,
-                            const Slice* begin, const Slice* end) {
+                            const Slice* begin_without_ts,
+                            const Slice* end_without_ts) {
+  const Comparator* const ucmp = column_family->GetComparator();
+  assert(ucmp);
+  size_t ts_sz = ucmp->timestamp_size();
+  if (ts_sz == 0) {
+    return CompactRangeInternal(options, column_family, begin_without_ts,
+                                end_without_ts);
+  }
+
+  std::string begin_str;
+  std::string end_str;
+
+  // CompactRange compact all keys: [begin, end] inclusively. Add maximum
+  // timestamp to include all `begin` keys, and add minimal timestamp to include
+  // all `end` keys.
+  if (begin_without_ts != nullptr) {
+    AppendKeyWithMaxTimestamp(&begin_str, *begin_without_ts, ts_sz);
+  }
+  if (end_without_ts != nullptr) {
+    AppendKeyWithMinTimestamp(&end_str, *end_without_ts, ts_sz);
+  }
+  Slice begin(begin_str);
+  Slice end(end_str);
+
+  Slice* begin_with_ts = begin_without_ts ? &begin : nullptr;
+  Slice* end_with_ts = end_without_ts ? &end : nullptr;
+
+  return CompactRangeInternal(options, column_family, begin_with_ts,
+                              end_with_ts);
+}
+
+Status DBImpl::CompactRangeInternal(const CompactRangeOptions& options,
+                                    ColumnFamilyHandle* column_family,
+                                    const Slice* begin, const Slice* end) {
  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();

@@ -790,20 +809,20 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
  }

  bool flush_needed = true;
+  Status s;
  if (begin != nullptr && end != nullptr) {
    // TODO(ajkr): We could also optimize away the flush in certain cases where
    // one/both sides of the interval are unbounded. But it requires more
    // changes to RangesOverlapWithMemtables.
    Range range(*begin, *end);
    SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
-    cfd->RangesOverlapWithMemtables({range}, super_version,
-                                    immutable_db_options_.allow_data_in_errors,
-                                    &flush_needed);
+    s = cfd->RangesOverlapWithMemtables(
+        {range}, super_version, immutable_db_options_.allow_data_in_errors,
+        &flush_needed);
    CleanupSuperVersion(super_version);
  }

-  Status s;
-  if (flush_needed) {
+  if (s.ok() && flush_needed) {
    FlushOptions fo;
    fo.allow_write_stall = options.allow_write_stall;
    if (immutable_db_options_.atomic_flush) {
@@ -1160,7 +1179,8 @@ Status DBImpl::CompactFilesImpl(
      c->mutable_cf_options()->paranoid_file_checks,
      c->mutable_cf_options()->report_bg_io_stats, dbname_,
      &compaction_job_stats, Env::Priority::USER, io_tracer_,
-      &manual_compaction_paused_, db_id_, db_session_id_);
+      &manual_compaction_paused_, db_id_, db_session_id_,
+      c->column_family_data()->GetFullHistoryTsLow());

  // Creating a compaction influences the compaction score because the score
  // takes running compactions into account (by skipping files that are already
@@ -1174,7 +1194,8 @@ Status DBImpl::CompactFilesImpl(
  mutex_.Unlock();
  TEST_SYNC_POINT("CompactFilesImpl:0");
  TEST_SYNC_POINT("CompactFilesImpl:1");
-  compaction_job.Run();
+  // Ignore the status here, as it will be checked in the Install down below...
+  compaction_job.Run().PermitUncheckedError();
  TEST_SYNC_POINT("CompactFilesImpl:2");
  TEST_SYNC_POINT("CompactFilesImpl:3");
  mutex_.Lock();
@@ -1224,11 +1245,9 @@ Status DBImpl::CompactFilesImpl(
                   job_context->job_id, status.ToString().c_str());
    IOStatus io_s = compaction_job.io_status();
    if (!io_s.ok()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kCompaction)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kCompaction);
    } else {
-      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
    }
  }

@@ -1373,8 +1392,6 @@ Status DBImpl::ReFitLevel(ColumnFamilyData* cfd, int level, int target_level) {

  SuperVersionContext sv_context(/* create_superversion */ true);

-  Status status;
-
  InstrumentedMutexLock guard_lock(&mutex_);

  // only allow one thread refitting
@@ -1438,8 +1455,9 @@ Status DBImpl::ReFitLevel(ColumnFamilyData* cfd, int level, int target_level) {
                    "[%s] Apply version edit:\n%s", cfd->GetName().c_str(),
                    edit.DebugString().data());

-    status = versions_->LogAndApply(cfd, mutable_cf_options, &edit, &mutex_,
-                                    directories_.GetDbDir());
+    Status status = versions_->LogAndApply(cfd, mutable_cf_options, &edit,
+                                           &mutex_, directories_.GetDbDir());
+
    InstallSuperVersionAndScheduleWork(cfd, &sv_context, mutable_cf_options);

    ROCKS_LOG_DEBUG(immutable_db_options_.info_log, "[%s] LogAndApply: %s\n",
@@ -1450,12 +1468,14 @@ Status DBImpl::ReFitLevel(ColumnFamilyData* cfd, int level, int target_level) {
                      "[%s] After refitting:\n%s", cfd->GetName().c_str(),
                      cfd->current()->DebugString().data());
    }
+    sv_context.Clean();
+    refitting_level_ = false;
+
+    return status;
  }

-  sv_context.Clean();
  refitting_level_ = false;
-
-  return status;
+  return Status::OK();
 }

 int DBImpl::NumberLevels(ColumnFamilyHandle* column_family) {
@@ -1697,8 +1717,9 @@ void DBImpl::GenerateFlushRequest(const autovector<ColumnFamilyData*>& cfds,
 Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
                             const FlushOptions& flush_options,
                             FlushReason flush_reason, bool writes_stopped) {
+  // This method should not be called if atomic_flush is true.
+  assert(!immutable_db_options_.atomic_flush);
  Status s;
-  uint64_t flush_memtable_id = 0;
  if (!flush_options.allow_write_stall) {
    bool flush_needed = true;
    s = WaitUntilFlushWouldNotStallWrites(cfd, &flush_needed);
@@ -1708,7 +1729,8 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
    }
  }

-  FlushRequest flush_req;
+  autovector<FlushRequest> flush_reqs;
+  autovector<uint64_t> memtable_ids_to_wait;
  {
    WriteContext context;
    InstrumentedMutexLock guard_lock(&mutex_);
@@ -1723,18 +1745,24 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
    }
    WaitForPendingWrites();

-    if (!cfd->mem()->IsEmpty() || !cached_recoverable_state_empty_.load()) {
-      if (flush_reason != FlushReason::kErrorRecoveryRetryFlush) {
-        s = SwitchMemtable(cfd, &context);
-      } else {
-        assert(cfd->imm()->NumNotFlushed() > 0);
-      }
+    if (flush_reason != FlushReason::kErrorRecoveryRetryFlush &&
+        (!cfd->mem()->IsEmpty() || !cached_recoverable_state_empty_.load())) {
+      // Note that, when flush reason is kErrorRecoveryRetryFlush, during the
+      // auto retry resume, we want to avoid creating new small memtables.
+      // Therefore, SwitchMemtable will not be called. Also, since ResumeImpl
+      // will iterate through all the CFs and call FlushMemtable during auto
+      // retry resume, it is possible that in some CFs,
+      // cfd->imm()->NumNotFlushed() = 0. In this case, so no flush request will
+      // be created and scheduled, status::OK() will be returned.
+      s = SwitchMemtable(cfd, &context);
    }
+    const uint64_t flush_memtable_id = port::kMaxUint64;
    if (s.ok()) {
      if (cfd->imm()->NumNotFlushed() != 0 || !cfd->mem()->IsEmpty() ||
          !cached_recoverable_state_empty_.load()) {
-        flush_memtable_id = cfd->imm()->GetLatestMemTableID();
-        flush_req.emplace_back(cfd, flush_memtable_id);
+        FlushRequest req{{cfd, flush_memtable_id}};
+        flush_reqs.emplace_back(std::move(req));
+        memtable_ids_to_wait.emplace_back(cfd->imm()->GetLatestMemTableID());
      }
      if (immutable_db_options_.persist_stats_to_disk &&
          flush_reason != FlushReason::kErrorRecoveryRetryFlush) {
@@ -1760,15 +1788,19 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
                           "to avoid holding old logs",
                           cfd->GetName().c_str());
            s = SwitchMemtable(cfd_stats, &context);
-            flush_memtable_id = cfd_stats->imm()->GetLatestMemTableID();
-            flush_req.emplace_back(cfd_stats, flush_memtable_id);
+            FlushRequest req{{cfd_stats, flush_memtable_id}};
+            flush_reqs.emplace_back(std::move(req));
+            memtable_ids_to_wait.emplace_back(
+                cfd->imm()->GetLatestMemTableID());
          }
        }
      }
    }
-    if (s.ok() && !flush_req.empty()) {
-      for (auto& elem : flush_req) {
-        ColumnFamilyData* loop_cfd = elem.first;
+
+    if (s.ok() && !flush_reqs.empty()) {
+      for (const auto& req : flush_reqs) {
+        assert(req.size() == 1);
+        ColumnFamilyData* loop_cfd = req[0].first;
        loop_cfd->imm()->FlushRequested();
      }
      // If the caller wants to wait for this flush to complete, it indicates
@@ -1776,12 +1808,15 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
      // other threads which may drop the column family concurrently.
      // Therefore, we increase the cfd's ref count.
      if (flush_options.wait) {
-        for (auto& elem : flush_req) {
-          ColumnFamilyData* loop_cfd = elem.first;
+        for (const auto& req : flush_reqs) {
+          assert(req.size() == 1);
+          ColumnFamilyData* loop_cfd = req[0].first;
          loop_cfd->Ref();
        }
      }
-      SchedulePendingFlush(flush_req, flush_reason);
+      for (const auto& req : flush_reqs) {
+        SchedulePendingFlush(req, flush_reason);
+      }
      MaybeScheduleFlushOrCompaction();
    }

@@ -1797,9 +1832,11 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
  if (s.ok() && flush_options.wait) {
    autovector<ColumnFamilyData*> cfds;
    autovector<const uint64_t*> flush_memtable_ids;
-    for (auto& iter : flush_req) {
-      cfds.push_back(iter.first);
-      flush_memtable_ids.push_back(&(iter.second));
+    assert(flush_reqs.size() == memtable_ids_to_wait.size());
+    for (size_t i = 0; i < flush_reqs.size(); ++i) {
+      assert(flush_reqs[i].size() == 1);
+      cfds.push_back(flush_reqs[i][0].first);
+      flush_memtable_ids.push_back(&(memtable_ids_to_wait[i]));
    }
    s = WaitForFlushMemTables(
        cfds, flush_memtable_ids,
@@ -2224,6 +2261,17 @@ DBImpl::FlushRequest DBImpl::PopFirstFromFlushQueue() {
  assert(!flush_queue_.empty());
  FlushRequest flush_req = flush_queue_.front();
  flush_queue_.pop_front();
+  if (!immutable_db_options_.atomic_flush) {
+    assert(flush_req.size() == 1);
+  }
+  for (const auto& elem : flush_req) {
+    if (!immutable_db_options_.atomic_flush) {
+      ColumnFamilyData* cfd = elem.first;
+      assert(cfd);
+      assert(cfd->queued_for_flush());
+      cfd->set_queued_for_flush(false);
+    }
+  }
  // TODO: need to unset flush reason?
  return flush_req;
 }
@@ -2256,19 +2304,36 @@ ColumnFamilyData* DBImpl::PickCompactionFromQueue(

 void DBImpl::SchedulePendingFlush(const FlushRequest& flush_req,
                                  FlushReason flush_reason) {
+  mutex_.AssertHeld();
  if (flush_req.empty()) {
    return;
  }
-  for (auto& iter : flush_req) {
-    ColumnFamilyData* cfd = iter.first;
-    cfd->Ref();
-    cfd->SetFlushReason(flush_reason);
+  if (!immutable_db_options_.atomic_flush) {
+    // For the non-atomic flush case, we never schedule multiple column
+    // families in the same flush request.
+    assert(flush_req.size() == 1);
+    ColumnFamilyData* cfd = flush_req[0].first;
+    assert(cfd);
+    if (!cfd->queued_for_flush() && cfd->imm()->IsFlushPending()) {
+      cfd->Ref();
+      cfd->set_queued_for_flush(true);
+      cfd->SetFlushReason(flush_reason);
+      ++unscheduled_flushes_;
+      flush_queue_.push_back(flush_req);
+    }
+  } else {
+    for (auto& iter : flush_req) {
+      ColumnFamilyData* cfd = iter.first;
+      cfd->Ref();
+      cfd->SetFlushReason(flush_reason);
+    }
+    ++unscheduled_flushes_;
+    flush_queue_.push_back(flush_req);
  }
-  ++unscheduled_flushes_;
-  flush_queue_.push_back(flush_req);
 }

 void DBImpl::SchedulePendingCompaction(ColumnFamilyData* cfd) {
+  mutex_.AssertHeld();
  if (!cfd->queued_for_compaction() && cfd->NeedsCompaction()) {
    AddToCompactionQueue(cfd);
    ++unscheduled_compactions_;
@@ -2563,7 +2628,8 @@ void DBImpl::BackgroundCallCompaction(PrepickedCompaction* prepicked_compaction,
    // failure). Thus, we force full scan in FindObsoleteFiles()
    FindObsoleteFiles(&job_context, !s.ok() && !s.IsShutdownInProgress() &&
                                        !s.IsManualCompactionPaused() &&
-                                        !s.IsColumnFamilyDropped());
+                                        !s.IsColumnFamilyDropped() &&
+                                        !s.IsBusy());
    TEST_SYNC_POINT("DBImpl::BackgroundCallCompaction:FoundObsoleteFiles");

    // delete unnecessary files if any, this is done outside the mutex
@@ -2969,7 +3035,7 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
        c->mutable_cf_options()->report_bg_io_stats, dbname_,
        &compaction_job_stats, thread_pri, io_tracer_,
        is_manual ? &manual_compaction_paused_ : nullptr, db_id_,
-        db_session_id_);
+        db_session_id_, c->column_family_data()->GetFullHistoryTsLow());
    compaction_job.Prepare();

    NotifyOnCompactionBegin(c->column_family_data(), c.get(), status,
@@ -3034,10 +3100,9 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
      auto err_reason = versions_->io_status().ok()
                            ? BackgroundErrorReason::kCompaction
                            : BackgroundErrorReason::kManifestWrite;
-      error_handler_.SetBGError(io_s, err_reason).PermitUncheckedError();
+      error_handler_.SetBGError(io_s, err_reason);
    } else {
-      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
    }
    if (c != nullptr && !is_manual && !error_handler_.IsBGWorkStopped()) {
      // Put this cfd back in the compaction queue so we can retry after some
@@ -22,12 +22,13 @@ uint64_t DBImpl::TEST_GetLevel0TotalSize() {
  return default_cf_handle_->cfd()->current()->storage_info()->NumLevelBytes(0);
 }

-void DBImpl::TEST_SwitchWAL() {
+Status DBImpl::TEST_SwitchWAL() {
  WriteContext write_context;
  InstrumentedMutexLock l(&mutex_);
  void* writer = TEST_BeginWrite();
-  SwitchWAL(&write_context);
+  auto s = SwitchWAL(&write_context);
  TEST_EndWrite(writer);
+  return s;
 }

 bool DBImpl::TEST_WALBufferIsEmpty(bool lock) {
@@ -170,7 +171,7 @@ Status DBImpl::TEST_WaitForCompact(bool wait_unscheduled) {
  while ((bg_bottom_compaction_scheduled_ || bg_compaction_scheduled_ ||
          bg_flush_scheduled_ ||
          (wait_unscheduled && unscheduled_compactions_)) &&
-         (error_handler_.GetBGError() == Status::OK())) {
+         (error_handler_.GetBGError().ok())) {
    bg_cv_.Wait();
  }
  return error_handler_.GetBGError();
@@ -6,16 +6,17 @@
 // Copyright (c) 2011 The LevelDB Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
-#include "db/db_impl/db_impl.h"
-
 #include <cinttypes>
 #include <set>
 #include <unordered_set>
+
+#include "db/db_impl/db_impl.h"
 #include "db/event_helpers.h"
 #include "db/memtable_list.h"
 #include "file/file_util.h"
 #include "file/filename.h"
 #include "file/sst_file_manager_impl.h"
+#include "port/port.h"
 #include "util/autovector.h"

 namespace ROCKSDB_NAMESPACE {
@@ -190,7 +191,8 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
      // set of all files in the directory. We'll exclude files that are still
      // alive in the subsequent processings.
      std::vector<std::string> files;
-      env_->GetChildren(path, &files).PermitUncheckedError();  // Ignore errors
+      Status s = env_->GetChildren(path, &files);
+      s.PermitUncheckedError();  // TODO: What should we do on error?
      for (const std::string& file : files) {
        uint64_t number;
        FileType type;
@@ -206,7 +208,8 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
          continue;
        }

-        // TODO(icanadi) clean up this mess to avoid having one-off "/" prefixes
+        // TODO(icanadi) clean up this mess to avoid having one-off "/"
+        // prefixes
        job_context->full_scan_candidate_files.emplace_back("/" + file, path);
      }
    }
@@ -214,9 +217,8 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
    // Add log files in wal_dir
    if (immutable_db_options_.wal_dir != dbname_) {
      std::vector<std::string> log_files;
-      env_->GetChildren(immutable_db_options_.wal_dir,
-                        &log_files)
-          .PermitUncheckedError();  // Ignore errors
+      Status s = env_->GetChildren(immutable_db_options_.wal_dir, &log_files);
+      s.PermitUncheckedError();  // TODO: What should we do on error?
      for (const std::string& log_file : log_files) {
        job_context->full_scan_candidate_files.emplace_back(
            log_file, immutable_db_options_.wal_dir);
@@ -226,9 +228,9 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
    if (!immutable_db_options_.db_log_dir.empty() &&
        immutable_db_options_.db_log_dir != dbname_) {
      std::vector<std::string> info_log_files;
-      // Ignore errors
-      env_->GetChildren(immutable_db_options_.db_log_dir, &info_log_files)
-          .PermitUncheckedError();
+      Status s =
+          env_->GetChildren(immutable_db_options_.db_log_dir, &info_log_files);
+      s.PermitUncheckedError();  // TODO: What should we do on error?
      for (std::string& log_file : info_log_files) {
        job_context->full_scan_candidate_files.emplace_back(
            log_file, immutable_db_options_.db_log_dir);
@@ -658,13 +660,15 @@ uint64_t FindMinPrepLogReferencedByMemTable(

  // we must look through the memtables for two phase transactions
  // that have been committed but not yet flushed
+  std::unordered_set<MemTable*> memtables_to_flush_set(
+      memtables_to_flush.begin(), memtables_to_flush.end());
  for (auto loop_cfd : *vset->GetColumnFamilySet()) {
    if (loop_cfd->IsDropped() || loop_cfd == cfd_to_flush) {
      continue;
    }

    auto log = loop_cfd->imm()->PrecomputeMinLogContainingPrepSection(
-        memtables_to_flush);
+        &memtables_to_flush_set);

    if (log > 0 && (min_log == 0 || log < min_log)) {
      min_log = log;
@@ -680,6 +684,37 @@ uint64_t FindMinPrepLogReferencedByMemTable(
  return min_log;
 }

+uint64_t FindMinPrepLogReferencedByMemTable(
+    VersionSet* vset, const autovector<ColumnFamilyData*>& cfds_to_flush,
+    const autovector<const autovector<MemTable*>*>& memtables_to_flush) {
+  uint64_t min_log = 0;
+
+  std::unordered_set<ColumnFamilyData*> cfds_to_flush_set(cfds_to_flush.begin(),
+                                                          cfds_to_flush.end());
+  std::unordered_set<MemTable*> memtables_to_flush_set;
+  for (const autovector<MemTable*>* memtables : memtables_to_flush) {
+    memtables_to_flush_set.insert(memtables->begin(), memtables->end());
+  }
+  for (auto loop_cfd : *vset->GetColumnFamilySet()) {
+    if (loop_cfd->IsDropped() || cfds_to_flush_set.count(loop_cfd)) {
+      continue;
+    }
+
+    auto log = loop_cfd->imm()->PrecomputeMinLogContainingPrepSection(
+        &memtables_to_flush_set);
+    if (log > 0 && (min_log == 0 || log < min_log)) {
+      min_log = log;
+    }
+
+    log = loop_cfd->mem()->GetMinLogContainingPrepSection();
+    if (log > 0 && (min_log == 0 || log < min_log)) {
+      min_log = log;
+    }
+  }
+
+  return min_log;
+}
+
 uint64_t PrecomputeMinLogNumberToKeepNon2PC(
    VersionSet* vset, const ColumnFamilyData& cfd_to_flush,
    const autovector<VersionEdit*>& edit_list) {
@@ -710,6 +745,42 @@ uint64_t PrecomputeMinLogNumberToKeepNon2PC(
  return min_log_number_to_keep;
 }

+uint64_t PrecomputeMinLogNumberToKeepNon2PC(
+    VersionSet* vset, const autovector<ColumnFamilyData*>& cfds_to_flush,
+    const autovector<autovector<VersionEdit*>>& edit_lists) {
+  assert(vset != nullptr);
+  assert(!cfds_to_flush.empty());
+  assert(cfds_to_flush.size() == edit_lists.size());
+
+  uint64_t min_log_number_to_keep = port::kMaxUint64;
+  for (const auto& edit_list : edit_lists) {
+    uint64_t log = 0;
+    for (const auto& e : edit_list) {
+      if (e->HasLogNumber()) {
+        log = std::max(log, e->GetLogNumber());
+      }
+    }
+    if (log != 0) {
+      min_log_number_to_keep = std::min(min_log_number_to_keep, log);
+    }
+  }
+  if (min_log_number_to_keep == port::kMaxUint64) {
+    min_log_number_to_keep = cfds_to_flush[0]->GetLogNumber();
+    for (size_t i = 1; i < cfds_to_flush.size(); i++) {
+      min_log_number_to_keep =
+          std::min(min_log_number_to_keep, cfds_to_flush[i]->GetLogNumber());
+    }
+  }
+
+  std::unordered_set<const ColumnFamilyData*> flushed_cfds(
+      cfds_to_flush.begin(), cfds_to_flush.end());
+  min_log_number_to_keep =
+      std::min(min_log_number_to_keep,
+               vset->PreComputeMinLogNumberWithUnflushedData(flushed_cfds));
+
+  return min_log_number_to_keep;
+}
+
 uint64_t PrecomputeMinLogNumberToKeep2PC(
    VersionSet* vset, const ColumnFamilyData& cfd_to_flush,
    const autovector<VersionEdit*>& edit_list,
@@ -751,6 +822,38 @@ uint64_t PrecomputeMinLogNumberToKeep2PC(
  return min_log_number_to_keep;
 }

+uint64_t PrecomputeMinLogNumberToKeep2PC(
+    VersionSet* vset, const autovector<ColumnFamilyData*>& cfds_to_flush,
+    const autovector<autovector<VersionEdit*>>& edit_lists,
+    const autovector<const autovector<MemTable*>*>& memtables_to_flush,
+    LogsWithPrepTracker* prep_tracker) {
+  assert(vset != nullptr);
+  assert(prep_tracker != nullptr);
+  assert(cfds_to_flush.size() == edit_lists.size());
+  assert(cfds_to_flush.size() == memtables_to_flush.size());
+
+  uint64_t min_log_number_to_keep =
+      PrecomputeMinLogNumberToKeepNon2PC(vset, cfds_to_flush, edit_lists);
+
+  uint64_t min_log_in_prep_heap =
+      prep_tracker->FindMinLogContainingOutstandingPrep();
+
+  if (min_log_in_prep_heap != 0 &&
+      min_log_in_prep_heap < min_log_number_to_keep) {
+    min_log_number_to_keep = min_log_in_prep_heap;
+  }
+
+  uint64_t min_log_refed_by_mem = FindMinPrepLogReferencedByMemTable(
+      vset, cfds_to_flush, memtables_to_flush);
+
+  if (min_log_refed_by_mem != 0 &&
+      min_log_refed_by_mem < min_log_number_to_keep) {
+    min_log_number_to_keep = min_log_refed_by_mem;
+  }
+
+  return min_log_number_to_keep;
+}
+
 Status DBImpl::SetDBId() {
  Status s;
  // Happens when immutable_db_options_.write_dbid_to_manifest is set to true
@@ -147,7 +147,8 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
    // DeleteScheduler::CleanupDirectory on the same dir later, it will be
    // safe
    std::vector<std::string> filenames;
-    result.env->GetChildren(result.wal_dir, &filenames).PermitUncheckedError();
+    Status s = result.env->GetChildren(result.wal_dir, &filenames);
+    s.PermitUncheckedError();  //**TODO: What to do on error?
    for (std::string& filename : filenames) {
      if (filename.find(".log.trash", filename.length() -
                                          std::string(".log.trash").length()) !=
@@ -163,7 +164,8 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
  // was not used)
  auto sfm = static_cast<SstFileManagerImpl*>(result.sst_file_manager.get());
  for (size_t i = 0; i < result.db_paths.size(); i++) {
-    DeleteScheduler::CleanupDirectory(result.env, sfm, result.db_paths[i].path);
+    DeleteScheduler::CleanupDirectory(result.env, sfm, result.db_paths[i].path)
+        .PermitUncheckedError();
  }

  // Create a default SstFileManager for purposes of tracking compaction size
@@ -1368,6 +1370,9 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
                      cfd->GetName().c_str(), meta.fd.GetNumber(),
                      meta.fd.GetFileSize(), s.ToString().c_str());
      mutex_.Lock();
+
+      io_s.PermitUncheckedError();  // TODO(AR) is this correct, or should we
+                                    // return io_s if not ok()?
    }
  }
  ReleaseFileNumberFromPendingOutputs(pending_outputs_inserted_elem);
@@ -1563,6 +1568,7 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
      InstrumentedMutexLock wl(&impl->log_write_mutex_);
      impl->logfile_number_ = new_log_number;
      assert(new_log != nullptr);
+      assert(impl->logs_.empty());
      impl->logs_.emplace_back(new_log_number, new_log);
    }

@@ -1735,9 +1741,8 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
    paths.erase(std::unique(paths.begin(), paths.end()), paths.end());
    for (auto& path : paths) {
      std::vector<std::string> existing_files;
-      // TODO: Check for errors here?
      impl->immutable_db_options_.env->GetChildren(path, &existing_files)
-          .PermitUncheckedError();
+          .PermitUncheckedError();  //**TODO: What do to on error?
      for (auto& file_name : existing_files) {
        uint64_t file_number;
        FileType file_type;
@@ -83,7 +83,7 @@ Iterator* DBImplReadOnly::NewIterator(const ReadOptions& read_options,
  ReadCallback* read_callback = nullptr;  // No read callback provided.
  auto db_iter = NewArenaWrappedDbIterator(
      env_, read_options, *cfd->ioptions(), super_version->mutable_cf_options,
-      read_seq,
+      super_version->current, read_seq,
      super_version->mutable_cf_options.max_sequential_skip_in_iterations,
      super_version->version_number, read_callback);
  auto internal_iter = NewInternalIterator(
@@ -115,7 +115,8 @@ Status DBImplReadOnly::NewIterators(
    auto* cfd = static_cast_with_check<ColumnFamilyHandleImpl>(cfh)->cfd();
    auto* sv = cfd->GetSuperVersion()->Ref();
    auto* db_iter = NewArenaWrappedDbIterator(
-        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options, read_seq,
+        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
+        sv->current, read_seq,
        sv->mutable_cf_options.max_sequential_skip_in_iterations,
        sv->version_number, read_callback);
    auto* internal_iter = NewInternalIterator(
@@ -421,7 +421,7 @@ ArenaWrappedDBIter* DBImplSecondary::NewIteratorImpl(
  SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
  auto db_iter = NewArenaWrappedDbIterator(
      env_, read_options, *cfd->ioptions(), super_version->mutable_cf_options,
-      snapshot,
+      super_version->current, snapshot,
      super_version->mutable_cf_options.max_sequential_skip_in_iterations,
      super_version->version_number, read_callback);
  auto internal_iter = NewInternalIterator(
@@ -163,7 +163,6 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
  StopWatch write_sw(env_, immutable_db_options_.statistics.get(), DB_WRITE);

  write_thread_.JoinBatchGroup(&w);
-  Status status;
  if (w.state == WriteThread::STATE_PARALLEL_MEMTABLE_WRITER) {
    // we are a non-leader in a parallel group

@@ -193,8 +192,6 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
    }
    assert(w.state == WriteThread::STATE_COMPLETED);
    // STATE_COMPLETED conditional below handles exit
-
-    status = w.FinalStatus();
  }
  if (w.state == WriteThread::STATE_COMPLETED) {
    if (log_used != nullptr) {
@@ -204,13 +201,11 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
      *seq_used = w.sequence;
    }
    // write is complete and leader has updated sequence
-    // Should we handle it?
-    status.PermitUncheckedError();
    return w.FinalStatus();
  }
  // else we are the leader of the write batch group
  assert(w.state == WriteThread::STATE_GROUP_LEADER);
-
+  Status status;
  // Once reaches this point, the current writer "w" will try to do its write
  // job.  It may also pick up some of the remaining writers in the "writers_"
  // when it finds suitable, and finish them in the same write batch.
@@ -531,6 +526,8 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
    PERF_TIMER_STOP(write_pre_and_post_process_time);

    IOStatus io_s;
+    io_s.PermitUncheckedError();  // Allow io_s to be uninitialized
+
    if (w.status.ok() && !write_options.disableWAL) {
      PERF_TIMER_GUARD(write_wal_time);
      stats->AddDBStats(InternalStats::kIntStatsWriteDoneBySelf, 1);
@@ -566,7 +563,12 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
    write_thread_.ExitAsBatchGroupLeader(wal_write_group, w.status);
  }

+  // NOTE: the memtable_write_group is declared before the following
+  // `if` statement because its lifetime needs to be longer
+  // that the inner context  of the `if` as a reference to it
+  // may be used further below within the outer _write_thread
  WriteThread::WriteGroup memtable_write_group;
+
  if (w.state == WriteThread::STATE_MEMTABLE_WRITER_LEADER) {
    PERF_TIMER_GUARD(write_memtable_time);
    assert(w.ShouldWriteToMemtable());
@@ -583,6 +585,10 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
      versions_->SetLastSequence(memtable_write_group.last_sequence);
      write_thread_.ExitAsMemTableWriter(&w, memtable_write_group);
    }
+  } else {
+    // NOTE: the memtable_write_group is never really used,
+    // so we need to set its status to pass ASSERT_STATUS_CHECKED
+    memtable_write_group.status.PermitUncheckedError();
  }

  if (w.state == WriteThread::STATE_PARALLEL_MEMTABLE_WRITER) {
@@ -666,7 +672,6 @@ Status DBImpl::WriteImplWALOnly(
    const uint64_t log_ref, uint64_t* seq_used, const size_t sub_batch_cnt,
    PreReleaseCallback* pre_release_callback, const AssignOrder assign_order,
    const PublishLastSeq publish_last_seq, const bool disable_memtable) {
-  Status status;
  PERF_TIMER_GUARD(write_pre_and_post_process_time);
  WriteThread::Writer w(write_options, my_batch, callback, log_ref,
                        disable_memtable, sub_batch_cnt, pre_release_callback);
@@ -688,6 +693,8 @@ Status DBImpl::WriteImplWALOnly(
  assert(w.state == WriteThread::STATE_GROUP_LEADER);

  if (publish_last_seq == kDoPublishLastSeq) {
+    Status status;
+
    // Currently we only use kDoPublishLastSeq in unordered_write
    assert(immutable_db_options_.unordered_write);
    WriteContext write_context;
@@ -764,7 +771,9 @@ Status DBImpl::WriteImplWALOnly(
    }
    seq_inc = total_batch_cnt;
  }
+  Status status;
  IOStatus io_s;
+  io_s.PermitUncheckedError();  // Allow io_s to be uninitialized
  if (!write_options.disableWAL) {
    io_s = ConcurrentWriteToWAL(write_group, log_used, &last_sequence, seq_inc);
    status = io_s;
@@ -850,8 +859,7 @@ void DBImpl::WriteStatusCheckOnLocked(const Status& status) {
  if (immutable_db_options_.paranoid_checks && !status.ok() &&
      !status.IsBusy() && !status.IsIncomplete()) {
    // Maybe change the return status to void?
-    error_handler_.SetBGError(status, BackgroundErrorReason::kWriteCallback)
-        .PermitUncheckedError();
+    error_handler_.SetBGError(status, BackgroundErrorReason::kWriteCallback);
  }
 }

@@ -863,8 +871,7 @@ void DBImpl::WriteStatusCheck(const Status& status) {
      !status.IsBusy() && !status.IsIncomplete()) {
    mutex_.Lock();
    // Maybe change the return status to void?
-    error_handler_.SetBGError(status, BackgroundErrorReason::kWriteCallback)
-        .PermitUncheckedError();
+    error_handler_.SetBGError(status, BackgroundErrorReason::kWriteCallback);
    mutex_.Unlock();
  }
 }
@@ -877,8 +884,7 @@ void DBImpl::IOStatusCheck(const IOStatus& io_status) {
      io_status.IsIOFenced()) {
    mutex_.Lock();
    // Maybe change the return status to void?
-    error_handler_.SetBGError(io_status, BackgroundErrorReason::kWriteCallback)
-        .PermitUncheckedError();
+    error_handler_.SetBGError(io_status, BackgroundErrorReason::kWriteCallback);
    mutex_.Unlock();
  }
 }
@@ -1335,10 +1341,17 @@ Status DBImpl::SwitchWAL(WriteContext* write_context) {
    }
    for (auto cfd : cfds) {
      cfd->imm()->FlushRequested();
+      if (!immutable_db_options_.atomic_flush) {
+        FlushRequest flush_req;
+        GenerateFlushRequest({cfd}, &flush_req);
+        SchedulePendingFlush(flush_req, FlushReason::kWriteBufferManager);
+      }
+    }
+    if (immutable_db_options_.atomic_flush) {
+      FlushRequest flush_req;
+      GenerateFlushRequest(cfds, &flush_req);
+      SchedulePendingFlush(flush_req, FlushReason::kWriteBufferManager);
    }
-    FlushRequest flush_req;
-    GenerateFlushRequest(cfds, &flush_req);
-    SchedulePendingFlush(flush_req, FlushReason::kWriteBufferManager);
    MaybeScheduleFlushOrCompaction();
  }
  return status;
@@ -1414,10 +1427,17 @@ Status DBImpl::HandleWriteBufferFull(WriteContext* write_context) {
    }
    for (const auto cfd : cfds) {
      cfd->imm()->FlushRequested();
+      if (!immutable_db_options_.atomic_flush) {
+        FlushRequest flush_req;
+        GenerateFlushRequest({cfd}, &flush_req);
+        SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
+      }
+    }
+    if (immutable_db_options_.atomic_flush) {
+      FlushRequest flush_req;
+      GenerateFlushRequest(cfds, &flush_req);
+      SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
    }
-    FlushRequest flush_req;
-    GenerateFlushRequest(cfds, &flush_req);
-    SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
    MaybeScheduleFlushOrCompaction();
  }
  return status;
@@ -1641,10 +1661,16 @@ Status DBImpl::ScheduleFlushes(WriteContext* context) {
  if (status.ok()) {
    if (immutable_db_options_.atomic_flush) {
      AssignAtomicFlushSeq(cfds);
+      FlushRequest flush_req;
+      GenerateFlushRequest(cfds, &flush_req);
+      SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
+    } else {
+      for (auto* cfd : cfds) {
+        FlushRequest flush_req;
+        GenerateFlushRequest({cfd}, &flush_req);
+        SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
+      }
    }
-    FlushRequest flush_req;
-    GenerateFlushRequest(cfds, &flush_req);
-    SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
    MaybeScheduleFlushOrCompaction();
  }
  return status;
@@ -1790,32 +1816,74 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
    }
    // We may have lost data from the WritableFileBuffer in-memory buffer for
    // the current log, so treat it as a fatal error and set bg_error
-    // Should handle return error?
    if (!io_s.ok()) {
-      // Should handle return error?
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kMemTable)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kMemTable);
    } else {
-      // Should handle return error?
-      error_handler_.SetBGError(s, BackgroundErrorReason::kMemTable)
-          .PermitUncheckedError();
+      error_handler_.SetBGError(s, BackgroundErrorReason::kMemTable);
    }
    // Read back bg_error in order to get the right severity
    s = error_handler_.GetBGError();
    return s;
  }

-  for (auto loop_cfd : *versions_->GetColumnFamilySet()) {
-    // all this is just optimization to delete logs that
-    // are no longer needed -- if CF is empty, that means it
-    // doesn't need that particular log to stay alive, so we just
-    // advance the log number. no need to persist this in the manifest
-    if (loop_cfd->mem()->GetFirstSequenceNumber() == 0 &&
-        loop_cfd->imm()->NumNotFlushed() == 0) {
-      if (creating_new_log) {
-        loop_cfd->SetLogNumber(logfile_number_);
+  bool empty_cf_updated = false;
+  if (immutable_db_options_.track_and_verify_wals_in_manifest &&
+      !immutable_db_options_.allow_2pc && creating_new_log) {
+    // In non-2pc mode, WALs become obsolete if they do not contain unflushed
+    // data. Updating the empty CF's log number might cause some WALs to become
+    // obsolete. So we should track the WAL obsoletion event before actually
+    // updating the empty CF's log number.
+    uint64_t min_wal_number_to_keep =
+        versions_->PreComputeMinLogNumberWithUnflushedData(logfile_number_);
+    if (min_wal_number_to_keep >
+        versions_->GetWalSet().GetMinWalNumberToKeep()) {
+      // Get a snapshot of the empty column families.
+      // LogAndApply may release and reacquire db
+      // mutex, during that period, column family may become empty (e.g. its
+      // flush succeeds), then it affects the computed min_log_number_to_keep,
+      // so we take a snapshot for consistency of column family data
+      // status. If a column family becomes non-empty afterwards, its active log
+      // should still be the created new log, so the min_log_number_to_keep is
+      // not affected.
+      autovector<ColumnFamilyData*> empty_cfs;
+      for (auto cf : *versions_->GetColumnFamilySet()) {
+        if (cf->IsEmpty()) {
+          empty_cfs.push_back(cf);
+        }
+      }
+
+      VersionEdit wal_deletion;
+      wal_deletion.DeleteWalsBefore(min_wal_number_to_keep);
+      s = versions_->LogAndApplyToDefaultColumnFamily(&wal_deletion, &mutex_);
+      if (!s.ok() && versions_->io_status().IsIOError()) {
+        s = error_handler_.SetBGError(versions_->io_status(),
+                                      BackgroundErrorReason::kManifestWrite);
+      }
+      if (!s.ok()) {
+        return s;
+      }
+
+      for (auto cf : empty_cfs) {
+        if (cf->IsEmpty()) {
+          cf->SetLogNumber(logfile_number_);
+          cf->mem()->SetCreationSeq(versions_->LastSequence());
+        }  // cf may become non-empty.
+      }
+      empty_cf_updated = true;
+    }
+  }
+  if (!empty_cf_updated) {
+    for (auto cf : *versions_->GetColumnFamilySet()) {
+      // all this is just optimization to delete logs that
+      // are no longer needed -- if CF is empty, that means it
+      // doesn't need that particular log to stay alive, so we just
+      // advance the log number. no need to persist this in the manifest
+      if (cf->IsEmpty()) {
+        if (creating_new_log) {
+          cf->SetLogNumber(logfile_number_);
+        }
+        cf->mem()->SetCreationSeq(versions_->LastSequence());
      }
-      loop_cfd->mem()->SetCreationSeq(versions_->LastSequence());
    }
  }

@@ -748,10 +748,12 @@ TEST_F(DBSecondaryTest, SwitchWALMultiColumnFamilies) {
    }
  };
  for (int k = 0; k != 8; ++k) {
-    ASSERT_OK(
-        Put(0 /*cf*/, "key" + std::to_string(k), "value" + std::to_string(k)));
-    ASSERT_OK(
-        Put(1 /*cf*/, "key" + std::to_string(k), "value" + std::to_string(k)));
+    for (int j = 0; j < 2; ++j) {
+      ASSERT_OK(Put(0 /*cf*/, "key" + std::to_string(k),
+                    "value" + std::to_string(k)));
+      ASSERT_OK(Put(1 /*cf*/, "key" + std::to_string(k),
+                    "value" + std::to_string(k)));
+    }
    TEST_SYNC_POINT(
        "DBSecondaryTest::SwitchWALMultipleColumnFamilies:BeforeCatchUp");
    ASSERT_OK(db_secondary_->TryCatchUpWithPrimary());
@@ -36,16 +36,18 @@ namespace ROCKSDB_NAMESPACE {
 DBIter::DBIter(Env* _env, const ReadOptions& read_options,
               const ImmutableCFOptions& cf_options,
               const MutableCFOptions& mutable_cf_options,
-               const Comparator* cmp, InternalIterator* iter, SequenceNumber s,
-               bool arena_mode, uint64_t max_sequential_skip_in_iterations,
+               const Comparator* cmp, InternalIterator* iter,
+               const Version* version, SequenceNumber s, bool arena_mode,
+               uint64_t max_sequential_skip_in_iterations,
               ReadCallback* read_callback, DBImpl* db_impl,
-               ColumnFamilyData* cfd, bool allow_blob)
+               ColumnFamilyData* cfd, bool expose_blob_index)
    : prefix_extractor_(mutable_cf_options.prefix_extractor.get()),
      env_(_env),
      logger_(cf_options.info_log),
      user_comparator_(cmp),
      merge_operator_(cf_options.merge_operator),
      iter_(iter),
+      version_(version),
      read_callback_(read_callback),
      sequence_(s),
      statistics_(cf_options.statistics),
@@ -65,7 +67,9 @@ DBIter::DBIter(Env* _env, const ReadOptions& read_options,
      expect_total_order_inner_iter_(prefix_extractor_ == nullptr ||
                                     read_options.total_order_seek ||
                                     read_options.auto_prefix_mode),
-      allow_blob_(allow_blob),
+      read_tier_(read_options.read_tier),
+      verify_checksums_(read_options.verify_checksums),
+      expose_blob_index_(expose_blob_index),
      is_blob_(false),
      arena_mode_(arena_mode),
      range_del_agg_(&cf_options.internal_comparator, s),
@@ -165,6 +169,40 @@ void DBIter::Next() {
  }
 }

+bool DBIter::SetBlobValueIfNeeded(const Slice& user_key,
+                                  const Slice& blob_index) {
+  assert(!is_blob_);
+
+  if (expose_blob_index_) {  // Stacked BlobDB implementation
+    is_blob_ = true;
+    return true;
+  }
+
+  if (!version_) {
+    status_ = Status::Corruption("Encountered unexpected blob index.");
+    valid_ = false;
+    return false;
+  }
+
+  // TODO: consider moving ReadOptions from ArenaWrappedDBIter to DBIter to
+  // avoid having to copy options back and forth.
+  ReadOptions read_options;
+  read_options.read_tier = read_tier_;
+  read_options.verify_checksums = verify_checksums_;
+
+  const Status s =
+      version_->GetBlob(read_options, user_key, blob_index, &blob_value_);
+
+  if (!s.ok()) {
+    status_ = s;
+    valid_ = false;
+    return false;
+  }
+
+  is_blob_ = true;
+  return true;
+}
+
 // PRE: saved_key_ has the current user key if skipping_saved_key
 // POST: saved_key_ should have the next user key if valid_,
 //       if the current entry is a result of merge
@@ -221,25 +259,28 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
      is_key_seqnum_zero_ = false;
      return false;
    }
+    Slice user_key_without_ts =
+        StripTimestampFromUserKey(ikey_.user_key, timestamp_size_);

    is_key_seqnum_zero_ = (ikey_.sequence == 0);

    assert(iterate_upper_bound_ == nullptr ||
           iter_.UpperBoundCheckResult() != IterBoundCheck::kInbound ||
           user_comparator_.CompareWithoutTimestamp(
-               ikey_.user_key, /*a_has_ts=*/true, *iterate_upper_bound_,
+               user_key_without_ts, /*a_has_ts=*/false, *iterate_upper_bound_,
               /*b_has_ts=*/false) < 0);
    if (iterate_upper_bound_ != nullptr &&
        iter_.UpperBoundCheckResult() != IterBoundCheck::kInbound &&
        user_comparator_.CompareWithoutTimestamp(
-            ikey_.user_key, /*a_has_ts=*/true, *iterate_upper_bound_,
+            user_key_without_ts, /*a_has_ts=*/false, *iterate_upper_bound_,
            /*b_has_ts=*/false) >= 0) {
      break;
    }

    assert(prefix == nullptr || prefix_extractor_ != nullptr);
    if (prefix != nullptr &&
-        prefix_extractor_->Transform(ikey_.user_key).compare(*prefix) != 0) {
+        prefix_extractor_->Transform(user_key_without_ts).compare(*prefix) !=
+            0) {
      assert(prefix_same_as_start_);
      break;
    }
@@ -316,8 +357,14 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
          case kTypeBlobIndex:
            if (start_seqnum_ > 0) {
              if (ikey_.sequence >= start_seqnum_) {
-                assert(ikey_.type != kTypeBlobIndex);
                saved_key_.SetInternalKey(ikey_);
+
+                if (ikey_.type == kTypeBlobIndex) {
+                  if (!SetBlobValueIfNeeded(ikey_.user_key, iter_.value())) {
+                    return false;
+                  }
+                }
+
                valid_ = true;
                return true;
              } else {
@@ -331,6 +378,13 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
              }
            } else if (timestamp_lb_) {
              saved_key_.SetInternalKey(ikey_);
+
+              if (ikey_.type == kTypeBlobIndex) {
+                if (!SetBlobValueIfNeeded(ikey_.user_key, iter_.value())) {
+                  return false;
+                }
+              }
+
              valid_ = true;
              return true;
            } else {
@@ -345,20 +399,13 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
                num_skipped = 0;
                reseek_done = false;
                PERF_COUNTER_ADD(internal_delete_skipped_count, 1);
-              } else if (ikey_.type == kTypeBlobIndex) {
-                if (!allow_blob_) {
-                  ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
-                  status_ = Status::NotSupported(
-                      "Encounter unexpected blob index. Please open DB with "
-                      "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
-                  valid_ = false;
-                  return false;
+              } else {
+                if (ikey_.type == kTypeBlobIndex) {
+                  if (!SetBlobValueIfNeeded(ikey_.user_key, iter_.value())) {
+                    return false;
+                  }
                }

-                is_blob_ = true;
-                valid_ = true;
-                return true;
-              } else {
                valid_ = true;
                return true;
              }
@@ -498,7 +545,6 @@ bool DBIter::MergeValuesNewToOld() {
  TEST_SYNC_POINT("DBIter::MergeValuesNewToOld:PushedFirstOperand");

  ParsedInternalKey ikey;
-  Status s;
  for (iter_.Next(); iter_.Valid(); iter_.Next()) {
    TEST_SYNC_POINT("DBIter::MergeValuesNewToOld:SteppedToNextOperand");
    if (!ParseKey(&ikey)) {
@@ -526,7 +572,7 @@ bool DBIter::MergeValuesNewToOld() {
      // hit a put, merge the put value with operands and store the
      // final result in saved_value_. We are done!
      const Slice val = iter_.value();
-      s = MergeHelper::TimedFullMerge(
+      Status s = MergeHelper::TimedFullMerge(
          merge_operator_, ikey.user_key, &val, merge_context_.GetOperands(),
          &saved_value_, logger_, statistics_, env_, &pinned_value_, true);
      if (!s.ok()) {
@@ -548,15 +594,7 @@ bool DBIter::MergeValuesNewToOld() {
          iter_.value(), iter_.iter()->IsValuePinned() /* operand_pinned */);
      PERF_COUNTER_ADD(internal_merge_count, 1);
    } else if (kTypeBlobIndex == ikey.type) {
-      if (!allow_blob_) {
-        ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
-        status_ = Status::NotSupported(
-            "Encounter unexpected blob index. Please open DB with "
-            "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
-      } else {
-        status_ =
-            Status::NotSupported("Blob DB does not support merge operator.");
-      }
+      status_ = Status::NotSupported("BlobDB does not support merge operator.");
      valid_ = false;
      return false;
    } else {
@@ -577,10 +615,10 @@ bool DBIter::MergeValuesNewToOld() {
  // a deletion marker.
  // feed null as the existing value to the merge operator, such that
  // client can differentiate this scenario and do things accordingly.
-  s = MergeHelper::TimedFullMerge(merge_operator_, saved_key_.GetUserKey(),
-                                  nullptr, merge_context_.GetOperands(),
-                                  &saved_value_, logger_, statistics_, env_,
-                                  &pinned_value_, true);
+  Status s = MergeHelper::TimedFullMerge(
+      merge_operator_, saved_key_.GetUserKey(), nullptr,
+      merge_context_.GetOperands(), &saved_value_, logger_, statistics_, env_,
+      &pinned_value_, true);
  if (!s.ok()) {
    valid_ = false;
    status_ = s;
@@ -704,7 +742,9 @@ void DBIter::PrevInternal(const Slice* prefix) {

    assert(prefix == nullptr || prefix_extractor_ != nullptr);
    if (prefix != nullptr &&
-        prefix_extractor_->Transform(saved_key_.GetUserKey())
+        prefix_extractor_
+                ->Transform(StripTimestampFromUserKey(saved_key_.GetUserKey(),
+                                                      timestamp_size_))
                .compare(*prefix) != 0) {
      assert(prefix_same_as_start_);
      // Current key does not have the same prefix as start
@@ -883,15 +923,8 @@ bool DBIter::FindValueForCurrentKey() {
            merge_context_.GetOperands(), &saved_value_, logger_, statistics_,
            env_, &pinned_value_, true);
      } else if (last_not_merge_type == kTypeBlobIndex) {
-        if (!allow_blob_) {
-          ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
-          status_ = Status::NotSupported(
-              "Encounter unexpected blob index. Please open DB with "
-              "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
-        } else {
-          status_ =
-              Status::NotSupported("Blob DB does not support merge operator.");
-        }
+        status_ =
+            Status::NotSupported("BlobDB does not support merge operator.");
        valid_ = false;
        return false;
      } else {
@@ -906,15 +939,10 @@ bool DBIter::FindValueForCurrentKey() {
      // do nothing - we've already has value in pinned_value_
      break;
    case kTypeBlobIndex:
-      if (!allow_blob_) {
-        ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
-        status_ = Status::NotSupported(
-            "Encounter unexpected blob index. Please open DB with "
-            "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
-        valid_ = false;
+      if (!SetBlobValueIfNeeded(saved_key_.GetUserKey(), pinned_value_)) {
        return false;
      }
-      is_blob_ = true;
+
      break;
    default:
      valid_ = false;
@@ -987,14 +1015,6 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
    valid_ = false;
    return true;
  }
-  if (ikey.type == kTypeBlobIndex && !allow_blob_) {
-    ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
-    status_ = Status::NotSupported(
-        "Encounter unexpected blob index. Please open DB with "
-        "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
-    valid_ = false;
-    return false;
-  }
  if (!iter_.PrepareValue()) {
    valid_ = false;
    return false;
@@ -1002,7 +1022,12 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
  if (ikey.type == kTypeValue || ikey.type == kTypeBlobIndex) {
    assert(iter_.iter()->IsValuePinned());
    pinned_value_ = iter_.value();
-    is_blob_ = (ikey.type == kTypeBlobIndex);
+    if (ikey.type == kTypeBlobIndex) {
+      if (!SetBlobValueIfNeeded(ikey.user_key, pinned_value_)) {
+        return false;
+      }
+    }
+
    valid_ = true;
    return true;
  }
@@ -1058,15 +1083,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
          iter_.value(), iter_.iter()->IsValuePinned() /* operand_pinned */);
      PERF_COUNTER_ADD(internal_merge_count, 1);
    } else if (ikey.type == kTypeBlobIndex) {
-      if (!allow_blob_) {
-        ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
-        status_ = Status::NotSupported(
-            "Encounter unexpected blob index. Please open DB with "
-            "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
-      } else {
-        status_ =
-            Status::NotSupported("Blob DB does not support merge operator.");
-      }
+      status_ = Status::NotSupported("BlobDB does not support merge operator.");
      valid_ = false;
      return false;
    } else {
@@ -1399,7 +1416,8 @@ void DBIter::SeekToFirst() {
  }
  if (valid_ && prefix_same_as_start_) {
    assert(prefix_extractor_ != nullptr);
-    prefix_.SetUserKey(prefix_extractor_->Transform(saved_key_.GetUserKey()));
+    prefix_.SetUserKey(prefix_extractor_->Transform(
+        StripTimestampFromUserKey(saved_key_.GetUserKey(), timestamp_size_)));
  }
 }

@@ -1450,7 +1468,8 @@ void DBIter::SeekToLast() {
  }
  if (valid_ && prefix_same_as_start_) {
    assert(prefix_extractor_ != nullptr);
-    prefix_.SetUserKey(prefix_extractor_->Transform(saved_key_.GetUserKey()));
+    prefix_.SetUserKey(prefix_extractor_->Transform(
+        StripTimestampFromUserKey(saved_key_.GetUserKey(), timestamp_size_)));
  }
 }

@@ -1458,15 +1477,16 @@ Iterator* NewDBIterator(Env* env, const ReadOptions& read_options,
                        const ImmutableCFOptions& cf_options,
                        const MutableCFOptions& mutable_cf_options,
                        const Comparator* user_key_comparator,
-                        InternalIterator* internal_iter,
+                        InternalIterator* internal_iter, const Version* version,
                        const SequenceNumber& sequence,
                        uint64_t max_sequential_skip_in_iterations,
                        ReadCallback* read_callback, DBImpl* db_impl,
-                        ColumnFamilyData* cfd, bool allow_blob) {
-  DBIter* db_iter = new DBIter(
-      env, read_options, cf_options, mutable_cf_options, user_key_comparator,
-      internal_iter, sequence, false, max_sequential_skip_in_iterations,
-      read_callback, db_impl, cfd, allow_blob);
+                        ColumnFamilyData* cfd, bool expose_blob_index) {
+  DBIter* db_iter =
+      new DBIter(env, read_options, cf_options, mutable_cf_options,
+                 user_key_comparator, internal_iter, version, sequence, false,
+                 max_sequential_skip_in_iterations, read_callback, db_impl, cfd,
+                 expose_blob_index);
  return db_iter;
 }

@@ -22,6 +22,8 @@

 namespace ROCKSDB_NAMESPACE {

+class Version;
+
 // This file declares the factory functions of DBIter, in its original form
 // or a wrapped form with class ArenaWrappedDBIter, which is defined here.
 // Class DBIter, which is declared and implemented inside db_iter.cc, is
@@ -114,10 +116,10 @@ class DBIter final : public Iterator {
  DBIter(Env* _env, const ReadOptions& read_options,
         const ImmutableCFOptions& cf_options,
         const MutableCFOptions& mutable_cf_options, const Comparator* cmp,
-         InternalIterator* iter, SequenceNumber s, bool arena_mode,
-         uint64_t max_sequential_skip_in_iterations,
+         InternalIterator* iter, const Version* version, SequenceNumber s,
+         bool arena_mode, uint64_t max_sequential_skip_in_iterations,
         ReadCallback* read_callback, DBImpl* db_impl, ColumnFamilyData* cfd,
-         bool allow_blob);
+         bool expose_blob_index);

  // No copying allowed
  DBIter(const DBIter&) = delete;
@@ -159,7 +161,10 @@ class DBIter final : public Iterator {
  }
  Slice value() const override {
    assert(valid_);
-    if (current_entry_is_merged_) {
+
+    if (!expose_blob_index_ && is_blob_) {
+      return blob_value_;
+    } else if (current_entry_is_merged_) {
      // If pinned_value_ is set then the result of merge operator is one of
      // the merge operands and we should return it.
      return pinned_value_.data() ? pinned_value_ : saved_value_;
@@ -185,7 +190,7 @@ class DBIter final : public Iterator {
    return ExtractTimestampFromUserKey(ukey_and_ts, timestamp_size_);
  }
  bool IsBlob() const {
-    assert(valid_ && (allow_blob_ || !is_blob_));
+    assert(valid_);
    return is_blob_;
  }

@@ -287,12 +292,17 @@ class DBIter final : public Iterator {
               : user_comparator_.CompareWithoutTimestamp(a, b);
  }

+  // Retrieves the blob value for the specified user key using the given blob
+  // index when using the integrated BlobDB implementation.
+  bool SetBlobValueIfNeeded(const Slice& user_key, const Slice& blob_index);
+
  const SliceTransform* prefix_extractor_;
  Env* const env_;
  Logger* logger_;
  UserComparatorWrapper user_comparator_;
  const MergeOperator* const merge_operator_;
  IteratorWrapper iter_;
+  const Version* version_;
  ReadCallback* read_callback_;
  // Max visible sequence number. It is normally the snapshot seq unless we have
  // uncommitted data in db as in WriteUnCommitted.
@@ -306,6 +316,7 @@ class DBIter final : public Iterator {
  std::string saved_value_;
  Slice pinned_value_;
  // for prefix seek mode to support prev()
+  PinnableSlice blob_value_;
  Statistics* statistics_;
  uint64_t max_skip_;
  uint64_t max_skippable_internal_keys_;
@@ -335,7 +346,11 @@ class DBIter final : public Iterator {
  // Expect the inner iterator to maintain a total order.
  // prefix_extractor_ must be non-NULL if the value is false.
  const bool expect_total_order_inner_iter_;
-  bool allow_blob_;
+  ReadTier read_tier_;
+  bool verify_checksums_;
+  // Whether the iterator is allowed to expose blob references. Set to true when
+  // the stacked BlobDB implementation is used, false otherwise.
+  bool expose_blob_index_;
  bool is_blob_;
  bool arena_mode_;
  // List of operands for merge operator.
@@ -367,8 +382,9 @@ extern Iterator* NewDBIterator(
    const ImmutableCFOptions& cf_options,
    const MutableCFOptions& mutable_cf_options,
    const Comparator* user_key_comparator, InternalIterator* internal_iter,
-    const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iterations,
-    ReadCallback* read_callback, DBImpl* db_impl = nullptr,
-    ColumnFamilyData* cfd = nullptr, bool allow_blob = false);
+    const Version* version, const SequenceNumber& sequence,
+    uint64_t max_sequential_skip_in_iterations, ReadCallback* read_callback,
+    DBImpl* db_impl = nullptr, ColumnFamilyData* cfd = nullptr,
+    bool expose_blob_index = false);

 }  // namespace ROCKSDB_NAMESPACE
@@ -513,7 +513,7 @@ TEST_F(DBIteratorStressTest, StressTest) {
                  db_iter.reset(NewDBIterator(
                      env_, ropt, ImmutableCFOptions(options),
                      MutableCFOptions(options), BytewiseComparator(),
-                      internal_iter, sequence,
+                      internal_iter, nullptr /* version */, sequence,
                      options.max_sequential_skip_in_iterations,
                      nullptr /*read_callback*/));
                }
@@ -253,8 +253,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
    ReadOptions ro;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -286,8 +287,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
    ReadOptions ro;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -313,8 +315,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -346,8 +349,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -382,11 +386,13 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());
  }
  // Test case to check SeekToLast with iterate_upper_bound set
  // (same key put may times - SeekToLast should start with the
@@ -412,8 +418,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 7, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 7 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    SetPerfLevel(kEnableCount);
    ASSERT_TRUE(GetPerfLevel() == kEnableCount);
@@ -450,8 +457,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 4, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 4 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -476,11 +484,13 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());
  }
  // Test to check the SeekToLast() with the iterate_upper_bound set
  // (Deletion cases)
@@ -499,8 +509,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -535,8 +546,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 7, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 7 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    SetPerfLevel(kEnableCount);
    ASSERT_TRUE(GetPerfLevel() == kEnableCount);
@@ -565,8 +577,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
    ReadOptions ro;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -585,6 +598,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {

    db_iter->Prev();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());
  }

  {
@@ -608,8 +622,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
    ReadOptions ro;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 2, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 2 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "b");
@@ -640,8 +655,9 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
    ReadOptions ro;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -671,10 +687,12 @@ TEST_F(DBIteratorTest, DBIteratorEmpty) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 0, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 0 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());
  }

  {
@@ -683,10 +701,12 @@ TEST_F(DBIteratorTest, DBIteratorEmpty) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 0, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 0 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToFirst();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());
  }
 }

@@ -706,8 +726,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkipCountSkips) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 2,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      2 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToLast();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -728,6 +749,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkipCountSkips) {

  db_iter->Prev();
  ASSERT_TRUE(!db_iter->Valid());
+  ASSERT_OK(db_iter->status());
  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 3u);
 }

@@ -751,8 +773,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, i + 2, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, i + 2 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));
      db_iter->SeekToLast();
      ASSERT_TRUE(db_iter->Valid());

@@ -771,6 +794,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      db_iter->Prev();

      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    }
  }

@@ -787,8 +811,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, i + 2, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, i + 2 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));
      db_iter->SeekToLast();
      ASSERT_TRUE(db_iter->Valid());

@@ -802,6 +827,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      db_iter->Prev();

      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    }

    {
@@ -816,8 +842,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, 202, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, 202 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));
      db_iter->SeekToLast();
      ASSERT_TRUE(db_iter->Valid());

@@ -836,6 +863,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      db_iter->Prev();

      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    }
  }

@@ -849,13 +877,16 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      internal_iter->Finish();
      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, i, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, i /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));
      db_iter->SeekToLast();
      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());

      db_iter->SeekToFirst();
      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    }

    TestIterator* internal_iter = new TestIterator(BytewiseComparator());
@@ -866,8 +897,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
    internal_iter->Finish();
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 200, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 200 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -875,6 +907,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

    db_iter->Prev();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -883,6 +916,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

    db_iter->Next();
    ASSERT_TRUE(!db_iter->Valid());
+    ASSERT_OK(db_iter->status());
  }

  {
@@ -901,8 +935,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, i + 2, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, i + 2 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));
      db_iter->SeekToLast();
      ASSERT_TRUE(db_iter->Valid());

@@ -921,6 +956,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      db_iter->Prev();

      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    }
  }

@@ -936,8 +972,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, i + 2, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, i + 2 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));
      db_iter->SeekToLast();
      ASSERT_TRUE(db_iter->Valid());

@@ -960,6 +997,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {

      db_iter->Prev();
      ASSERT_TRUE(!db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    }
  }
 }
@@ -986,8 +1024,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 0;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -1015,7 +1054,7 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {

    db_iter->Prev();
    ASSERT_TRUE(!db_iter->Valid());
-    ASSERT_TRUE(db_iter->status().ok());
+    ASSERT_OK(db_iter->status());
  }

  // Test to make sure that the request will *not* fail as incomplete if
@@ -1033,8 +1072,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 2;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -1078,8 +1118,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 2;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -1117,8 +1158,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 2;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -1153,8 +1195,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 2;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -1184,8 +1227,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 2;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -1222,8 +1266,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
    ro.max_skippable_internal_keys = 2;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    ASSERT_TRUE(db_iter->Valid());
@@ -1260,8 +1305,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
      ro.max_skippable_internal_keys = i;
      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, 2 * i + 1, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, 2 * i + 1 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));

      db_iter->SeekToFirst();
      ASSERT_TRUE(db_iter->Valid());
@@ -1314,8 +1360,9 @@ TEST_F(DBIteratorTest, DBIteratorSkipInternalKeys) {
      ro.max_skippable_internal_keys = i;
      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-          internal_iter, 2 * i + 1, options.max_sequential_skip_in_iterations,
-          nullptr /*read_callback*/));
+          internal_iter, nullptr /* version */, 2 * i + 1 /* sequence */,
+          options.max_sequential_skip_in_iterations,
+          nullptr /* read_callback */));

      db_iter->SeekToFirst();
      ASSERT_TRUE(db_iter->Valid());
@@ -1353,8 +1400,9 @@ TEST_F(DBIteratorTest, DBIterator1) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 1,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      1 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToFirst();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1381,8 +1429,9 @@ TEST_F(DBIteratorTest, DBIterator2) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 0,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      0 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToFirst();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1406,8 +1455,9 @@ TEST_F(DBIteratorTest, DBIterator3) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 2,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      2 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToFirst();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1431,8 +1481,9 @@ TEST_F(DBIteratorTest, DBIterator4) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 4,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      4 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToFirst();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1465,8 +1516,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 0, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 0 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1488,8 +1540,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 1, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 1 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1511,8 +1564,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 2, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 2 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1534,8 +1588,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 3, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 3 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1557,8 +1612,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 4, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 4 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1580,8 +1636,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 5, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 5 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1603,8 +1660,9 @@ TEST_F(DBIteratorTest, DBIterator5) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 6, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 6 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1624,8 +1682,9 @@ TEST_F(DBIteratorTest, DBIterator5) {
    internal_iter->Finish();
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 10, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 10 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->Seek("b");
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "b");
@@ -1655,8 +1714,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 0, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 0 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1678,8 +1738,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 1, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 1 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1701,8 +1762,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 2, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 2 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1724,8 +1786,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 3, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 3 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(!db_iter->Valid());
  }
@@ -1743,8 +1806,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 4, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 4 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1766,8 +1830,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 5, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 5 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1789,8 +1854,9 @@ TEST_F(DBIteratorTest, DBIterator6) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 6, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 6 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1832,8 +1898,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 0, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 0 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
    ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1867,8 +1934,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 2, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 2 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -1908,8 +1976,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 4, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 4 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -1949,8 +2018,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 5, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 5 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -1995,8 +2065,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 6, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 6 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -2042,8 +2113,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 7, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 7 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -2083,8 +2155,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 9, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 9 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -2130,8 +2203,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 13, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 13 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -2178,8 +2252,9 @@ TEST_F(DBIteratorTest, DBIterator7) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
-        internal_iter, 14, options.max_sequential_skip_in_iterations,
-        nullptr /*read_callback*/));
+        internal_iter, nullptr /* version */, 14 /* sequence */,
+        options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());

@@ -2209,8 +2284,9 @@ TEST_F(DBIteratorTest, DBIterator8) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToLast();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "b");
@@ -2240,8 +2316,9 @@ TEST_F(DBIteratorTest, DBIterator9) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-        BytewiseComparator(), internal_iter, 10,
-        options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+        BytewiseComparator(), internal_iter, nullptr /* version */,
+        10 /* sequence */, options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));

    db_iter->SeekToLast();
    ASSERT_TRUE(db_iter->Valid());
@@ -2307,8 +2384,9 @@ TEST_F(DBIteratorTest, DBIterator10) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));

  db_iter->Seek("c");
  ASSERT_TRUE(db_iter->Valid());
@@ -2347,8 +2425,8 @@ TEST_F(DBIteratorTest, SeekToLastOccurrenceSeq0) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10, 0 /* force seek */,
-      nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, 0 /* force seek */, nullptr /* read_callback */));
  db_iter->SeekToFirst();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -2376,8 +2454,9 @@ TEST_F(DBIteratorTest, DBIterator11) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 1,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      1 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));
  db_iter->SeekToFirst();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -2403,7 +2482,8 @@ TEST_F(DBIteratorTest, DBIterator12) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10, 0, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, 0 /* force seek */, nullptr /* read_callback */));
  db_iter->SeekToLast();
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -2440,7 +2520,9 @@ TEST_F(DBIteratorTest, DBIterator13) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 2, 3, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      2 /* sequence */, 3 /* max_sequential_skip_in_iterations */,
+      nullptr /* read_callback */));
  db_iter->Seek("b");
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), key);
@@ -2468,7 +2550,9 @@ TEST_F(DBIteratorTest, DBIterator14) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 4, 1, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      4 /* sequence */, 1 /* max_sequential_skip_in_iterations */,
+      nullptr /* read_callback */));
  db_iter->Seek("b");
  ASSERT_TRUE(db_iter->Valid());
  ASSERT_EQ(db_iter->key().ToString(), "b");
@@ -2495,8 +2579,9 @@ TEST_F(DBIteratorTest, DBIteratorTestDifferentialSnapshots) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-        BytewiseComparator(), internal_iter, 13,
-        options.max_sequential_skip_in_iterations, nullptr));
+        BytewiseComparator(), internal_iter, nullptr /* version */,
+        13 /* sequence */, options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    // Expecting InternalKeys in [5,8] range with correct type
    int seqnums[4] = {5,8,11,13};
    std::string user_keys[4] = {"1","2","3","4"};
@@ -2530,8 +2615,9 @@ TEST_F(DBIteratorTest, DBIteratorTestDifferentialSnapshots) {

    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-        BytewiseComparator(), internal_iter, 13,
-        options.max_sequential_skip_in_iterations, nullptr));
+        BytewiseComparator(), internal_iter, nullptr /* version */,
+        13 /* sequence */, options.max_sequential_skip_in_iterations,
+        nullptr /* read_callback */));
    // Expecting InternalKeys in [5,8] range with correct type
    int seqnums[4] = {5,8,11,13};
    EntryType key_types[4] = {EntryType::kEntryDelete,EntryType::kEntryDelete,
@@ -2580,9 +2666,9 @@ class DBIterWithMergeIterTest : public testing::Test {

    db_iter_.reset(NewDBIterator(
        env_, ro_, ImmutableCFOptions(options_), MutableCFOptions(options_),
-        BytewiseComparator(), merge_iter,
+        BytewiseComparator(), merge_iter, nullptr /* version */,
        8 /* read data earlier than seqId 8 */,
-        3 /* max iterators before reseek */, nullptr /*read_callback*/));
+        3 /* max iterators before reseek */, nullptr /* read_callback */));
  }

  Env* env_;
@@ -3020,8 +3106,9 @@ TEST_F(DBIteratorTest, SeekPrefixTombstones) {
  ro.prefix_same_as_start = true;
  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));

  int skipped_keys = 0;

@@ -3056,14 +3143,15 @@ TEST_F(DBIteratorTest, SeekToFirstLowerBound) {
    Options options;
    std::unique_ptr<Iterator> db_iter(NewDBIterator(
        env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-        BytewiseComparator(), internal_iter, 10 /* sequence */,
-        options.max_sequential_skip_in_iterations,
+        BytewiseComparator(), internal_iter, nullptr /* version */,
+        10 /* sequence */, options.max_sequential_skip_in_iterations,
        nullptr /* read_callback */));

    db_iter->SeekToFirst();
    if (i == kNumKeys + 1) {
      // lower bound was beyond the last key
      ASSERT_FALSE(db_iter->Valid());
+      ASSERT_OK(db_iter->status());
    } else {
      ASSERT_TRUE(db_iter->Valid());
      int expected;
@@ -3095,8 +3183,9 @@ TEST_F(DBIteratorTest, PrevLowerBound) {
  Options options;
  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10 /* sequence */,
-      options.max_sequential_skip_in_iterations, nullptr /* read_callback */));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));

  db_iter->SeekToLast();
  for (int i = kNumKeys; i >= kLowerBound; --i) {
@@ -3123,8 +3212,9 @@ TEST_F(DBIteratorTest, SeekLessLowerBound) {
  Options options;
  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ro, ImmutableCFOptions(options), MutableCFOptions(options),
-      BytewiseComparator(), internal_iter, 10 /* sequence */,
-      options.max_sequential_skip_in_iterations, nullptr /* read_callback */));
+      BytewiseComparator(), internal_iter, nullptr /* version */,
+      10 /* sequence */, options.max_sequential_skip_in_iterations,
+      nullptr /* read_callback */));

  auto before_lower_bound_str = std::to_string(kLowerBound - 1);
  Slice before_lower_bound(lower_bound_str);
@@ -3148,8 +3238,9 @@ TEST_F(DBIteratorTest, ReverseToForwardWithDisappearingKeys) {

  std::unique_ptr<Iterator> db_iter(NewDBIterator(
      env_, ReadOptions(), ImmutableCFOptions(options),
-      MutableCFOptions(options), BytewiseComparator(), internal_iter, 10,
-      options.max_sequential_skip_in_iterations, nullptr /*read_callback*/));
+      MutableCFOptions(options), BytewiseComparator(), internal_iter,
+      nullptr /* version */, 10 /* sequence */,
+      options.max_sequential_skip_in_iterations, nullptr /* read_callback */));

  db_iter->SeekForPrev("a");
  ASSERT_TRUE(db_iter->Valid());
@@ -67,8 +67,8 @@ TEST_P(DBIteratorTest, IteratorProperty) {
  // The test needs to be changed if kPersistedTier is supported in iterator.
  Options options = CurrentOptions();
  CreateAndReopenWithCF({"pikachu"}, options);
-  Put(1, "1", "2");
-  Delete(1, "2");
+  ASSERT_OK(Put(1, "1", "2"));
+  ASSERT_OK(Delete(1, "2"));
  ReadOptions ropt;
  ropt.pin_data = false;
  {
@@ -172,10 +172,10 @@ TEST_P(DBIteratorTest, NonBlockingIteration) {
 TEST_P(DBIteratorTest, IterSeekBeforePrev) {
  ASSERT_OK(Put("a", "b"));
  ASSERT_OK(Put("c", "d"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("0", "f"));
  ASSERT_OK(Put("1", "h"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("2", "j"));
  auto iter = NewIterator(ReadOptions());
  iter->Seek(Slice("c"));
@@ -199,7 +199,7 @@ TEST_P(DBIteratorTest, IterReseekNewUpperBound) {
  ASSERT_OK(Put("aabb", rnd.RandomString(400)));
  ASSERT_OK(Put("aaef", rnd.RandomString(400)));
  ASSERT_OK(Put("b", rnd.RandomString(400)));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ReadOptions opts;
  Slice ub = Slice("aa");
  opts.iterate_upper_bound = &ub;
@@ -215,10 +215,10 @@ TEST_P(DBIteratorTest, IterReseekNewUpperBound) {
 TEST_P(DBIteratorTest, IterSeekForPrevBeforeNext) {
  ASSERT_OK(Put("a", "b"));
  ASSERT_OK(Put("c", "d"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("0", "f"));
  ASSERT_OK(Put("1", "h"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("2", "j"));
  auto iter = NewIterator(ReadOptions());
  iter->SeekForPrev(Slice("0"));
@@ -238,7 +238,7 @@ TEST_P(DBIteratorTest, IterLongKeys) {
  ASSERT_OK(Put(MakeLongKey(20, 0), "0"));
  ASSERT_OK(Put(MakeLongKey(32, 2), "2"));
  ASSERT_OK(Put("a", "b"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put(MakeLongKey(50, 1), "1"));
  ASSERT_OK(Put(MakeLongKey(127, 3), "3"));
  ASSERT_OK(Put(MakeLongKey(64, 4), "4"));
@@ -276,7 +276,7 @@ TEST_P(DBIteratorTest, IterLongKeys) {

 TEST_P(DBIteratorTest, IterNextWithNewerSeq) {
  ASSERT_OK(Put("0", "0"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("a", "b"));
  ASSERT_OK(Put("c", "d"));
  ASSERT_OK(Put("d", "e"));
@@ -302,7 +302,7 @@ TEST_P(DBIteratorTest, IterNextWithNewerSeq) {

 TEST_P(DBIteratorTest, IterPrevWithNewerSeq) {
  ASSERT_OK(Put("0", "0"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("a", "b"));
  ASSERT_OK(Put("c", "d"));
  ASSERT_OK(Put("d", "e"));
@@ -333,7 +333,7 @@ TEST_P(DBIteratorTest, IterPrevWithNewerSeq) {

 TEST_P(DBIteratorTest, IterPrevWithNewerSeq2) {
  ASSERT_OK(Put("0", "0"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("a", "b"));
  ASSERT_OK(Put("c", "d"));
  ASSERT_OK(Put("e", "f"));
@@ -377,6 +377,8 @@ TEST_P(DBIteratorTest, IterEmpty) {
    iter->SeekForPrev("foo");
    ASSERT_EQ(IterStatus(iter), "(invalid)");

+    ASSERT_OK(iter->status());
+
    delete iter;
  } while (ChangeCompactOptions());
 }
@@ -783,18 +785,18 @@ TEST_P(DBIteratorTest, IterWithSnapshot) {
 TEST_P(DBIteratorTest, IteratorPinsRef) {
  do {
    CreateAndReopenWithCF({"pikachu"}, CurrentOptions());
-    Put(1, "foo", "hello");
+    ASSERT_OK(Put(1, "foo", "hello"));

    // Get iterator that will yield the current contents of the DB.
    Iterator* iter = NewIterator(ReadOptions(), handles_[1]);

    // Write to force compactions
-    Put(1, "foo", "newvalue1");
+    ASSERT_OK(Put(1, "foo", "newvalue1"));
    for (int i = 0; i < 100; i++) {
      // 100K values
      ASSERT_OK(Put(1, Key(i), Key(i) + std::string(100000, 'v')));
    }
-    Put(1, "foo", "newvalue2");
+    ASSERT_OK(Put(1, "foo", "newvalue2"));

    iter->SeekToFirst();
    ASSERT_TRUE(iter->Valid());
@@ -809,8 +811,8 @@ TEST_P(DBIteratorTest, IteratorPinsRef) {
 TEST_P(DBIteratorTest, IteratorDeleteAfterCfDelete) {
  CreateAndReopenWithCF({"pikachu"}, CurrentOptions());

-  Put(1, "foo", "delete-cf-then-delete-iter");
-  Put(1, "hello", "value2");
+  ASSERT_OK(Put(1, "foo", "delete-cf-then-delete-iter"));
+  ASSERT_OK(Put(1, "hello", "value2"));

  ColumnFamilyHandle* cf = handles_[1];
  ReadOptions ro;
@@ -820,7 +822,7 @@ TEST_P(DBIteratorTest, IteratorDeleteAfterCfDelete) {
  ASSERT_EQ(IterStatus(iter), "foo->delete-cf-then-delete-iter");

  // delete CF handle
-  db_->DestroyColumnFamilyHandle(cf);
+  EXPECT_OK(db_->DestroyColumnFamilyHandle(cf));
  handles_.erase(std::begin(handles_) + 1);

  // delete Iterator after CF handle is deleted
@@ -832,7 +834,7 @@ TEST_P(DBIteratorTest, IteratorDeleteAfterCfDelete) {
 TEST_P(DBIteratorTest, IteratorDeleteAfterCfDrop) {
  CreateAndReopenWithCF({"pikachu"}, CurrentOptions());

-  Put(1, "foo", "drop-cf-then-delete-iter");
+  ASSERT_OK(Put(1, "foo", "drop-cf-then-delete-iter"));

  ReadOptions ro;
  ColumnFamilyHandle* cf = handles_[1];
@@ -842,8 +844,8 @@ TEST_P(DBIteratorTest, IteratorDeleteAfterCfDrop) {
  ASSERT_EQ(IterStatus(iter), "foo->drop-cf-then-delete-iter");

  // drop and delete CF
-  db_->DropColumnFamily(cf);
-  db_->DestroyColumnFamilyHandle(cf);
+  EXPECT_OK(db_->DropColumnFamily(cf));
+  EXPECT_OK(db_->DestroyColumnFamilyHandle(cf));
  handles_.erase(std::begin(handles_) + 1);

  // delete Iterator after CF handle is dropped
@@ -1307,9 +1309,9 @@ TEST_P(DBIteratorTest, PrevAfterAndNextAfterMerge) {

  // write three entries with different keys using Merge()
  WriteOptions wopts;
-  db_->Merge(wopts, "1", "data1");
-  db_->Merge(wopts, "2", "data2");
-  db_->Merge(wopts, "3", "data3");
+  ASSERT_OK(db_->Merge(wopts, "1", "data1"));
+  ASSERT_OK(db_->Merge(wopts, "2", "data2"));
+  ASSERT_OK(db_->Merge(wopts, "3", "data3"));

  std::unique_ptr<Iterator> it(NewIterator(ReadOptions()));

@@ -1393,7 +1395,7 @@ class DBIteratorTestForPinnedData : public DBIteratorTest {

      if (run_config == TestConfig::FLUSH_EVERY_1000) {
        if (i && i % 1000 == 0) {
-          Flush();
+          ASSERT_OK(Flush());
        }
      }
    }
@@ -1402,7 +1404,7 @@ class DBIteratorTestForPinnedData : public DBIteratorTest {
      Close();
      Reopen(options);
    } else if (run_config == TestConfig::COMPACT_BEFORE_READ) {
-      db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+      ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
    }

    ReadOptions ro;
@@ -1768,6 +1770,7 @@ TEST_P(DBIteratorTest, IterSeekForPrevCrossingFiles) {
    Iterator* iter = NewIterator(ro);
    iter->SeekForPrev("c2");
    ASSERT_TRUE(!iter->Valid());
+    ASSERT_OK(iter->status());
    delete iter;
  }
 }
@@ -1823,6 +1826,7 @@ TEST_P(DBIteratorTest, IterSeekForPrevCrossingFilesCustomPrefixExtractor) {
    Iterator* iter = NewIterator(ro);
    iter->SeekForPrev("c2");
    ASSERT_TRUE(!iter->Valid());
+    ASSERT_OK(iter->status());
    delete iter;
  }
 }
@@ -2151,19 +2155,19 @@ TEST_P(DBIteratorTest, ReadAhead) {

  std::string value(1024, 'a');
  for (int i = 0; i < 100; i++) {
-    Put(Key(i), value);
+    ASSERT_OK(Put(Key(i), value));
  }
  ASSERT_OK(Flush());
  MoveFilesToLevel(2);

  for (int i = 0; i < 100; i++) {
-    Put(Key(i), value);
+    ASSERT_OK(Put(Key(i), value));
  }
  ASSERT_OK(Flush());
  MoveFilesToLevel(1);

  for (int i = 0; i < 100; i++) {
-    Put(Key(i), value);
+    ASSERT_OK(Put(Key(i), value));
  }
  ASSERT_OK(Flush());
 #ifndef ROCKSDB_LITE
@@ -2270,6 +2274,7 @@ TEST_P(DBIteratorTest, Refresh) {
  ASSERT_OK(Put("x", "y"));

  std::unique_ptr<Iterator> iter(NewIterator(ReadOptions()));
+  ASSERT_OK(iter->status());
  iter->Seek(Slice("a"));
  ASSERT_TRUE(iter->Valid());
  ASSERT_EQ(iter->key().compare(Slice("x")), 0);
@@ -2284,7 +2289,8 @@ TEST_P(DBIteratorTest, Refresh) {
  iter->Next();
  ASSERT_FALSE(iter->Valid());

-  iter->Refresh();
+  ASSERT_OK(iter->status());
+  ASSERT_OK(iter->Refresh());

  iter->Seek(Slice("a"));
  ASSERT_TRUE(iter->Valid());
@@ -2295,7 +2301,7 @@ TEST_P(DBIteratorTest, Refresh) {
  iter->Next();
  ASSERT_FALSE(iter->Valid());

-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));

  ASSERT_OK(Put("m", "n"));

@@ -2308,7 +2314,8 @@ TEST_P(DBIteratorTest, Refresh) {
  iter->Next();
  ASSERT_FALSE(iter->Valid());

-  iter->Refresh();
+  ASSERT_OK(iter->status());
+  ASSERT_OK(iter->Refresh());

  iter->Seek(Slice("a"));
  ASSERT_TRUE(iter->Valid());
@@ -2331,6 +2338,7 @@ TEST_P(DBIteratorTest, RefreshWithSnapshot) {
  ReadOptions options;
  options.snapshot = snapshot;
  Iterator* iter = NewIterator(options);
+  ASSERT_OK(iter->status());

  iter->Seek(Slice("a"));
  ASSERT_TRUE(iter->Valid());
@@ -2346,8 +2354,8 @@ TEST_P(DBIteratorTest, RefreshWithSnapshot) {
  iter->Next();
  ASSERT_FALSE(iter->Valid());

-  Status s;
-  s = iter->Refresh();
+  ASSERT_OK(iter->status());
+  Status s = iter->Refresh();
  ASSERT_TRUE(s.IsNotSupported());
  db_->ReleaseSnapshot(snapshot);
  delete iter;
@@ -2405,14 +2413,14 @@ TEST_P(DBIteratorTest, UpperBoundWithChangeDirection) {

 TEST_P(DBIteratorTest, TableFilter) {
  ASSERT_OK(Put("a", "1"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("b", "2"));
  ASSERT_OK(Put("c", "3"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));
  ASSERT_OK(Put("d", "4"));
  ASSERT_OK(Put("e", "5"));
  ASSERT_OK(Put("f", "6"));
-  dbfull()->Flush(FlushOptions());
+  EXPECT_OK(dbfull()->Flush(FlushOptions()));

  // Ensure the table_filter callback is called once for each table.
  {
@@ -2597,13 +2605,13 @@ TEST_P(DBIteratorTest, SeekAfterHittingManyInternalKeys) {
  ReadOptions ropts;
  ropts.max_skippable_internal_keys = 2;

-  Put("1", "val_1");
+  ASSERT_OK(Put("1", "val_1"));
  // Add more tombstones than max_skippable_internal_keys so that Next() fails.
-  Delete("2");
-  Delete("3");
-  Delete("4");
-  Delete("5");
-  Put("6", "val_6");
+  ASSERT_OK(Delete("2"));
+  ASSERT_OK(Delete("3"));
+  ASSERT_OK(Delete("4"));
+  ASSERT_OK(Delete("5"));
+  ASSERT_OK(Put("6", "val_6"));

  std::unique_ptr<Iterator> iter(NewIterator(ropts));
  iter->SeekToFirst();
@@ -2645,9 +2653,9 @@ TEST_P(DBIteratorTest, NonBlockingIterationBugRepro) {
  DestroyAndReopen(options);

  // Two records in sst file, each in its own block.
-  Put("b", "");
-  Put("d", "");
-  Flush();
+  ASSERT_OK(Put("b", ""));
+  ASSERT_OK(Put("d", ""));
+  ASSERT_OK(Flush());

  // Create a nonblocking iterator before writing to memtable.
  ReadOptions ropt;
@@ -2657,7 +2665,7 @@ TEST_P(DBIteratorTest, NonBlockingIterationBugRepro) {
  // Overwrite a key in memtable many times to hit
  // max_sequential_skip_in_iterations (which is 8 by default).
  for (int i = 0; i < 20; ++i) {
-    Put("c", "");
+    ASSERT_OK(Put("c", ""));
  }

  // Load the second block in sst file into the block cache.
@@ -2674,9 +2682,9 @@ TEST_P(DBIteratorTest, NonBlockingIterationBugRepro) {
 }

 TEST_P(DBIteratorTest, SeekBackwardAfterOutOfUpperBound) {
-  Put("a", "");
-  Put("b", "");
-  Flush();
+  ASSERT_OK(Put("a", ""));
+  ASSERT_OK(Put("b", ""));
+  ASSERT_OK(Flush());

  ReadOptions ropt;
  Slice ub = "b";
@@ -2883,6 +2891,127 @@ TEST_P(DBIteratorTest, IterateWithLowerBoundAcrossFileBoundary) {
  ASSERT_OK(iter->status());
 }

+TEST_P(DBIteratorTest, Blob) {
+  Options options = CurrentOptions();
+  options.enable_blob_files = true;
+  options.max_sequential_skip_in_iterations = 2;
+  options.statistics = CreateDBStatistics();
+
+  Reopen(options);
+
+  // Note: we have 4 KVs (3 of which are hidden) for key "b" and
+  // max_sequential_skip_in_iterations is set to 2. Thus, we need to do a reseek
+  // anytime we move from "b" to "c" or vice versa.
+  ASSERT_OK(Put("a", "va"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("b", "vb0"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("b", "vb1"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("b", "vb2"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("b", "vb3"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("c", "vc"));
+  ASSERT_OK(Flush());
+
+  std::unique_ptr<Iterator> iter_guard(NewIterator(ReadOptions()));
+  Iterator* const iter = iter_guard.get();
+
+  iter->SeekToFirst();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 0);
+  ASSERT_EQ(IterStatus(iter), "a->va");
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 0);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 1);
+  ASSERT_EQ(IterStatus(iter), "c->vc");
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 1);
+  ASSERT_EQ(IterStatus(iter), "(invalid)");
+  iter->SeekToFirst();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 1);
+  ASSERT_EQ(IterStatus(iter), "a->va");
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 1);
+  ASSERT_EQ(IterStatus(iter), "(invalid)");
+
+  iter->SeekToLast();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 1);
+  ASSERT_EQ(IterStatus(iter), "c->vc");
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "a->va");
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "(invalid)");
+  iter->SeekToLast();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "c->vc");
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "(invalid)");
+
+  iter->Seek("");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "a->va");
+  iter->Seek("a");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "a->va");
+  iter->Seek("ax");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+
+  iter->SeekForPrev("d");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "c->vc");
+  iter->SeekForPrev("c");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 2);
+  ASSERT_EQ(IterStatus(iter), "c->vc");
+  iter->SeekForPrev("bx");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 3);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+
+  iter->Seek("b");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 3);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+  iter->Seek("z");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 3);
+  ASSERT_EQ(IterStatus(iter), "(invalid)");
+  iter->SeekForPrev("b");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 4);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+  iter->SeekForPrev("");
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 4);
+  ASSERT_EQ(IterStatus(iter), "(invalid)");
+
+  // Switch from reverse to forward
+  iter->SeekToLast();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 4);
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+
+  // Switch from forward to reverse
+  iter->SeekToFirst();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  iter->Next();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 6);
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 7);
+  ASSERT_EQ(IterStatus(iter), "b->vb3");
+}
+
 INSTANTIATE_TEST_CASE_P(DBIteratorTestInstance, DBIteratorTest,
                        testing::Values(true, false));

@@ -33,9 +33,8 @@ class DBTestXactLogIterator : public DBTestBase {
 };

 namespace {
-SequenceNumber ReadRecords(
-    std::unique_ptr<TransactionLogIterator>& iter,
-    int& count) {
+SequenceNumber ReadRecords(std::unique_ptr<TransactionLogIterator>& iter,
+                           int& count, bool expect_ok = true) {
  count = 0;
  SequenceNumber lastSequence = 0;
  BatchResult res;
@@ -47,6 +46,11 @@ SequenceNumber ReadRecords(
    EXPECT_OK(iter->status());
    iter->Next();
  }
+  if (expect_ok) {
+    EXPECT_OK(iter->status());
+  } else {
+    EXPECT_NOK(iter->status());
+  }
  return res.sequence;
 }

@@ -64,9 +68,9 @@ TEST_F(DBTestXactLogIterator, TransactionLogIterator) {
    Options options = OptionsForLogIterTest();
    DestroyAndReopen(options);
    CreateAndReopenWithCF({"pikachu"}, options);
-    Put(0, "key1", DummyString(1024));
-    Put(1, "key2", DummyString(1024));
-    Put(1, "key2", DummyString(1024));
+    ASSERT_OK(Put(0, "key1", DummyString(1024)));
+    ASSERT_OK(Put(1, "key2", DummyString(1024)));
+    ASSERT_OK(Put(1, "key2", DummyString(1024)));
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 3U);
    {
      auto iter = OpenTransactionLogIter(0);
@@ -75,9 +79,9 @@ TEST_F(DBTestXactLogIterator, TransactionLogIterator) {
    ReopenWithColumnFamilies({"default", "pikachu"}, options);
    env_->SleepForMicroseconds(2 * 1000 * 1000);
    {
-      Put(0, "key4", DummyString(1024));
-      Put(1, "key5", DummyString(1024));
-      Put(0, "key6", DummyString(1024));
+      ASSERT_OK(Put(0, "key4", DummyString(1024)));
+      ASSERT_OK(Put(1, "key5", DummyString(1024)));
+      ASSERT_OK(Put(0, "key6", DummyString(1024)));
    }
    {
      auto iter = OpenTransactionLogIter(0);
@@ -109,15 +113,15 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorRace) {
      ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
      Options options = OptionsForLogIterTest();
      DestroyAndReopen(options);
-      Put("key1", DummyString(1024));
-      dbfull()->Flush(FlushOptions());
-      Put("key2", DummyString(1024));
-      dbfull()->Flush(FlushOptions());
-      Put("key3", DummyString(1024));
-      dbfull()->Flush(FlushOptions());
-      Put("key4", DummyString(1024));
+      ASSERT_OK(Put("key1", DummyString(1024)));
+      ASSERT_OK(dbfull()->Flush(FlushOptions()));
+      ASSERT_OK(Put("key2", DummyString(1024)));
+      ASSERT_OK(dbfull()->Flush(FlushOptions()));
+      ASSERT_OK(Put("key3", DummyString(1024)));
+      ASSERT_OK(dbfull()->Flush(FlushOptions()));
+      ASSERT_OK(Put("key4", DummyString(1024)));
      ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 4U);
-      dbfull()->FlushWAL(false);
+      ASSERT_OK(dbfull()->FlushWAL(false));

      {
        auto iter = OpenTransactionLogIter(0);
@@ -130,11 +134,11 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorRace) {
      // condition
      FlushOptions flush_options;
      flush_options.wait = false;
-      dbfull()->Flush(flush_options);
+      ASSERT_OK(dbfull()->Flush(flush_options));

      // "key5" would be written in a new memtable and log
-      Put("key5", DummyString(1024));
-      dbfull()->FlushWAL(false);
+      ASSERT_OK(Put("key5", DummyString(1024)));
+      ASSERT_OK(dbfull()->FlushWAL(false));
      {
        // this iter would miss "key4" if not fixed
        auto iter = OpenTransactionLogIter(0);
@@ -149,14 +153,14 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorStallAtLastRecord) {
  do {
    Options options = OptionsForLogIterTest();
    DestroyAndReopen(options);
-    Put("key1", DummyString(1024));
+    ASSERT_OK(Put("key1", DummyString(1024)));
    auto iter = OpenTransactionLogIter(0);
    ASSERT_OK(iter->status());
    ASSERT_TRUE(iter->Valid());
    iter->Next();
    ASSERT_TRUE(!iter->Valid());
    ASSERT_OK(iter->status());
-    Put("key2", DummyString(1024));
+    ASSERT_OK(Put("key2", DummyString(1024)));
    iter->Next();
    ASSERT_OK(iter->status());
    ASSERT_TRUE(iter->Valid());
@@ -167,9 +171,9 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorCheckAfterRestart) {
  do {
    Options options = OptionsForLogIterTest();
    DestroyAndReopen(options);
-    Put("key1", DummyString(1024));
-    Put("key2", DummyString(1023));
-    dbfull()->Flush(FlushOptions());
+    ASSERT_OK(Put("key1", DummyString(1024)));
+    ASSERT_OK(Put("key2", DummyString(1023)));
+    ASSERT_OK(dbfull()->Flush(FlushOptions()));
    Reopen(options);
    auto iter = OpenTransactionLogIter(0);
    ExpectRecords(2, iter);
@@ -181,10 +185,10 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorCorruptedLog) {
    Options options = OptionsForLogIterTest();
    DestroyAndReopen(options);
    for (int i = 0; i < 1024; i++) {
-      Put("key"+ToString(i), DummyString(10));
+      ASSERT_OK(Put("key" + ToString(i), DummyString(10)));
    }
-    dbfull()->Flush(FlushOptions());
-    dbfull()->FlushWAL(false);
+    ASSERT_OK(dbfull()->Flush(FlushOptions()));
+    ASSERT_OK(dbfull()->FlushWAL(false));
    // Corrupt this log to create a gap
    ROCKSDB_NAMESPACE::VectorLogPtr wal_files;
    ASSERT_OK(dbfull()->GetSortedWalFiles(wal_files));
@@ -197,13 +201,13 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorCorruptedLog) {
    }

    // Insert a new entry to a new log file
-    Put("key1025", DummyString(10));
-    dbfull()->FlushWAL(false);
+    ASSERT_OK(Put("key1025", DummyString(10)));
+    ASSERT_OK(dbfull()->FlushWAL(false));
    // Try to read from the beginning. Should stop before the gap and read less
    // than 1025 entries
    auto iter = OpenTransactionLogIter(0);
    int count;
-    SequenceNumber last_sequence_read = ReadRecords(iter, count);
+    SequenceNumber last_sequence_read = ReadRecords(iter, count, false);
    ASSERT_LT(last_sequence_read, 1025U);
    // Try to read past the gap, should be able to seek to key1025
    auto iter2 = OpenTransactionLogIter(last_sequence_read + 1);
@@ -217,15 +221,15 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorBatchOperations) {
    DestroyAndReopen(options);
    CreateAndReopenWithCF({"pikachu"}, options);
    WriteBatch batch;
-    batch.Put(handles_[1], "key1", DummyString(1024));
-    batch.Put(handles_[0], "key2", DummyString(1024));
-    batch.Put(handles_[1], "key3", DummyString(1024));
-    batch.Delete(handles_[0], "key2");
-    dbfull()->Write(WriteOptions(), &batch);
-    Flush(1);
-    Flush(0);
+    ASSERT_OK(batch.Put(handles_[1], "key1", DummyString(1024)));
+    ASSERT_OK(batch.Put(handles_[0], "key2", DummyString(1024)));
+    ASSERT_OK(batch.Put(handles_[1], "key3", DummyString(1024)));
+    ASSERT_OK(batch.Delete(handles_[0], "key2"));
+    ASSERT_OK(dbfull()->Write(WriteOptions(), &batch));
+    ASSERT_OK(Flush(1));
+    ASSERT_OK(Flush(0));
    ReopenWithColumnFamilies({"default", "pikachu"}, options);
-    Put(1, "key4", DummyString(1024));
+    ASSERT_OK(Put(1, "key4", DummyString(1024)));
    auto iter = OpenTransactionLogIter(3);
    ExpectRecords(2, iter);
  } while (ChangeCompactOptions());
@@ -237,13 +241,13 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorBlobs) {
  CreateAndReopenWithCF({"pikachu"}, options);
  {
    WriteBatch batch;
-    batch.Put(handles_[1], "key1", DummyString(1024));
-    batch.Put(handles_[0], "key2", DummyString(1024));
-    batch.PutLogData(Slice("blob1"));
-    batch.Put(handles_[1], "key3", DummyString(1024));
-    batch.PutLogData(Slice("blob2"));
-    batch.Delete(handles_[0], "key2");
-    dbfull()->Write(WriteOptions(), &batch);
+    ASSERT_OK(batch.Put(handles_[1], "key1", DummyString(1024)));
+    ASSERT_OK(batch.Put(handles_[0], "key2", DummyString(1024)));
+    ASSERT_OK(batch.PutLogData(Slice("blob1")));
+    ASSERT_OK(batch.Put(handles_[1], "key3", DummyString(1024)));
+    ASSERT_OK(batch.PutLogData(Slice("blob2")));
+    ASSERT_OK(batch.Delete(handles_[0], "key2"));
+    ASSERT_OK(dbfull()->Write(WriteOptions(), &batch));
    ReopenWithColumnFamilies({"default", "pikachu"}, options);
  }

@@ -268,7 +272,7 @@ TEST_F(DBTestXactLogIterator, TransactionLogIteratorBlobs) {
      return Status::OK();
    }
  } handler;
-  res.writeBatchPtr->Iterate(&handler);
+  ASSERT_OK(res.writeBatchPtr->Iterate(&handler));
  ASSERT_EQ(
      "Put(1, key1, 1024)"
      "Put(0, key2, 1024)"
@@ -401,7 +401,7 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithDifferentPaths) {
  ColumnFamilyOptions cf_options0;
  cf_options0.cf_paths = {{cf_path_0_, 1024}};
  ColumnFamilyHandle* cf0;
-  db0->CreateColumnFamily(cf_options0, "cf", &cf0);
+  ASSERT_OK(db0->CreateColumnFamily(cf_options0, "cf", &cf0));
  ASSERT_EQ(2, cache_->Size());
  ASSERT_TRUE(cache_->Contains(data_path_0_));
  ASSERT_EQ(1, cache_->GetRefCount(data_path_0_));
@@ -421,7 +421,7 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithDifferentPaths) {
  ColumnFamilyOptions cf_options1;
  cf_options1.cf_paths = {{cf_path_1_, 1024}};
  ColumnFamilyHandle* cf1;
-  db1->CreateColumnFamily(cf_options1, "cf", &cf1);
+  ASSERT_OK(db1->CreateColumnFamily(cf_options1, "cf", &cf1));
  ASSERT_EQ(4, cache_->Size());
  ASSERT_TRUE(cache_->Contains(data_path_0_));
  ASSERT_EQ(1, cache_->GetRefCount(data_path_0_));
@@ -432,7 +432,7 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithDifferentPaths) {
  ASSERT_TRUE(cache_->Contains(cf_path_1_));
  ASSERT_EQ(1, cache_->GetRefCount(cf_path_1_));

-  db0->DestroyColumnFamilyHandle(cf0);
+  ASSERT_OK(db0->DestroyColumnFamilyHandle(cf0));
  delete db0;
  ASSERT_EQ(2, cache_->Size());
  ASSERT_TRUE(cache_->Contains(data_path_1_));
@@ -441,7 +441,7 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithDifferentPaths) {
  ASSERT_EQ(1, cache_->GetRefCount(cf_path_1_));
  ASSERT_OK(DestroyDB(data_path_0_, options, {{"cf", cf_options0}}));

-  db1->DestroyColumnFamilyHandle(cf1);
+  ASSERT_OK(db1->DestroyColumnFamilyHandle(cf1));
  delete db1;
  ASSERT_EQ(0, cache_->Size());
  ASSERT_OK(DestroyDB(data_path_1_, options, {{"cf", cf_options1}}));
@@ -466,7 +466,7 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithSamePaths) {
  ASSERT_EQ(1, cache_->GetRefCount(data_path_0_));

  ColumnFamilyHandle* cf0;
-  db0->CreateColumnFamily(cf_options, "cf", &cf0);
+  ASSERT_OK(db0->CreateColumnFamily(cf_options, "cf", &cf0));
  ASSERT_EQ(2, cache_->Size());
  ASSERT_TRUE(cache_->Contains(data_path_0_));
  ASSERT_EQ(1, cache_->GetRefCount(data_path_0_));
@@ -482,14 +482,14 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithSamePaths) {
  ASSERT_EQ(1, cache_->GetRefCount(cf_path_0_));

  ColumnFamilyHandle* cf1;
-  db1->CreateColumnFamily(cf_options, "cf", &cf1);
+  ASSERT_OK(db1->CreateColumnFamily(cf_options, "cf", &cf1));
  ASSERT_EQ(2, cache_->Size());
  ASSERT_TRUE(cache_->Contains(data_path_0_));
  ASSERT_EQ(2, cache_->GetRefCount(data_path_0_));
  ASSERT_TRUE(cache_->Contains(cf_path_0_));
  ASSERT_EQ(2, cache_->GetRefCount(cf_path_0_));

-  db0->DestroyColumnFamilyHandle(cf0);
+  ASSERT_OK(db0->DestroyColumnFamilyHandle(cf0));
  delete db0;
  ASSERT_EQ(2, cache_->Size());
  ASSERT_TRUE(cache_->Contains(data_path_0_));
@@ -498,7 +498,7 @@ TEST_F(DBLogicalBlockSizeCacheTest, MultiDBWithSamePaths) {
  ASSERT_EQ(1, cache_->GetRefCount(cf_path_0_));
  ASSERT_OK(DestroyDB(dbname_ + "/db0", options, {{"cf", cf_options}}));

-  db1->DestroyColumnFamilyHandle(cf1);
+  ASSERT_OK(db1->DestroyColumnFamilyHandle(cf1));
  delete db1;
  ASSERT_EQ(0, cache_->Size());
  ASSERT_OK(DestroyDB(dbname_ + "/db1", options, {{"cf", cf_options}}));
@@ -129,7 +129,6 @@ class TestPrefixExtractor : public SliceTransform {
 TEST_F(DBMemTableTest, DuplicateSeq) {
  SequenceNumber seq = 123;
  std::string value;
-  Status s;
  MergeContext merge_context;
  Options options;
  InternalKeyComparator ikey_cmp(options.comparator);
@@ -146,22 +145,15 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
                               kMaxSequenceNumber, 0 /* column_family_id */);

  // Write some keys and make sure it returns false on duplicates
-  bool res;
-  res = mem->Add(seq, kTypeValue, "key", "value2");
-  ASSERT_TRUE(res);
-  res = mem->Add(seq, kTypeValue, "key", "value2");
-  ASSERT_FALSE(res);
+  ASSERT_OK(mem->Add(seq, kTypeValue, "key", "value2"));
+  ASSERT_TRUE(mem->Add(seq, kTypeValue, "key", "value2").IsTryAgain());
  // Changing the type should still cause the duplicatae key
-  res = mem->Add(seq, kTypeMerge, "key", "value2");
-  ASSERT_FALSE(res);
+  ASSERT_TRUE(mem->Add(seq, kTypeMerge, "key", "value2").IsTryAgain());
  // Changing the seq number will make the key fresh
-  res = mem->Add(seq + 1, kTypeMerge, "key", "value2");
-  ASSERT_TRUE(res);
+  ASSERT_OK(mem->Add(seq + 1, kTypeMerge, "key", "value2"));
  // Test with different types for duplicate keys
-  res = mem->Add(seq, kTypeDeletion, "key", "");
-  ASSERT_FALSE(res);
-  res = mem->Add(seq, kTypeSingleDeletion, "key", "");
-  ASSERT_FALSE(res);
+  ASSERT_TRUE(mem->Add(seq, kTypeDeletion, "key", "").IsTryAgain());
+  ASSERT_TRUE(mem->Add(seq, kTypeSingleDeletion, "key", "").IsTryAgain());

  // Test the duplicate keys under stress
  for (int i = 0; i < 10000; i++) {
@@ -169,11 +161,11 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
    if (!insert_dup) {
      seq++;
    }
-    res = mem->Add(seq, kTypeValue, "foo", "value" + ToString(seq));
+    Status s = mem->Add(seq, kTypeValue, "foo", "value" + ToString(seq));
    if (insert_dup) {
-      ASSERT_FALSE(res);
+      ASSERT_TRUE(s.IsTryAgain());
    } else {
-      ASSERT_TRUE(res);
+      ASSERT_OK(s);
    }
  }
  delete mem;
@@ -185,10 +177,8 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
  mem = new MemTable(cmp, ioptions, MutableCFOptions(options), &wb,
                     kMaxSequenceNumber, 0 /* column_family_id */);
  // Insert a duplicate key with _ in it
-  res = mem->Add(seq, kTypeValue, "key_1", "value");
-  ASSERT_TRUE(res);
-  res = mem->Add(seq, kTypeValue, "key_1", "value");
-  ASSERT_FALSE(res);
+  ASSERT_OK(mem->Add(seq, kTypeValue, "key_1", "value"));
+  ASSERT_TRUE(mem->Add(seq, kTypeValue, "key_1", "value").IsTryAgain());
  delete mem;

  // Test when InsertConcurrently will be invoked
@@ -197,10 +187,11 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
  mem = new MemTable(cmp, ioptions, MutableCFOptions(options), &wb,
                     kMaxSequenceNumber, 0 /* column_family_id */);
  MemTablePostProcessInfo post_process_info;
-  res = mem->Add(seq, kTypeValue, "key", "value", true, &post_process_info);
-  ASSERT_TRUE(res);
-  res = mem->Add(seq, kTypeValue, "key", "value", true, &post_process_info);
-  ASSERT_FALSE(res);
+  ASSERT_OK(
+      mem->Add(seq, kTypeValue, "key", "value", true, &post_process_info));
+  ASSERT_TRUE(
+      mem->Add(seq, kTypeValue, "key", "value", true, &post_process_info)
+          .IsTryAgain());
  delete mem;
 }

@@ -208,7 +199,6 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
 TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
  int num_ops = 1000;
  std::string value;
-  Status s;
  MergeContext merge_context;
  Options options;
  // A merge operator that is not sensitive to concurrent writes since in this
@@ -227,8 +217,7 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {

  // Put 0 as the base
  PutFixed64(&value, static_cast<uint64_t>(0));
-  bool res = mem->Add(0, kTypeValue, "key", value);
-  ASSERT_TRUE(res);
+  ASSERT_OK(mem->Add(0, kTypeValue, "key", value));
  value.clear();

  // Write Merge concurrently
@@ -237,9 +226,8 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
    std::string v1;
    for (int seq = 1; seq < num_ops / 2; seq++) {
      PutFixed64(&v1, seq);
-      bool res1 =
-          mem->Add(seq, kTypeMerge, "key", v1, true, &post_process_info1);
-      ASSERT_TRUE(res1);
+      ASSERT_OK(
+          mem->Add(seq, kTypeMerge, "key", v1, true, &post_process_info1));
      v1.clear();
    }
  });
@@ -248,9 +236,8 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
    std::string v2;
    for (int seq = num_ops / 2; seq < num_ops; seq++) {
      PutFixed64(&v2, seq);
-      bool res2 =
-          mem->Add(seq, kTypeMerge, "key", v2, true, &post_process_info2);
-      ASSERT_TRUE(res2);
+      ASSERT_OK(
+          mem->Add(seq, kTypeMerge, "key", v2, true, &post_process_info2));
      v2.clear();
    }
  });
@@ -261,8 +248,9 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
  ReadOptions roptions;
  SequenceNumber max_covering_tombstone_seq = 0;
  LookupKey lkey("key", kMaxSequenceNumber);
-  res = mem->Get(lkey, &value, /*timestamp=*/nullptr, &status, &merge_context,
-                 &max_covering_tombstone_seq, roptions);
+  bool res = mem->Get(lkey, &value, /*timestamp=*/nullptr, &status,
+                      &merge_context, &max_covering_tombstone_seq, roptions);
+  ASSERT_OK(status);
  ASSERT_TRUE(res);
  uint64_t ivalue = DecodeFixed64(Slice(value).data());
  uint64_t sum = 0;
@@ -59,29 +59,29 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  merge_operands_info.expected_max_number_of_operands = num_records;

  // k0 value in memtable
-  Put("k0", "PutARock");
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k0",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(Put("k0", "PutARock"));
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k0", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "PutARock");

  // k0.1 value in SST
-  Put("k0.1", "RockInSST");
+  ASSERT_OK(Put("k0.1", "RockInSST"));
  ASSERT_OK(Flush());
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k0.1",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k0.1", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "RockInSST");

  // All k1 values are in memtable.
  ASSERT_OK(Merge("k1", "a"));
-  Put("k1", "x");
+  ASSERT_OK(Put("k1", "x"));
  ASSERT_OK(Merge("k1", "b"));
  ASSERT_OK(Merge("k1", "c"));
  ASSERT_OK(Merge("k1", "d"));
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k1",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k1", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "x");
  ASSERT_EQ(values[1], "b");
  ASSERT_EQ(values[2], "c");
@@ -98,13 +98,13 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {

  // All k1.1 values are in memtable.
  ASSERT_OK(Merge("k1.1", "r"));
-  Delete("k1.1");
+  ASSERT_OK(Delete("k1.1"));
  ASSERT_OK(Merge("k1.1", "c"));
  ASSERT_OK(Merge("k1.1", "k"));
  ASSERT_OK(Merge("k1.1", "s"));
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k1.1",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k1.1", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "c");
  ASSERT_EQ(values[1], "k");
  ASSERT_EQ(values[2], "s");
@@ -115,9 +115,9 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  ASSERT_OK(Merge("k2", "e"));
  ASSERT_OK(Merge("k2", "r"));
  ASSERT_OK(Flush());
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k2",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k2", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "q");
  ASSERT_EQ(values[1], "w");
  ASSERT_EQ(values[2], "e");
@@ -125,30 +125,30 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {

  // All k2.1 values are flushed to L0 into a single file.
  ASSERT_OK(Merge("k2.1", "m"));
-  Put("k2.1", "l");
+  ASSERT_OK(Put("k2.1", "l"));
  ASSERT_OK(Merge("k2.1", "n"));
  ASSERT_OK(Merge("k2.1", "o"));
  ASSERT_OK(Flush());
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k2.1",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k2.1", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "l,n,o");

  // All k2.2 values are flushed to L0 into a single file.
  ASSERT_OK(Merge("k2.2", "g"));
-  Delete("k2.2");
+  ASSERT_OK(Delete("k2.2"));
  ASSERT_OK(Merge("k2.2", "o"));
  ASSERT_OK(Merge("k2.2", "t"));
  ASSERT_OK(Flush());
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k2.2",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k2.2", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "o,t");

  // Do some compaction that will make the following tests more predictable
  //  Slice start("PutARock");
  //  Slice end("t");
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  // All k3 values are flushed and are in different files.
  ASSERT_OK(Merge("k3", "ab"));
@@ -158,9 +158,9 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  ASSERT_OK(Merge("k3", "cd"));
  ASSERT_OK(Flush());
  ASSERT_OK(Merge("k3", "de"));
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k3",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k3", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "ab");
  ASSERT_EQ(values[1], "bc");
  ASSERT_EQ(values[2], "cd");
@@ -169,14 +169,14 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  // All k3.1 values are flushed and are in different files.
  ASSERT_OK(Merge("k3.1", "ab"));
  ASSERT_OK(Flush());
-  Put("k3.1", "bc");
+  ASSERT_OK(Put("k3.1", "bc"));
  ASSERT_OK(Flush());
  ASSERT_OK(Merge("k3.1", "cd"));
  ASSERT_OK(Flush());
  ASSERT_OK(Merge("k3.1", "de"));
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k3.1",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k3.1", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "bc");
  ASSERT_EQ(values[1], "cd");
  ASSERT_EQ(values[2], "de");
@@ -184,14 +184,14 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  // All k3.2 values are flushed and are in different files.
  ASSERT_OK(Merge("k3.2", "ab"));
  ASSERT_OK(Flush());
-  Delete("k3.2");
+  ASSERT_OK(Delete("k3.2"));
  ASSERT_OK(Flush());
  ASSERT_OK(Merge("k3.2", "cd"));
  ASSERT_OK(Flush());
  ASSERT_OK(Merge("k3.2", "de"));
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k3.2",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k3.2", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "cd");
  ASSERT_EQ(values[1], "de");

@@ -206,9 +206,9 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  ASSERT_OK(Flush());
  MoveFilesToLevel(1);
  ASSERT_OK(Merge("k4", "ed"));
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k4",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k4", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "ba");
  ASSERT_EQ(values[1], "cb");
  ASSERT_EQ(values[2], "dc");
@@ -219,14 +219,14 @@ TEST_F(DBMergeOperandTest, GetMergeOperandsBasic) {
  ASSERT_OK(Merge("k5", "am"));
  ASSERT_OK(Merge("k5", "i"));
  ASSERT_OK(Flush());
-  Put("k5", "remember");
+  ASSERT_OK(Put("k5", "remember"));
  ASSERT_OK(Merge("k5", "i"));
  ASSERT_OK(Merge("k5", "am"));
  ASSERT_OK(Merge("k5", "rocks"));
-  dbfull()->TEST_SwitchMemtable();
-  db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(), "k5",
-                        values.data(), &merge_operands_info,
-                        &number_of_operands);
+  ASSERT_OK(dbfull()->TEST_SwitchMemtable());
+  ASSERT_OK(db_->GetMergeOperands(ReadOptions(), db_->DefaultColumnFamily(),
+                                  "k5", values.data(), &merge_operands_info,
+                                  &number_of_operands));
  ASSERT_EQ(values[0], "remember");
  ASSERT_EQ(values[1], "i");
  ASSERT_EQ(values[2], "am");
@@ -94,7 +94,7 @@ TEST_F(DBMergeOperatorTest, LimitMergeOperands) {
  ASSERT_OK(Merge("k1", "c"));
  ASSERT_OK(Merge("k1", "d"));
  std::string value;
-  ASSERT_TRUE(db_->Get(ReadOptions(), "k1", &value).ok());
+  ASSERT_OK(db_->Get(ReadOptions(), "k1", &value));
  // Make sure that only the latest two merge operands are used. If this was
  // not the case the value would be "a,b,c,d".
  ASSERT_EQ(value, "c,d");
@@ -105,7 +105,7 @@ TEST_F(DBMergeOperatorTest, LimitMergeOperands) {
  ASSERT_OK(Merge("k2", "c"));
  ASSERT_OK(Merge("k2", "d"));
  ASSERT_OK(Flush());
-  ASSERT_TRUE(db_->Get(ReadOptions(), "k2", &value).ok());
+  ASSERT_OK(db_->Get(ReadOptions(), "k2", &value));
  ASSERT_EQ(value, "c,d");

  // All K3 values are flushed and are in different files.
@@ -116,7 +116,7 @@ TEST_F(DBMergeOperatorTest, LimitMergeOperands) {
  ASSERT_OK(Merge("k3", "cd"));
  ASSERT_OK(Flush());
  ASSERT_OK(Merge("k3", "de"));
-  ASSERT_TRUE(db_->Get(ReadOptions(), "k3", &value).ok());
+  ASSERT_OK(db_->Get(ReadOptions(), "k3", &value));
  ASSERT_EQ(value, "cd,de");

  // All K4 values are in different levels
@@ -130,7 +130,7 @@ TEST_F(DBMergeOperatorTest, LimitMergeOperands) {
  ASSERT_OK(Flush());
  MoveFilesToLevel(1);
  ASSERT_OK(Merge("k4", "de"));
-  ASSERT_TRUE(db_->Get(ReadOptions(), "k4", &value).ok());
+  ASSERT_OK(db_->Get(ReadOptions(), "k4", &value));
  ASSERT_EQ(value, "cd,de");
 }

@@ -344,8 +344,9 @@ TEST_P(MergeOperatorPinningTest, EvictCacheBeforeMerge) {
  // Code executed before merge operation
  merge_hook->before_merge_ = [&]() {
    // Evict all tables from cache before every merge operation
+    auto* table_cache = dbfull()->TEST_table_cache();
    for (uint64_t num : file_numbers) {
-      TableCache::Evict(dbfull()->TEST_table_cache(), num);
+      TableCache::Evict(table_cache, num);
    }
    // Decrease cache capacity to force all unrefed blocks to be evicted
    if (bbto.block_cache) {
@@ -366,7 +367,7 @@ TEST_P(MergeOperatorPinningTest, EvictCacheBeforeMerge) {
  VerifyDBFromMap(true_data, &total_reads);
  ASSERT_EQ(merge_cnt, total_reads);

-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  VerifyDBFromMap(true_data, &total_reads);
 }
@@ -385,7 +386,7 @@ TEST_P(MergeOperatorPinningTest, TailingIterator) {
  std::function<void()> writer_func = [&]() {
    int k = 0;
    for (int i = 0; i < kNumWrites; i++) {
-      db_->Merge(WriteOptions(), Key(k), Key(k));
+      ASSERT_OK(db_->Merge(WriteOptions(), Key(k), Key(k)));

      if (i && i % kNumOperands == 0) {
        k++;
@@ -403,7 +404,7 @@ TEST_P(MergeOperatorPinningTest, TailingIterator) {
    ReadOptions ro;
    ro.tailing = true;
    Iterator* iter = db_->NewIterator(ro);
-
+    ASSERT_OK(iter->status());
    iter->SeekToFirst();
    for (int i = 0; i < (kNumWrites / kNumOperands); i++) {
      while (!iter->Valid()) {
@@ -416,6 +417,7 @@ TEST_P(MergeOperatorPinningTest, TailingIterator) {

      iter->Next();
    }
+    ASSERT_OK(iter->status());

    delete iter;
  };
@@ -449,12 +451,13 @@ TEST_F(DBMergeOperatorTest, TailingIteratorMemtableUnrefedBySomeoneElse) {
  //    ForwardIterator to not pin it in some circumstances. This test
  //    reproduces it.

-  db_->Merge(WriteOptions(), "key", "sst");
-  db_->Flush(FlushOptions()); // Switch to SuperVersion A
-  db_->Merge(WriteOptions(), "key", "memtable");
+  ASSERT_OK(db_->Merge(WriteOptions(), "key", "sst"));
+  ASSERT_OK(db_->Flush(FlushOptions()));  // Switch to SuperVersion A
+  ASSERT_OK(db_->Merge(WriteOptions(), "key", "memtable"));

  // Pin SuperVersion A
  std::unique_ptr<Iterator> someone_else(db_->NewIterator(ReadOptions()));
+  ASSERT_OK(someone_else->status());

  bool pushed_first_operand = false;
  bool stepped_to_next_operand = false;
@@ -462,7 +465,7 @@ TEST_F(DBMergeOperatorTest, TailingIteratorMemtableUnrefedBySomeoneElse) {
      "DBIter::MergeValuesNewToOld:PushedFirstOperand", [&](void*) {
        EXPECT_FALSE(pushed_first_operand);
        pushed_first_operand = true;
-        db_->Flush(FlushOptions()); // Switch to SuperVersion B
+        EXPECT_OK(db_->Flush(FlushOptions()));  // Switch to SuperVersion B
      });
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBIter::MergeValuesNewToOld:SteppedToNextOperand", [&](void*) {
@@ -477,7 +480,7 @@ TEST_F(DBMergeOperatorTest, TailingIteratorMemtableUnrefedBySomeoneElse) {
  std::unique_ptr<Iterator> iter(db_->NewIterator(ro));
  iter->Seek("key");

-  ASSERT_TRUE(iter->status().ok());
+  ASSERT_OK(iter->status());
  ASSERT_TRUE(iter->Valid());
  EXPECT_EQ(std::string("sst,memtable"), iter->value().ToString());
  EXPECT_TRUE(pushed_first_operand);
@@ -729,7 +729,7 @@ TEST_F(DBOptionsTest, SetFIFOCompactionOptions) {
    for (int j = 0; j < 10; j++) {
      ASSERT_OK(Put(ToString(i * 20 + j), rnd.RandomString(980)));
    }
-    Flush();
+    ASSERT_OK(Flush());
  }
  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(NumTableFilesAtLevel(0), 10);
@@ -56,7 +56,7 @@ TEST_F(DBRangeDelTest, EndSameAsStartCoversNothing) {
 }

 TEST_F(DBRangeDelTest, EndComesBeforeStartInvalidArgument) {
-  db_->Put(WriteOptions(), "b", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "b", "val"));
  ASSERT_TRUE(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "b", "a")
          .IsInvalidArgument());
@@ -82,13 +82,14 @@ TEST_F(DBRangeDelTest, CompactionOutputHasOnlyRangeTombstone) {

    // snapshot protects range tombstone from dropping due to becoming obsolete.
    const Snapshot* snapshot = db_->GetSnapshot();
-    db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z");
-    db_->Flush(FlushOptions());
+    ASSERT_OK(
+        db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));
+    ASSERT_OK(db_->Flush(FlushOptions()));

    ASSERT_EQ(1, NumTableFilesAtLevel(0));
    ASSERT_EQ(0, NumTableFilesAtLevel(1));
-    dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                                true /* disallow_trivial_move */);
+    ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                          true /* disallow_trivial_move */));
    ASSERT_EQ(0, NumTableFilesAtLevel(0));
    ASSERT_EQ(1, NumTableFilesAtLevel(1));
    ASSERT_EQ(0, TestGetTickerCount(opts, COMPACTION_RANGE_DEL_DROP_OBSOLETE));
@@ -118,7 +119,8 @@ TEST_F(DBRangeDelTest, CompactionOutputFilesExactlyFilled) {

  // snapshot protects range tombstone from dropping due to becoming obsolete.
  const Snapshot* snapshot = db_->GetSnapshot();
-  db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0), Key(1));
+  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0),
+                             Key(1)));

  Random rnd(301);
  for (int i = 0; i < kNumFiles; ++i) {
@@ -128,18 +130,18 @@ TEST_F(DBRangeDelTest, CompactionOutputFilesExactlyFilled) {
      values.push_back(rnd.RandomString(3 << 10));
      ASSERT_OK(Put(Key(i * kNumPerFile + j), values[j]));
      if (j == 0 && i > 0) {
-        dbfull()->TEST_WaitForFlushMemTable();
+        ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
      }
    }
  }
  // put extra key to trigger final flush
  ASSERT_OK(Put("", ""));
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(0));
  ASSERT_EQ(0, NumTableFilesAtLevel(1));

-  dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                        true /* disallow_trivial_move */));
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
  ASSERT_EQ(2, NumTableFilesAtLevel(1));
  db_->ReleaseSnapshot(snapshot);
@@ -178,12 +180,12 @@ TEST_F(DBRangeDelTest, MaxCompactionBytesCutsOutputFiles) {
    }
    // extra entry to trigger SpecialSkipListFactory's flush
    ASSERT_OK(Put(GetNumericStr(kNumPerFile), ""));
-    dbfull()->TEST_WaitForFlushMemTable();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
    ASSERT_EQ(i + 1, NumTableFilesAtLevel(0));
  }

-  dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                        true /* disallow_trivial_move */));
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
  ASSERT_GE(NumTableFilesAtLevel(1), 2);

@@ -221,10 +223,10 @@ TEST_F(DBRangeDelTest, SentinelsOmittedFromOutputFile) {
 }

 TEST_F(DBRangeDelTest, FlushRangeDelsSameStartKey) {
-  db_->Put(WriteOptions(), "b1", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "b1", "val"));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "c"));
-  db_->Put(WriteOptions(), "b2", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "b2", "val"));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "b"));
  // first iteration verifies query correctness in memtable, second verifies
@@ -241,8 +243,9 @@ TEST_F(DBRangeDelTest, FlushRangeDelsSameStartKey) {
 }

 TEST_F(DBRangeDelTest, CompactRangeDelsSameStartKey) {
-  db_->Put(WriteOptions(), "unused", "val");  // prevents empty after compaction
-  db_->Put(WriteOptions(), "b1", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "unused",
+                     "val"));  // prevents empty after compaction
+  ASSERT_OK(db_->Put(WriteOptions(), "b1", "val"));
  ASSERT_OK(db_->Flush(FlushOptions()));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "c"));
@@ -254,8 +257,8 @@ TEST_F(DBRangeDelTest, CompactRangeDelsSameStartKey) {

  for (int i = 0; i < 2; ++i) {
    if (i > 0) {
-      dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                                  true /* disallow_trivial_move */);
+      ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                            true /* disallow_trivial_move */));
      ASSERT_EQ(0, NumTableFilesAtLevel(0));
      ASSERT_EQ(1, NumTableFilesAtLevel(1));
    }
@@ -279,12 +282,13 @@ TEST_F(DBRangeDelTest, FlushRemovesCoveredKeys) {
    if (i == kNum / 3) {
      snapshot = db_->GetSnapshot();
    } else if (i == 2 * kNum / 3) {
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                       GetNumericStr(kRangeBegin), GetNumericStr(kRangeEnd));
+      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                 GetNumericStr(kRangeBegin),
+                                 GetNumericStr(kRangeEnd)));
    }
-    db_->Put(WriteOptions(), GetNumericStr(i), "val");
+    ASSERT_OK(db_->Put(WriteOptions(), GetNumericStr(i), "val"));
  }
-  db_->Flush(FlushOptions());
+  ASSERT_OK(db_->Flush(FlushOptions()));

  for (int i = 0; i < kNum; ++i) {
    ReadOptions read_opts;
@@ -314,24 +318,27 @@ TEST_F(DBRangeDelTest, CompactionRemovesCoveredKeys) {
  for (int i = 0; i < kNumFiles; ++i) {
    if (i > 0) {
      // range tombstone covers first half of the previous file
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                       GetNumericStr((i - 1) * kNumPerFile),
-                       GetNumericStr((i - 1) * kNumPerFile + kNumPerFile / 2));
+      ASSERT_OK(db_->DeleteRange(
+          WriteOptions(), db_->DefaultColumnFamily(),
+          GetNumericStr((i - 1) * kNumPerFile),
+          GetNumericStr((i - 1) * kNumPerFile + kNumPerFile / 2)));
    }
    // Make sure a given key appears in each file so compaction won't be able to
    // use trivial move, which would happen if the ranges were non-overlapping.
    // Also, we need an extra element since flush is only triggered when the
    // number of keys is one greater than SpecialSkipListFactory's limit.
    // We choose a key outside the key-range used by the test to avoid conflict.
-    db_->Put(WriteOptions(), GetNumericStr(kNumPerFile * kNumFiles), "val");
+    ASSERT_OK(db_->Put(WriteOptions(), GetNumericStr(kNumPerFile * kNumFiles),
+                       "val"));

    for (int j = 0; j < kNumPerFile; ++j) {
-      db_->Put(WriteOptions(), GetNumericStr(i * kNumPerFile + j), "val");
+      ASSERT_OK(
+          db_->Put(WriteOptions(), GetNumericStr(i * kNumPerFile + j), "val"));
    }
-    dbfull()->TEST_WaitForFlushMemTable();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
    ASSERT_EQ(i + 1, NumTableFilesAtLevel(0));
  }
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
  ASSERT_GT(NumTableFilesAtLevel(1), 0);
  ASSERT_EQ((kNumFiles - 1) * kNumPerFile / 2,
@@ -373,8 +380,8 @@ TEST_F(DBRangeDelTest, ValidLevelSubcompactionBoundaries) {
      if (i > 0) {
        // delete [95,105) in two files, [295,305) in next two
        int mid = (j + (1 - j % 2)) * kNumPerFile;
-        db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                         Key(mid - 5), Key(mid + 5));
+        ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                   Key(mid - 5), Key(mid + 5)));
      }
      std::vector<std::string> values;
      // Write 100KB (100 values, each 1K)
@@ -384,7 +391,7 @@ TEST_F(DBRangeDelTest, ValidLevelSubcompactionBoundaries) {
      }
      // put extra key to trigger flush
      ASSERT_OK(Put("", ""));
-      dbfull()->TEST_WaitForFlushMemTable();
+      ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
      if (j < kNumFiles - 1) {
        // background compaction may happen early for kNumFiles'th file
        ASSERT_EQ(NumTableFilesAtLevel(0), j + 1);
@@ -400,7 +407,7 @@ TEST_F(DBRangeDelTest, ValidLevelSubcompactionBoundaries) {
        // oversized L0 (relative to base_level) causes the compaction to run
        // earlier.
        ASSERT_OK(db_->EnableAutoCompaction({db_->DefaultColumnFamily()}));
-        dbfull()->TEST_WaitForCompact();
+        ASSERT_OK(dbfull()->TEST_WaitForCompact());
        ASSERT_OK(db_->SetOptions(db_->DefaultColumnFamily(),
                                  {{"disable_auto_compactions", "true"}}));
        ASSERT_EQ(NumTableFilesAtLevel(0), 0);
@@ -433,8 +440,8 @@ TEST_F(DBRangeDelTest, ValidUniversalSubcompactionBoundaries) {
        // insert range deletions [95,105) in two files, [295,305) in next two
        // to prepare L1 for later manual compaction.
        int mid = (j + (1 - j % 2)) * kNumPerFile;
-        db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                         Key(mid - 5), Key(mid + 5));
+        ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                   Key(mid - 5), Key(mid + 5)));
      }
      std::vector<std::string> values;
      // Write 100KB (100 values, each 1K)
@@ -444,13 +451,13 @@ TEST_F(DBRangeDelTest, ValidUniversalSubcompactionBoundaries) {
      }
      // put extra key to trigger flush
      ASSERT_OK(Put("", ""));
-      dbfull()->TEST_WaitForFlushMemTable();
+      ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
      if (j < kFilesPerLevel - 1) {
        // background compaction may happen early for kFilesPerLevel'th file
        ASSERT_EQ(NumTableFilesAtLevel(0), j + 1);
      }
    }
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
    ASSERT_EQ(NumTableFilesAtLevel(0), 0);
    ASSERT_GT(NumTableFilesAtLevel(kNumLevels - 1 - i), kFilesPerLevel - 1);
  }
@@ -483,17 +490,17 @@ TEST_F(DBRangeDelTest, CompactionRemovesCoveredMergeOperands) {
  for (int i = 0; i <= kNumFiles * kNumPerFile; ++i) {
    if (i % kNumPerFile == 0 && i / kNumPerFile == kNumFiles - 1) {
      // Delete merge operands from all but the last file
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "key",
-                       "key_");
+      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                 "key", "key_"));
    }
    std::string val;
    PutFixed64(&val, i);
-    db_->Merge(WriteOptions(), "key", val);
+    ASSERT_OK(db_->Merge(WriteOptions(), "key", val));
    // we need to prevent trivial move using Puts so compaction will actually
    // process the merge operands.
-    db_->Put(WriteOptions(), "prevent_trivial_move", "");
+    ASSERT_OK(db_->Put(WriteOptions(), "prevent_trivial_move", ""));
    if (i > 0 && i % kNumPerFile == 0) {
-      dbfull()->TEST_WaitForFlushMemTable();
+      ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
    }
  }

@@ -504,7 +511,7 @@ TEST_F(DBRangeDelTest, CompactionRemovesCoveredMergeOperands) {
  PutFixed64(&expected, 45);  // 1+2+...+9
  ASSERT_EQ(expected, actual);

-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  expected.clear();
  ASSERT_OK(db_->Get(read_opts, "key", &actual));
@@ -550,19 +557,19 @@ TEST_F(DBRangeDelTest, ObsoleteTombstoneCleanup) {
  opts.statistics = CreateDBStatistics();
  Reopen(opts);

-  db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr1",
-                   "dr10");  // obsolete after compaction
-  db_->Put(WriteOptions(), "key", "val");
-  db_->Flush(FlushOptions());
+  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr1",
+                             "dr10"));  // obsolete after compaction
+  ASSERT_OK(db_->Put(WriteOptions(), "key", "val"));
+  ASSERT_OK(db_->Flush(FlushOptions()));
  const Snapshot* snapshot = db_->GetSnapshot();
-  db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr2",
-                   "dr20");  // protected by snapshot
-  db_->Put(WriteOptions(), "key", "val");
-  db_->Flush(FlushOptions());
+  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr2",
+                             "dr20"));  // protected by snapshot
+  ASSERT_OK(db_->Put(WriteOptions(), "key", "val"));
+  ASSERT_OK(db_->Flush(FlushOptions()));

  ASSERT_EQ(2, NumTableFilesAtLevel(0));
  ASSERT_EQ(0, NumTableFilesAtLevel(1));
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
  ASSERT_EQ(1, NumTableFilesAtLevel(1));
  ASSERT_EQ(1, TestGetTickerCount(opts, COMPACTION_RANGE_DEL_DROP_OBSOLETE));
@@ -609,22 +616,24 @@ TEST_F(DBRangeDelTest, TableEvictedDuringScan) {
  // to bottommost level (i.e., L1).
  const Snapshot* snapshot = db_->GetSnapshot();
  for (int i = 0; i < kNum; ++i) {
-    db_->Put(WriteOptions(), GetNumericStr(i), "val");
+    ASSERT_OK(db_->Put(WriteOptions(), GetNumericStr(i), "val"));
    if (i > 0) {
-      dbfull()->TEST_WaitForFlushMemTable();
+      ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
    }
    if (i >= kNum / 2 && i < kNum / 2 + kNumRanges) {
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                       GetNumericStr(kRangeBegin), GetNumericStr(kRangeEnd));
+      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                 GetNumericStr(kRangeBegin),
+                                 GetNumericStr(kRangeEnd)));
    }
  }
  // Must be > 1 so the first L1 file can be closed before scan finishes
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_GT(NumTableFilesAtLevel(1), 1);
  std::vector<uint64_t> file_numbers = ListTableFiles(env_, dbname_);

  ReadOptions read_opts;
  auto* iter = db_->NewIterator(read_opts);
+  ASSERT_OK(iter->status());
  int expected = kRangeEnd;
  iter->SeekToFirst();
  for (auto file_number : file_numbers) {
@@ -647,7 +656,7 @@ TEST_F(DBRangeDelTest, TableEvictedDuringScan) {
 TEST_F(DBRangeDelTest, GetCoveredKeyFromMutableMemtable) {
  do {
    DestroyAndReopen(CurrentOptions());
-    db_->Put(WriteOptions(), "key", "val");
+    ASSERT_OK(db_->Put(WriteOptions(), "key", "val"));
    ASSERT_OK(
        db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));

@@ -669,10 +678,10 @@ TEST_F(DBRangeDelTest, GetCoveredKeyFromImmutableMemtable) {
    opts.memtable_factory.reset(new SpecialSkipListFactory(1));
    DestroyAndReopen(opts);

-    db_->Put(WriteOptions(), "key", "val");
+    ASSERT_OK(db_->Put(WriteOptions(), "key", "val"));
    ASSERT_OK(
        db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));
-    db_->Put(WriteOptions(), "blah", "val");
+    ASSERT_OK(db_->Put(WriteOptions(), "blah", "val"));

    ReadOptions read_opts;
    std::string value;
@@ -683,7 +692,7 @@ TEST_F(DBRangeDelTest, GetCoveredKeyFromImmutableMemtable) {
 TEST_F(DBRangeDelTest, GetCoveredKeyFromSst) {
  do {
    DestroyAndReopen(CurrentOptions());
-    db_->Put(WriteOptions(), "key", "val");
+    ASSERT_OK(db_->Put(WriteOptions(), "key", "val"));
    // snapshot prevents key from being deleted during flush
    const Snapshot* snapshot = db_->GetSnapshot();
    ASSERT_OK(
@@ -706,11 +715,11 @@ TEST_F(DBRangeDelTest, GetCoveredMergeOperandFromMemtable) {
  for (int i = 0; i < kNumMergeOps; ++i) {
    std::string val;
    PutFixed64(&val, i);
-    db_->Merge(WriteOptions(), "key", val);
+    ASSERT_OK(db_->Merge(WriteOptions(), "key", val));
    if (i == kNumMergeOps / 2) {
      // deletes [0, 5]
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "key",
-                       "key_");
+      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                 "key", "key_"));
    }
  }

@@ -734,16 +743,16 @@ TEST_F(DBRangeDelTest, GetIgnoresRangeDeletions) {
  opts.memtable_factory.reset(new SpecialSkipListFactory(1));
  Reopen(opts);

-  db_->Put(WriteOptions(), "sst_key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "sst_key", "val"));
  // snapshot prevents key from being deleted during flush
  const Snapshot* snapshot = db_->GetSnapshot();
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));
  ASSERT_OK(db_->Flush(FlushOptions()));
-  db_->Put(WriteOptions(), "imm_key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "imm_key", "val"));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));
-  db_->Put(WriteOptions(), "mem_key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "mem_key", "val"));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));

@@ -768,13 +777,15 @@ TEST_F(DBRangeDelTest, IteratorRemovesCoveredKeys) {
  // should be deleted.
  for (int i = 0; i < kNum; ++i) {
    if (i == kNum / 2) {
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                       GetNumericStr(kRangeBegin), GetNumericStr(kRangeEnd));
+      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                 GetNumericStr(kRangeBegin),
+                                 GetNumericStr(kRangeEnd)));
    }
-    db_->Put(WriteOptions(), GetNumericStr(i), "val");
+    ASSERT_OK(db_->Put(WriteOptions(), GetNumericStr(i), "val"));
  }
  ReadOptions read_opts;
  auto* iter = db_->NewIterator(read_opts);
+  ASSERT_OK(iter->status());

  int expected = 0;
  for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
@@ -802,14 +813,16 @@ TEST_F(DBRangeDelTest, IteratorOverUserSnapshot) {
  for (int i = 0; i < kNum; ++i) {
    if (i == kNum / 2) {
      snapshot = db_->GetSnapshot();
-      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                       GetNumericStr(kRangeBegin), GetNumericStr(kRangeEnd));
+      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                                 GetNumericStr(kRangeBegin),
+                                 GetNumericStr(kRangeEnd)));
    }
-    db_->Put(WriteOptions(), GetNumericStr(i), "val");
+    ASSERT_OK(db_->Put(WriteOptions(), GetNumericStr(i), "val"));
  }
  ReadOptions read_opts;
  read_opts.snapshot = snapshot;
  auto* iter = db_->NewIterator(read_opts);
+  ASSERT_OK(iter->status());

  int expected = 0;
  for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
@@ -828,22 +841,23 @@ TEST_F(DBRangeDelTest, IteratorIgnoresRangeDeletions) {
  opts.memtable_factory.reset(new SpecialSkipListFactory(1));
  Reopen(opts);

-  db_->Put(WriteOptions(), "sst_key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "sst_key", "val"));
  // snapshot prevents key from being deleted during flush
  const Snapshot* snapshot = db_->GetSnapshot();
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));
  ASSERT_OK(db_->Flush(FlushOptions()));
-  db_->Put(WriteOptions(), "imm_key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "imm_key", "val"));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));
-  db_->Put(WriteOptions(), "mem_key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "mem_key", "val"));
  ASSERT_OK(
      db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z"));

  ReadOptions read_opts;
  read_opts.ignore_range_deletions = true;
  auto* iter = db_->NewIterator(read_opts);
+  ASSERT_OK(iter->status());
  int i = 0;
  std::string expected[] = {"imm_key", "mem_key", "sst_key"};
  for (iter->SeekToFirst(); iter->Valid(); iter->Next(), ++i) {
@@ -857,7 +871,7 @@ TEST_F(DBRangeDelTest, IteratorIgnoresRangeDeletions) {

 #ifndef ROCKSDB_UBSAN_RUN
 TEST_F(DBRangeDelTest, TailingIteratorRangeTombstoneUnsupported) {
-  db_->Put(WriteOptions(), "key", "val");
+  ASSERT_OK(db_->Put(WriteOptions(), "key", "val"));
  // snapshot prevents key from being deleted during flush
  const Snapshot* snapshot = db_->GetSnapshot();
  ASSERT_OK(
@@ -873,6 +887,7 @@ TEST_F(DBRangeDelTest, TailingIteratorRangeTombstoneUnsupported) {
      iter->SeekToFirst();
    }
    ASSERT_TRUE(iter->status().IsNotSupported());
+
    delete iter;
    if (i == 0) {
      ASSERT_OK(db_->Flush(FlushOptions()));
@@ -882,7 +897,6 @@ TEST_F(DBRangeDelTest, TailingIteratorRangeTombstoneUnsupported) {
  }
  db_->ReleaseSnapshot(snapshot);
 }
-
 #endif  // !ROCKSDB_UBSAN_RUN

 TEST_F(DBRangeDelTest, SubcompactionHasEmptyDedicatedRangeDelFile) {
@@ -926,8 +940,8 @@ TEST_F(DBRangeDelTest, SubcompactionHasEmptyDedicatedRangeDelFile) {
  ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(0));
  ASSERT_EQ(1, NumTableFilesAtLevel(1));

-  db_->EnableAutoCompaction({db_->DefaultColumnFamily()});
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(db_->EnableAutoCompaction({db_->DefaultColumnFamily()}));
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  db_->ReleaseSnapshot(snapshot);
 }

@@ -949,7 +963,7 @@ TEST_F(DBRangeDelTest, MemtableBloomFilter) {
  for (int i = 0; i < kNumKeys; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0),
                             Key(kNumKeys)));
  for (int i = 0; i < kNumKeys; ++i) {
@@ -987,8 +1001,8 @@ TEST_F(DBRangeDelTest, CompactionTreatsSplitInputLevelDeletionAtomically) {

    // snapshot protects range tombstone from dropping due to becoming obsolete.
    const Snapshot* snapshot = db_->GetSnapshot();
-    db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0),
-                     Key(2 * kNumFilesPerLevel));
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(0), Key(2 * kNumFilesPerLevel)));

    Random rnd(301);
    std::string value = rnd.RandomString(kValueBytes);
@@ -997,14 +1011,14 @@ TEST_F(DBRangeDelTest, CompactionTreatsSplitInputLevelDeletionAtomically) {
      ASSERT_OK(Put(Key(j), value));
      ASSERT_OK(Put(Key(2 * kNumFilesPerLevel - 1 - j), value));
      if (j > 0) {
-        dbfull()->TEST_WaitForFlushMemTable();
+        ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
        ASSERT_EQ(j, NumTableFilesAtLevel(0));
      }
    }
    // put extra key to trigger final flush
    ASSERT_OK(Put("", ""));
-    dbfull()->TEST_WaitForFlushMemTable();
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
    ASSERT_EQ(0, NumTableFilesAtLevel(0));
    ASSERT_EQ(kNumFilesPerLevel, NumTableFilesAtLevel(1));

@@ -1022,7 +1036,7 @@ TEST_F(DBRangeDelTest, CompactionTreatsSplitInputLevelDeletionAtomically) {
    } else if (i == 2) {
      ASSERT_OK(db_->SetOptions(db_->DefaultColumnFamily(),
                                {{"max_bytes_for_level_base", "10000"}}));
-      dbfull()->TEST_WaitForCompact();
+      ASSERT_OK(dbfull()->TEST_WaitForCompact());
      ASSERT_EQ(1, NumTableFilesAtLevel(1));
    }
    ASSERT_GT(NumTableFilesAtLevel(2), 0);
@@ -1056,8 +1070,8 @@ TEST_F(DBRangeDelTest, RangeTombstoneEndKeyAsSstableUpperBound) {
  // A snapshot protects the range tombstone from dropping due to
  // becoming obsolete.
  const Snapshot* snapshot = db_->GetSnapshot();
-  db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                   Key(0), Key(2 * kNumFilesPerLevel));
+  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0),
+                             Key(2 * kNumFilesPerLevel)));

  // Create 2 additional sstables in L0. Note that the first sstable
  // contains the range tombstone.
@@ -1096,7 +1110,7 @@ TEST_F(DBRangeDelTest, RangeTombstoneEndKeyAsSstableUpperBound) {
    ASSERT_EQ(value, Get(Key(2)));
    auto begin_str = Key(3);
    const ROCKSDB_NAMESPACE::Slice begin = begin_str;
-    dbfull()->TEST_CompactRange(1, &begin, nullptr);
+    ASSERT_OK(dbfull()->TEST_CompactRange(1, &begin, nullptr));
    ASSERT_EQ(1, NumTableFilesAtLevel(1));
    ASSERT_EQ(2, NumTableFilesAtLevel(2));
    ASSERT_EQ(value, Get(Key(2)));
@@ -1115,7 +1129,7 @@ TEST_F(DBRangeDelTest, RangeTombstoneEndKeyAsSstableUpperBound) {
    //     [key000002#6,1, key000004#72057594037927935,15]
    auto begin_str = Key(0);
    const ROCKSDB_NAMESPACE::Slice begin = begin_str;
-    dbfull()->TEST_CompactRange(1, &begin, &begin);
+    ASSERT_OK(dbfull()->TEST_CompactRange(1, &begin, &begin));
    ASSERT_EQ(0, NumTableFilesAtLevel(1));
    ASSERT_EQ(3, NumTableFilesAtLevel(2));
  }
@@ -1216,9 +1230,9 @@ TEST_F(DBRangeDelTest, KeyAtOverlappingEndpointReappears) {
  std::string value;
  ASSERT_TRUE(db_->Get(ReadOptions(), "key", &value).IsNotFound());

-  dbfull()->TEST_CompactRange(0 /* level */, nullptr /* begin */,
-                              nullptr /* end */, nullptr /* column_family */,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(
+      0 /* level */, nullptr /* begin */, nullptr /* end */,
+      nullptr /* column_family */, true /* disallow_trivial_move */));
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
  // Now we have multiple files at L1 all containing a single user key, thus
  // guaranteeing overlap in the file endpoints.
@@ -1229,9 +1243,9 @@ TEST_F(DBRangeDelTest, KeyAtOverlappingEndpointReappears) {

  // Compact and verify again. It's worthwhile because now the files have
  // tighter endpoints, so we can verify that doesn't mess anything up.
-  dbfull()->TEST_CompactRange(1 /* level */, nullptr /* begin */,
-                              nullptr /* end */, nullptr /* column_family */,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(
+      1 /* level */, nullptr /* begin */, nullptr /* end */,
+      nullptr /* column_family */, true /* disallow_trivial_move */));
  ASSERT_GT(NumTableFilesAtLevel(2), 1);
  ASSERT_TRUE(db_->Get(ReadOptions(), "key", &value).IsNotFound());

@@ -1307,6 +1321,7 @@ TEST_F(DBRangeDelTest, UntruncatedTombstoneDoesNotDeleteNewerKey) {

  auto get_key_count = [this]() -> int {
    auto* iter = db_->NewIterator(ReadOptions());
+    assert(iter->status().ok());
    iter->SeekToFirst();
    int keys_found = 0;
    for (; iter->Valid(); iter->Next()) {
@@ -1409,6 +1424,7 @@ TEST_F(DBRangeDelTest, DeletedMergeOperandReappearsIterPrev) {
  ASSERT_GT(NumTableFilesAtLevel(1), 1);

  auto* iter = db_->NewIterator(ReadOptions());
+  ASSERT_OK(iter->status());
  iter->SeekToLast();
  int keys_found = 0;
  for (; iter->Valid(); iter->Prev()) {
@@ -1435,11 +1451,12 @@ TEST_F(DBRangeDelTest, SnapshotPreventsDroppedKeys) {
  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0),
                             Key(10)));

-  db_->Flush(FlushOptions());
+  ASSERT_OK(db_->Flush(FlushOptions()));

  ReadOptions read_opts;
  read_opts.snapshot = snapshot;
  auto* iter = db_->NewIterator(read_opts);
+  ASSERT_OK(iter->status());

  iter->SeekToFirst();
  ASSERT_TRUE(iter->Valid());
@@ -1482,6 +1499,7 @@ TEST_F(DBRangeDelTest, SnapshotPreventsDroppedKeysInImmMemTables) {
  ReadOptions read_opts;
  read_opts.snapshot = snapshot.get();
  std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts));
+  ASSERT_OK(iter->status());

  TEST_SYNC_POINT("SnapshotPreventsDroppedKeysInImmMemTables:AfterNewIterator");

@@ -1519,7 +1537,7 @@ TEST_F(DBRangeDelTest, RangeTombstoneWrittenToMinimalSsts) {
      std::string value = rnd.RandomString(kValueBytes);
      ASSERT_OK(Put(key, value));
    }
-    db_->Flush(FlushOptions());
+    ASSERT_OK(db_->Flush(FlushOptions()));
    MoveFilesToLevel(2);
  }
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
@@ -1538,7 +1556,7 @@ TEST_F(DBRangeDelTest, RangeTombstoneWrittenToMinimalSsts) {
  // TODO(ajkr): remove this `Put` after file cutting accounts for range
  // tombstones (#3977).
  ASSERT_OK(Put("c" + Key(1), "value"));
-  db_->Flush(FlushOptions());
+  ASSERT_OK(db_->Flush(FlushOptions()));

  // Ensure manual L0->L1 compaction cuts the outputs before the range tombstone
  // and the range tombstone is only placed in the second SST.
@@ -1546,9 +1564,9 @@ TEST_F(DBRangeDelTest, RangeTombstoneWrittenToMinimalSsts) {
  Slice begin_key(begin_key_storage);
  std::string end_key_storage("d");
  Slice end_key(end_key_storage);
-  dbfull()->TEST_CompactRange(0 /* level */, &begin_key /* begin */,
-                              &end_key /* end */, nullptr /* column_family */,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(
+      0 /* level */, &begin_key /* begin */, &end_key /* end */,
+      nullptr /* column_family */, true /* disallow_trivial_move */));
  ASSERT_EQ(2, NumTableFilesAtLevel(1));

  std::vector<LiveFileMetaData> all_metadata;
@@ -1613,15 +1631,15 @@ TEST_F(DBRangeDelTest, OverlappedTombstones) {

  ASSERT_EQ(1, NumTableFilesAtLevel(0));

-  dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                        true /* disallow_trivial_move */));

  // The tombstone range is not broken up into multiple SSTs which may incur a
  // large compaction with L2.
  ASSERT_EQ(1, NumTableFilesAtLevel(1));
  std::vector<std::vector<FileMetaData>> files;
-  dbfull()->TEST_CompactRange(1, nullptr, nullptr, nullptr,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(1, nullptr, nullptr, nullptr,
+                                        true /* disallow_trivial_move */));
  ASSERT_EQ(1, NumTableFilesAtLevel(2));
  ASSERT_EQ(0, NumTableFilesAtLevel(1));
 }
@@ -1654,13 +1672,13 @@ TEST_F(DBRangeDelTest, OverlappedKeys) {

  // The key range is broken up into three SSTs to avoid a future big compaction
  // with the grandparent
-  dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
+                                        true /* disallow_trivial_move */));
  ASSERT_EQ(3, NumTableFilesAtLevel(1));

  std::vector<std::vector<FileMetaData>> files;
-  dbfull()->TEST_CompactRange(1, nullptr, nullptr, nullptr,
-                              true /* disallow_trivial_move */);
+  ASSERT_OK(dbfull()->TEST_CompactRange(1, nullptr, nullptr, nullptr,
+                                        true /* disallow_trivial_move */));
  ASSERT_EQ(1, NumTableFilesAtLevel(2));
  ASSERT_EQ(0, NumTableFilesAtLevel(1));
 }
@@ -170,7 +170,7 @@ TEST_F(DBSSTTest, DontDeleteMovedFile) {
    ASSERT_OK(Flush());
  }
  // this should execute both L0->L1 and L1->(move)->L2 compactions
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ("0,0,1", FilesPerLevel(0));

  // If the moved file is actually deleted (the move-safeguard in
@@ -218,7 +218,7 @@ TEST_F(DBSSTTest, DeleteObsoleteFilesPendingOutputs) {
    ASSERT_OK(Flush());
  }
  // this should execute both L0->L1 and L1->(move)->L2 compactions
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ("0,0,1", FilesPerLevel(0));

  test::SleepingBackgroundTask blocking_thread;
@@ -264,9 +264,9 @@ TEST_F(DBSSTTest, DeleteObsoleteFilesPendingOutputs) {
  // finish the flush!
  blocking_thread.WakeUp();
  blocking_thread.WaitUntilDone();
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  // File just flushed is too big for L0 and L1 so gets moved to L2.
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ("0,0,1,0,1", FilesPerLevel(0));

  metadata.clear();
@@ -302,8 +302,8 @@ TEST_F(DBSSTTest, DBWithSstFileManager) {
  for (int i = 0; i < 25; i++) {
    GenerateNewRandomFile(&rnd);
    ASSERT_OK(Flush());
-    dbfull()->TEST_WaitForFlushMemTable();
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
    // Verify that we are tracking all sst files in dbname_
    std::unordered_map<std::string, uint64_t> files_in_db;
    ASSERT_OK(GetAllSSTFiles(&files_in_db));
@@ -544,6 +544,11 @@ TEST_P(DBWALTestWithParam, WALTrashCleanupOnOpen) {

  // Create 4 files in L0
  for (char v = 'a'; v <= 'd'; v++) {
+    if (v == 'd') {
+      // Maximize the change that the last log file will be preserved in trash
+      // before restarting the DB.
+      options.sst_file_manager->SetDeleteRateBytesPerSecond(1);
+    }
    ASSERT_OK(Put("Key2", DummyString(1024, v)));
    ASSERT_OK(Put("Key3", DummyString(1024, v)));
    ASSERT_OK(Put("Key4", DummyString(1024, v)));
@@ -603,7 +608,7 @@ TEST_F(DBSSTTest, OpenDBWithExistingTrash) {
  Destroy(last_options_);

  // Add some trash files to the db directory so the DB can clean them up
-  env_->CreateDirIfMissing(dbname_);
+  ASSERT_OK(env_->CreateDirIfMissing(dbname_));
  ASSERT_OK(WriteStringToFile(env_, "abc", dbname_ + "/" + "001.sst.trash"));
  ASSERT_OK(WriteStringToFile(env_, "abc", dbname_ + "/" + "002.sst.trash"));
  ASSERT_OK(WriteStringToFile(env_, "abc", dbname_ + "/" + "003.sst.trash"));
@@ -728,7 +733,7 @@ TEST_F(DBSSTTest, DestroyDBWithRateLimitedDelete) {
  int num_sst_files = 0;
  int num_wal_files = 0;
  std::vector<std::string> db_files;
-  env_->GetChildren(dbname_, &db_files);
+  ASSERT_OK(env_->GetChildren(dbname_, &db_files));
  for (std::string f : db_files) {
    if (f.substr(f.find_last_of(".") + 1) == "sst") {
      num_sst_files++;
@@ -817,7 +822,7 @@ TEST_F(DBSSTTest, CancellingCompactionsWorks) {
    ASSERT_OK(Put(Key(i), rnd.RandomString(50)));
  }
  ASSERT_OK(Flush());
-  dbfull()->TEST_WaitForCompact(true);
+  ASSERT_OK(dbfull()->TEST_WaitForCompact(true));

  // Because we set a callback in CancelledCompaction, we actually
  // let the compaction run
@@ -137,7 +137,7 @@ TEST_F(DBStatisticsTest, ResetStats) {
      ASSERT_EQ(1, TestGetTickerCount(options, NUMBER_KEYS_WRITTEN));
      options.statistics->histogramData(DB_WRITE, &histogram_data);
      ASSERT_GT(histogram_data.max, 0.0);
-      options.statistics->Reset();
+      ASSERT_OK(options.statistics->Reset());
    }
  }
 }
@@ -65,9 +65,9 @@ TEST_F(DBTablePropertiesTest, GetPropertiesOfAllTablesTest) {
  // Create 4 tables
  for (int table = 0; table < 4; ++table) {
    for (int i = 0; i < 10 + table; ++i) {
-      db_->Put(WriteOptions(), ToString(table * 100 + i), "val");
+      ASSERT_OK(db_->Put(WriteOptions(), ToString(table * 100 + i), "val"));
    }
-    db_->Flush(FlushOptions());
+    ASSERT_OK(db_->Flush(FlushOptions()));
  }

  // 1. Read table properties directly from file
@@ -161,14 +161,14 @@ TEST_F(DBTablePropertiesTest, GetPropertiesOfTablesInRange) {
  for (int i = 0; i < 10000; i++) {
    ASSERT_OK(Put(test::RandomKey(&rnd, 5), rnd.RandomString(102)));
  }
-  Flush();
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(Flush());
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  if (NumTableFilesAtLevel(0) == 0) {
    ASSERT_OK(Put(test::RandomKey(&rnd, 5), rnd.RandomString(102)));
-    Flush();
+    ASSERT_OK(Flush());
  }

-  db_->PauseBackgroundWork();
+  ASSERT_OK(db_->PauseBackgroundWork());

  // Ensure that we have at least L0, L1 and L2
  ASSERT_GT(NumTableFilesAtLevel(0), 0);
@@ -236,8 +236,8 @@ TEST_F(DBTablePropertiesTest, GetColumnFamilyNameProperty) {
  // Create one table per CF, then verify it was created with the column family
  // name property.
  for (uint32_t cf = 0; cf < 2; ++cf) {
-    Put(cf, "key", "val");
-    Flush(cf);
+    ASSERT_OK(Put(cf, "key", "val"));
+    ASSERT_OK(Flush(cf));

    TablePropertiesCollection fname_to_props;
    ASSERT_OK(db_->GetPropertiesOfAllTables(handles_[cf], &fname_to_props));
@@ -260,17 +260,17 @@ TEST_F(DBTablePropertiesTest, GetDbIdentifiersProperty) {
  CreateAndReopenWithCF({"goku"}, CurrentOptions());

  for (uint32_t cf = 0; cf < 2; ++cf) {
-    Put(cf, "key", "val");
-    Put(cf, "foo", "bar");
-    Flush(cf);
+    ASSERT_OK(Put(cf, "key", "val"));
+    ASSERT_OK(Put(cf, "foo", "bar"));
+    ASSERT_OK(Flush(cf));

    TablePropertiesCollection fname_to_props;
    ASSERT_OK(db_->GetPropertiesOfAllTables(handles_[cf], &fname_to_props));
    ASSERT_EQ(1U, fname_to_props.size());

    std::string id, sid;
-    db_->GetDbIdentity(id);
-    db_->GetDbSessionId(sid);
+    ASSERT_OK(db_->GetDbIdentity(id));
+    ASSERT_OK(db_->GetDbSessionId(sid));
    ASSERT_EQ(id, fname_to_props.begin()->second->db_id);
    ASSERT_EQ(sid, fname_to_props.begin()->second->db_session_id);
  }
@@ -298,9 +298,9 @@ TEST_P(DBTableHostnamePropertyTest, DbHostLocationProperty) {
  CreateAndReopenWithCF({"goku"}, opts);

  for (uint32_t cf = 0; cf < 2; ++cf) {
-    Put(cf, "key", "val");
-    Put(cf, "foo", "bar");
-    Flush(cf);
+    ASSERT_OK(Put(cf, "key", "val"));
+    ASSERT_OK(Put(cf, "foo", "bar"));
+    ASSERT_OK(Flush(cf));

    TablePropertiesCollection fname_to_props;
    ASSERT_OK(db_->GetPropertiesOfAllTables(handles_[cf], &fname_to_props));
@@ -356,8 +356,8 @@ TEST_P(DBTablePropertiesTest, DeletionTriggeredCompactionMarking) {

  // add an L1 file to prevent tombstones from dropping due to obsolescence
  // during flush
-  Put(Key(0), "val");
-  Flush();
+  ASSERT_OK(Put(Key(0), "val"));
+  ASSERT_OK(Flush());
  MoveFilesToLevel(1);

  DeletionTriggeredCompactionTestListener *listener =
@@ -368,14 +368,14 @@ TEST_P(DBTablePropertiesTest, DeletionTriggeredCompactionMarking) {
  for (int i = 0; i < kNumKeys; ++i) {
    if (i >= kNumKeys - kWindowSize &&
        i < kNumKeys - kWindowSize + kNumDelsTrigger) {
-      Delete(Key(i));
+      ASSERT_OK(Delete(Key(i)));
    } else {
-      Put(Key(i), "val");
+      ASSERT_OK(Put(Key(i), "val"));
    }
  }
-  Flush();
+  ASSERT_OK(Flush());

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));

  // Change the window size and deletion trigger and ensure new values take
@@ -389,14 +389,14 @@ TEST_P(DBTablePropertiesTest, DeletionTriggeredCompactionMarking) {
  for (int i = 0; i < kNumKeys; ++i) {
    if (i >= kNumKeys - kWindowSize &&
        i < kNumKeys - kWindowSize + kNumDelsTrigger) {
-      Delete(Key(i));
+      ASSERT_OK(Delete(Key(i)));
    } else {
-      Put(Key(i), "val");
+      ASSERT_OK(Put(Key(i), "val"));
    }
  }
-  Flush();
+  ASSERT_OK(Flush());

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));

  // Change the window size to disable delete triggered compaction
@@ -408,14 +408,14 @@ TEST_P(DBTablePropertiesTest, DeletionTriggeredCompactionMarking) {
  for (int i = 0; i < kNumKeys; ++i) {
    if (i >= kNumKeys - kWindowSize &&
        i < kNumKeys - kWindowSize + kNumDelsTrigger) {
-      Delete(Key(i));
+      ASSERT_OK(Delete(Key(i)));
    } else {
-      Put(Key(i), "val");
+      ASSERT_OK(Put(Key(i), "val"));
    }
  }
-  Flush();
+  ASSERT_OK(Flush());

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(1, NumTableFilesAtLevel(0));
  ASSERT_LT(0, opts.statistics->getTickerCount(COMPACT_WRITE_BYTES_MARKED));
  ASSERT_LT(0, opts.statistics->getTickerCount(COMPACT_READ_BYTES_MARKED));
@@ -438,8 +438,8 @@ TEST_P(DBTablePropertiesTest, RatioBasedDeletionTriggeredCompactionMarking) {

  // Add an L2 file to prevent tombstones from dropping due to obsolescence
  // during flush
-  Put(Key(0), "val");
-  Flush();
+  ASSERT_OK(Put(Key(0), "val"));
+  ASSERT_OK(Flush());
  MoveFilesToLevel(2);

  auto* listener = new DeletionTriggeredCompactionTestListener();
@@ -179,7 +179,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {

    if (i % 100 == 99) {
      ASSERT_OK(Flush(1));
-      dbfull()->TEST_WaitForCompact();
+      ASSERT_OK(dbfull()->TEST_WaitForCompact());
      if (i == 299) {
        file_iters_deleted = true;
      }
@@ -411,7 +411,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorUpperBound) {
  it->Next();
  // Not valid since "21" is over the upper bound.
  ASSERT_FALSE(it->Valid());
-
+  ASSERT_OK(it->status());
  // This keeps track of the number of times NeedToSeekImmutable() was true.
  int immutable_seeks = 0;
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
@@ -424,6 +424,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorUpperBound) {
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();

  ASSERT_FALSE(it->Valid());
+  ASSERT_OK(it->status());
  ASSERT_EQ(0, immutable_seeks);
 }

@@ -1305,51 +1305,6 @@ TEST_F(DBTest, DISABLED_RepeatedWritesToSameKey) {
 }
 #endif  // ROCKSDB_LITE

-TEST_F(DBTest, SparseMerge) {
-  do {
-    Options options = CurrentOptions();
-    options.compression = kNoCompression;
-    CreateAndReopenWithCF({"pikachu"}, options);
-
-    FillLevels("A", "Z", 1);
-
-    // Suppose there is:
-    //    small amount of data with prefix A
-    //    large amount of data with prefix B
-    //    small amount of data with prefix C
-    // and that recent updates have made small changes to all three prefixes.
-    // Check that we do not do a compaction that merges all of B in one shot.
-    const std::string value(1000, 'x');
-    Put(1, "A", "va");
-    // Write approximately 100MB of "B" values
-    for (int i = 0; i < 100000; i++) {
-      char key[100];
-      snprintf(key, sizeof(key), "B%010d", i);
-      Put(1, key, value);
-    }
-    Put(1, "C", "vc");
-    ASSERT_OK(Flush(1));
-    dbfull()->TEST_CompactRange(0, nullptr, nullptr, handles_[1]);
-
-    // Make sparse update
-    Put(1, "A", "va2");
-    Put(1, "B100", "bvalue2");
-    Put(1, "C", "vc2");
-    ASSERT_OK(Flush(1));
-
-    // Compactions should not cause us to create a situation where
-    // a file overlaps too much data at the next level.
-    ASSERT_LE(dbfull()->TEST_MaxNextLevelOverlappingBytes(handles_[1]),
-              20 * 1048576);
-    dbfull()->TEST_CompactRange(0, nullptr, nullptr);
-    ASSERT_LE(dbfull()->TEST_MaxNextLevelOverlappingBytes(handles_[1]),
-              20 * 1048576);
-    dbfull()->TEST_CompactRange(1, nullptr, nullptr);
-    ASSERT_LE(dbfull()->TEST_MaxNextLevelOverlappingBytes(handles_[1]),
-              20 * 1048576);
-  } while (ChangeCompactOptions());
-}
-
 #ifndef ROCKSDB_LITE
 static bool Between(uint64_t val, uint64_t low, uint64_t high) {
  bool result = (val >= low) && (val <= high);
@@ -174,7 +174,9 @@ TEST_F(DBTest2, PartitionedIndexUserToInternalKey) {
    ASSERT_OK(Put("keykey_" + std::to_string(j), value));
    snapshots.push_back(db_->GetSnapshot());
  }
-  Flush();
+  ASSERT_OK(Flush());
+  dbfull()->TEST_WaitForFlushMemTable();
+
  for (auto s : snapshots) {
    db_->ReleaseSnapshot(s);
  }
@@ -200,28 +200,28 @@ bool DBTestBase::ChangeCompactOptions() {
    Destroy(last_options_);
    auto options = CurrentOptions();
    options.create_if_missing = true;
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else if (option_config_ == kUniversalCompaction) {
    option_config_ = kUniversalCompactionMultiLevel;
    Destroy(last_options_);
    auto options = CurrentOptions();
    options.create_if_missing = true;
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else if (option_config_ == kUniversalCompactionMultiLevel) {
    option_config_ = kLevelSubcompactions;
    Destroy(last_options_);
    auto options = CurrentOptions();
    assert(options.max_subcompactions > 1);
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else if (option_config_ == kLevelSubcompactions) {
    option_config_ = kUniversalSubcompactions;
    Destroy(last_options_);
    auto options = CurrentOptions();
    assert(options.max_subcompactions > 1);
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else {
    return false;
@@ -236,7 +236,7 @@ bool DBTestBase::ChangeWalOptions() {
    auto options = CurrentOptions();
    Destroy(options);
    options.create_if_missing = true;
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else if (option_config_ == kDBLogDir) {
    option_config_ = kWalDirAndMmapReads;
@@ -244,14 +244,14 @@ bool DBTestBase::ChangeWalOptions() {
    auto options = CurrentOptions();
    Destroy(options);
    options.create_if_missing = true;
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else if (option_config_ == kWalDirAndMmapReads) {
    option_config_ = kRecycleLogFiles;
    Destroy(last_options_);
    auto options = CurrentOptions();
    Destroy(options);
-    TryReopen(options);
+    Reopen(options);
    return true;
  } else {
    return false;
@@ -680,7 +680,7 @@ void DBTestBase::Close() {
 void DBTestBase::DestroyAndReopen(const Options& options) {
  // Destroy using last options
  Destroy(last_options_);
-  ASSERT_OK(TryReopen(options));
+  Reopen(options);
 }

 void DBTestBase::Destroy(const Options& options, bool delete_cf_paths) {
@@ -1129,11 +1129,12 @@ std::string DBTestBase::FilesPerLevel(int cf) {

 size_t DBTestBase::CountFiles() {
  std::vector<std::string> files;
-  env_->GetChildren(dbname_, &files);
+  EXPECT_OK(env_->GetChildren(dbname_, &files));

  std::vector<std::string> logfiles;
  if (dbname_ != last_options_.wal_dir) {
-    env_->GetChildren(last_options_.wal_dir, &logfiles);
+    Status s = env_->GetChildren(last_options_.wal_dir, &logfiles);
+    EXPECT_TRUE(s.ok() || s.IsNotFound());
  }

  return files.size() + logfiles.size();
@@ -1266,8 +1267,8 @@ void DBTestBase::GenerateNewRandomFile(Random* rnd, bool nowait) {
  }
  ASSERT_OK(Put("key" + rnd->RandomString(7), rnd->RandomString(200)));
  if (!nowait) {
-    dbfull()->TEST_WaitForFlushMemTable();
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
  }
 }

@@ -996,7 +996,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionStopStyleSimilarSize) {
  //   First, clean up memtable before inserting new data. This will generate
  //   a level-0 file, with size around 0.4 (according to previously written
  //   data amount).
-  dbfull()->Flush(FlushOptions());
+  ASSERT_OK(dbfull()->Flush(FlushOptions()));
  for (int num = 0; num < options.level0_file_num_compaction_trigger - 3;
       num++) {
    // Write 110KB (11 values, each 10K)
@@ -1781,7 +1781,7 @@ TEST_P(DBTestUniversalCompaction, FinalSortedRunCompactFilesConflict) {
      dbfull()->TEST_write_controler().GetCompactionPressureToken();

  ASSERT_OK(Put("key", "val"));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  ASSERT_EQ(NumTableFilesAtLevel(num_levels_ - 1), 1);
  ColumnFamilyMetaData cf_meta;
@@ -1807,7 +1807,7 @@ TEST_P(DBTestUniversalCompaction, FinalSortedRunCompactFilesConflict) {
      "DBTestUniversalCompaction:FinalSortedRunCompactFilesConflict:0");
  for (int i = 0; i < 2; ++i) {
    ASSERT_OK(Put("key", "val"));
-    Flush();
+    ASSERT_OK(Flush());
  }
  ASSERT_OK(dbfull()->TEST_WaitForCompact());

@@ -1911,7 +1911,7 @@ TEST_F(DBTestUniversalCompaction2, BasicL0toL1) {
  for (i = 0; i < 2000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  //  MoveFilesToLevel(6);
  ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));

@@ -1923,7 +1923,7 @@ TEST_F(DBTestUniversalCompaction2, BasicL0toL1) {
      ASSERT_OK(Put(Key(i), "val"));
    }
  }
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
@@ -1954,7 +1954,7 @@ TEST_F(DBTestUniversalCompaction2, SingleLevel) {
  for (i = 0; i < 2000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());

  for (i = 1999; i < kNumKeys; ++i) {
    if (i >= kNumKeys - kWindowSize &&
@@ -1964,7 +1964,7 @@ TEST_F(DBTestUniversalCompaction2, SingleLevel) {
      ASSERT_OK(Put(Key(i), "val"));
    }
  }
-  Flush();
+  ASSERT_OK(Flush()(;

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(1, NumTableFilesAtLevel(0));
@@ -1992,19 +1992,19 @@ TEST_F(DBTestUniversalCompaction2, MultipleLevels) {
  for (i = 0; i < 500; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 500; i < 1000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 1000; i < 1500; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 1500; i < 2000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
@@ -2013,15 +2013,15 @@ TEST_F(DBTestUniversalCompaction2, MultipleLevels) {
  for (i = 1999; i < 2333; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 2333; i < 2666; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 2666; i < 2999; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
@@ -2031,7 +2031,7 @@ TEST_F(DBTestUniversalCompaction2, MultipleLevels) {
  for (i = 1900; i < 2100; ++i) {
    ASSERT_OK(Delete(Key(i)));
  }
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
@@ -2064,19 +2064,19 @@ TEST_F(DBTestUniversalCompaction2, OverlappingL0) {
  for (i = 0; i < 2000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 2000; i < 3000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 3500; i < 4000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  for (i = 2900; i < 3100; ++i) {
    ASSERT_OK(Delete(Key(i)));
  }
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(2, NumTableFilesAtLevel(0));
@@ -2106,7 +2106,7 @@ TEST_F(DBTestUniversalCompaction2, IngestBehind) {
  for (i = 0; i < 2000; ++i) {
    ASSERT_OK(Put(Key(i), "val"));
  }
-  Flush();
+  ASSERT_OK(Flush());
  //  MoveFilesToLevel(6);
  ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));

@@ -2118,7 +2118,7 @@ TEST_F(DBTestUniversalCompaction2, IngestBehind) {
      ASSERT_OK(Put(Key(i), "val"));
    }
  }
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
@@ -2184,7 +2184,7 @@ TEST_F(DBTestUniversalCompaction2, PeriodicCompaction) {

  // Case 1: Oldest flushed file excceeds periodic compaction threshold.
  ASSERT_OK(Put("foo", "bar"));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_EQ(0, periodic_compactions);
  // Move clock forward so that the flushed file would qualify periodic
  // compaction.
@@ -2192,7 +2192,7 @@ TEST_F(DBTestUniversalCompaction2, PeriodicCompaction) {

  // Another flush would trigger compaction the oldest file.
  ASSERT_OK(Put("foo", "bar2"));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  ASSERT_EQ(1, periodic_compactions);
@@ -2203,7 +2203,7 @@ TEST_F(DBTestUniversalCompaction2, PeriodicCompaction) {
  periodic_compactions = 0;
  // A flush doesn't trigger a periodic compaction when threshold not hit
  ASSERT_OK(Put("foo", "bar2"));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(0, periodic_compactions);

@@ -2211,7 +2211,7 @@ TEST_F(DBTestUniversalCompaction2, PeriodicCompaction) {
  // a compaction
  ASSERT_OK(Put("foo", "bar2"));
  env_->MockSleepForSeconds(48 * 60 * 60 + 100);
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_EQ(1, periodic_compactions);
  ASSERT_EQ(0, start_level);
@@ -358,16 +358,16 @@ TEST_F(DBWALTest, RecoverWithBlob) {
  // There should be no files just yet since we haven't flushed.
  {
    VersionSet* const versions = dbfull()->TEST_GetVersionSet();
-    assert(versions);
+    ASSERT_NE(versions, nullptr);

    ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
-    assert(cfd);
+    ASSERT_NE(cfd, nullptr);

    Version* const current = cfd->current();
-    assert(current);
+    ASSERT_NE(current, nullptr);

    const VersionStorageInfo* const storage_info = current->storage_info();
-    assert(storage_info);
+    ASSERT_NE(storage_info, nullptr);

    ASSERT_EQ(storage_info->num_non_empty_levels(), 0);
    ASSERT_TRUE(storage_info->GetBlobFiles().empty());
@@ -388,28 +388,28 @@ TEST_F(DBWALTest, RecoverWithBlob) {
  ASSERT_EQ(Get("key2"), long_value);

  VersionSet* const versions = dbfull()->TEST_GetVersionSet();
-  assert(versions);
+  ASSERT_NE(versions, nullptr);

  ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
-  assert(cfd);
+  ASSERT_NE(cfd, nullptr);

  Version* const current = cfd->current();
-  assert(current);
+  ASSERT_NE(current, nullptr);

  const VersionStorageInfo* const storage_info = current->storage_info();
-  assert(storage_info);
+  ASSERT_NE(storage_info, nullptr);

  const auto& l0_files = storage_info->LevelFiles(0);
  ASSERT_EQ(l0_files.size(), 1);

  const FileMetaData* const table_file = l0_files[0];
-  assert(table_file);
+  ASSERT_NE(table_file, nullptr);

  const auto& blob_files = storage_info->GetBlobFiles();
  ASSERT_EQ(blob_files.size(), 1);

  const auto& blob_file = blob_files.begin()->second;
-  assert(blob_file);
+  ASSERT_NE(blob_file, nullptr);

  ASSERT_EQ(table_file->smallest.user_key(), "key1");
  ASSERT_EQ(table_file->largest.user_key(), "key2");
@@ -422,7 +422,7 @@ TEST_F(DBWALTest, RecoverWithBlob) {

 #ifndef ROCKSDB_LITE
  const InternalStats* const internal_stats = cfd->internal_stats();
-  assert(internal_stats);
+  ASSERT_NE(internal_stats, nullptr);

  const uint64_t expected_bytes =
      table_file->fd.GetFileSize() + blob_file->GetTotalBlobBytes();
@@ -502,12 +502,12 @@ TEST_F(DBWALTest, IgnoreRecoveredLog) {

  do {
    // delete old files in backup_logs directory
-    env_->CreateDirIfMissing(backup_logs);
+    ASSERT_OK(env_->CreateDirIfMissing(backup_logs));
    std::vector<std::string> old_files;
-    env_->GetChildren(backup_logs, &old_files);
+    ASSERT_OK(env_->GetChildren(backup_logs, &old_files));
    for (auto& file : old_files) {
      if (file != "." && file != "..") {
-        env_->DeleteFile(backup_logs + "/" + file);
+        ASSERT_OK(env_->DeleteFile(backup_logs + "/" + file));
      }
    }
    Options options = CurrentOptions();
@@ -526,7 +526,7 @@ TEST_F(DBWALTest, IgnoreRecoveredLog) {

    // copy the logs to backup
    std::vector<std::string> logs;
-    env_->GetChildren(options.wal_dir, &logs);
+    ASSERT_OK(env_->GetChildren(options.wal_dir, &logs));
    for (auto& log : logs) {
      if (log != ".." && log != ".") {
        CopyFile(options.wal_dir + "/" + log, backup_logs + "/" + log);
@@ -557,7 +557,7 @@ TEST_F(DBWALTest, IgnoreRecoveredLog) {
    Close();

    // copy the logs from backup back to wal dir
-    env_->CreateDirIfMissing(options.wal_dir);
+    ASSERT_OK(env_->CreateDirIfMissing(options.wal_dir));
    for (auto& log : logs) {
      if (log != ".." && log != ".") {
        CopyFile(backup_logs + "/" + log, options.wal_dir + "/" + log);
@@ -572,16 +572,16 @@ TEST_F(DBWALTest, IgnoreRecoveredLog) {
    // Recovery will fail if DB directory doesn't exist.
    Destroy(options);
    // copy the logs from backup back to wal dir
-    env_->CreateDirIfMissing(options.wal_dir);
+    ASSERT_OK(env_->CreateDirIfMissing(options.wal_dir));
    for (auto& log : logs) {
      if (log != ".." && log != ".") {
        CopyFile(backup_logs + "/" + log, options.wal_dir + "/" + log);
        // we won't be needing this file no more
-        env_->DeleteFile(backup_logs + "/" + log);
+        ASSERT_OK(env_->DeleteFile(backup_logs + "/" + log));
      }
    }
    Status s = TryReopen(options);
-    ASSERT_TRUE(!s.ok());
+    ASSERT_NOK(s);
    Destroy(options);
  } while (ChangeWalOptions());
 }
@@ -619,9 +619,9 @@ TEST_F(DBWALTest, PreallocateBlock) {
        called.fetch_add(1);
      });
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
-  Put("", "");
-  Flush();
-  Put("", "");
+  ASSERT_OK(Put("", ""));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("", ""));
  Close();
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_EQ(2, called.load());
@@ -638,9 +638,9 @@ TEST_F(DBWALTest, PreallocateBlock) {
        called.fetch_add(1);
      });
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
-  Put("", "");
-  Flush();
-  Put("", "");
+  ASSERT_OK(Put("", ""));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("", ""));
  Close();
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_EQ(2, called.load());
@@ -658,9 +658,9 @@ TEST_F(DBWALTest, PreallocateBlock) {
        called.fetch_add(1);
      });
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
-  Put("", "");
-  Flush();
-  Put("", "");
+  ASSERT_OK(Put("", ""));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("", ""));
  Close();
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_EQ(2, called.load());
@@ -679,9 +679,9 @@ TEST_F(DBWALTest, PreallocateBlock) {
        called.fetch_add(1);
      });
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
-  Put("", "");
-  Flush();
-  Put("", "");
+  ASSERT_OK(Put("", ""));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("", ""));
  Close();
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_EQ(2, called.load());
@@ -907,7 +907,7 @@ TEST_F(DBWALTest, RecoverCheckFileAmountWithSmallWriteBuffer) {
  // Make 'dobrynia' to be flushed and new WAL file to be created
  ASSERT_OK(Put(2, Key(10), DummyString(7500000)));
  ASSERT_OK(Put(2, Key(1), DummyString(1)));
-  dbfull()->TEST_WaitForFlushMemTable(handles_[2]);
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(handles_[2]));
  {
    auto tables = ListTableFiles(env_, dbname_);
    ASSERT_EQ(tables.size(), static_cast<size_t>(1));
@@ -961,7 +961,7 @@ TEST_F(DBWALTest, RecoverCheckFileAmount) {
  // Make 'nikitich' memtable to be flushed
  ASSERT_OK(Put(3, Key(10), DummyString(1002400)));
  ASSERT_OK(Put(3, Key(1), DummyString(1)));
-  dbfull()->TEST_WaitForFlushMemTable(handles_[3]);
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(handles_[3]));
  // 4 memtable are not flushed, 1 sst file
  {
    auto tables = ListTableFiles(env_, dbname_);
@@ -981,7 +981,7 @@ TEST_F(DBWALTest, RecoverCheckFileAmount) {
  ASSERT_OK(Put(3, Key(10), DummyString(1002400)));
  // make it flush
  ASSERT_OK(Put(3, Key(1), DummyString(1)));
-  dbfull()->TEST_WaitForFlushMemTable(handles_[3]);
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(handles_[3]));
  // There are still 4 memtable not flushed, and 2 sst tables
  ASSERT_OK(Put(0, Key(1), DummyString(1)));
  ASSERT_OK(Put(1, Key(1), DummyString(1)));
@@ -1029,10 +1029,10 @@ TEST_F(DBWALTest, SyncMultipleLogs) {
  for (uint64_t b = 0; b < kNumBatches; b++) {
    batch.Clear();
    for (int i = 0; i < kBatchSize; i++) {
-      batch.Put(Key(i), DummyString(128));
+      ASSERT_OK(batch.Put(Key(i), DummyString(128)));
    }

-    dbfull()->Write(wo, &batch);
+    ASSERT_OK(dbfull()->Write(wo, &batch));
  }

  ASSERT_OK(dbfull()->SyncWAL());
@@ -1060,7 +1060,7 @@ TEST_F(DBWALTest, PartOfWritesWithWALDisabled) {
  ASSERT_OK(Flush(0));
  ASSERT_OK(Put(0, "key", "v5", wal_on));  // seq id 5
  ASSERT_EQ("v5", Get(0, "key"));
-  dbfull()->FlushWAL(false);
+  ASSERT_OK(dbfull()->FlushWAL(false));
  // Simulate a crash.
  fault_env->SetFilesystemActive(false);
  Close();
@@ -1128,12 +1128,13 @@ class RecoveryTestHelper {
      for (int i = 0; i < kKeysPerWALFile; i++) {
        std::string key = "key" + ToString((*count)++);
        std::string value = test->DummyString(kValueSize);
-        assert(current_log_writer.get() != nullptr);
+        ASSERT_NE(current_log_writer.get(), nullptr);
        uint64_t seq = versions->LastSequence() + 1;
        batch.Clear();
-        batch.Put(key, value);
+        ASSERT_OK(batch.Put(key, value));
        WriteBatchInternal::SetSequence(&batch, seq);
-        current_log_writer->AddRecord(WriteBatchInternal::Contents(&batch));
+        ASSERT_OK(current_log_writer->AddRecord(
+            WriteBatchInternal::Contents(&batch)));
        versions->SetLastAllocatedSequence(seq);
        versions->SetLastPublishedSequence(seq);
        versions->SetLastSequence(seq);
@@ -1309,10 +1310,11 @@ TEST_F(DBWALTest, kPointInTimeRecoveryCFConsistency) {

  ASSERT_OK(Put(1, "key3", "val3"));
  // Corrupt WAL at location of key3
-  test::CorruptFile(env, fname, static_cast<int>(offset_to_corrupt), 4, false);
+  ASSERT_OK(test::CorruptFile(env, fname, static_cast<int>(offset_to_corrupt),
+                              4, false));
  ASSERT_OK(Put(2, "key4", "val4"));
  ASSERT_OK(Put(1, "key5", "val5"));
-  Flush(2);
+  ASSERT_OK(Flush(2));

  // PIT recovery & verify
  options.wal_recovery_mode = WALRecoveryMode::kPointInTimeRecovery;
@@ -1348,13 +1350,19 @@ TEST_P(DBWALTestWithParams, kPointInTimeRecovery) {
  size_t recovered_row_count = RecoveryTestHelper::GetData(this);
  ASSERT_LT(recovered_row_count, row_count);

-  bool expect_data = true;
-  for (size_t k = 0; k < maxkeys; ++k) {
-    bool found = Get("key" + ToString(corrupt_offset)) != "NOT_FOUND";
-    if (expect_data && !found) {
-      expect_data = false;
+  // Verify a prefix of keys were recovered. But not in the case of full WAL
+  // truncation, because we have no way to know there was a corruption when
+  // truncation happened on record boundaries (preventing recovery holes in
+  // that case requires using `track_and_verify_wals_in_manifest`).
+  if (!trunc || corrupt_offset != 0) {
+    bool expect_data = true;
+    for (size_t k = 0; k < maxkeys; ++k) {
+      bool found = Get("key" + ToString(k)) != "NOT_FOUND";
+      if (expect_data && !found) {
+        expect_data = false;
+      }
+      ASSERT_EQ(found, expect_data);
    }
-    ASSERT_EQ(found, expect_data);
  }

  const size_t min = RecoveryTestHelper::kKeysPerWALFile *
@@ -1460,7 +1468,7 @@ TEST_F(DBWALTest, WalCleanupAfterAvoidFlushDuringRecovery) {
  for (int i = 0; i < 2; ++i) {
    if (i > 0) {
      // Flush() triggers deletion of obsolete tracked files
-      Flush();
+      ASSERT_OK(Flush());
    }
    VectorLogPtr log_files;
    ASSERT_OK(dbfull()->GetSortedWalFiles(log_files));
@@ -1502,7 +1510,7 @@ TEST_F(DBWALTest, RecoverWithoutFlush) {
  ASSERT_EQ(Get("foo"), "foo_v2");
  ASSERT_EQ(Get("bar"), "bar_v2");
  // manual flush and insert again
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_EQ(Get("foo"), "foo_v2");
  ASSERT_EQ(Get("bar"), "bar_v2");
  ASSERT_OK(Put("foo", "foo_v3"));
@@ -1523,7 +1531,9 @@ TEST_F(DBWALTest, RecoverWithoutFlushMultipleCF) {

  auto countWalFiles = [this]() {
    VectorLogPtr log_files;
-    dbfull()->GetSortedWalFiles(log_files);
+    if (!dbfull()->GetSortedWalFiles(log_files).ok()) {
+      return size_t{0};
+    }
    return log_files.size();
  };

@@ -1531,11 +1541,11 @@ TEST_F(DBWALTest, RecoverWithoutFlushMultipleCF) {
  CreateAndReopenWithCF({"one", "two"}, options);
  ASSERT_OK(Put(0, "key1", kSmallValue));
  ASSERT_OK(Put(1, "key2", kLargeValue));
-  Flush(1);
+  ASSERT_OK(Flush(1));
  ASSERT_EQ(1, countWalFiles());
  ASSERT_OK(Put(0, "key3", kSmallValue));
  ASSERT_OK(Put(2, "key4", kLargeValue));
-  Flush(2);
+  ASSERT_OK(Flush(2));
  ASSERT_EQ(2, countWalFiles());

  // Reopen, insert and flush.
@@ -1549,9 +1559,9 @@ TEST_F(DBWALTest, RecoverWithoutFlushMultipleCF) {
  ASSERT_OK(Put(0, "key5", kLargeValue));
  ASSERT_OK(Put(1, "key6", kLargeValue));
  ASSERT_EQ(3, countWalFiles());
-  Flush(1);
+  ASSERT_OK(Flush(1));
  ASSERT_OK(Put(2, "key7", kLargeValue));
-  dbfull()->FlushWAL(false);
+  ASSERT_OK(dbfull()->FlushWAL(false));
  ASSERT_EQ(4, countWalFiles());

  // Reopen twice and validate.
@@ -1640,7 +1650,7 @@ TEST_F(DBWALTest, RestoreTotalLogSizeAfterRecoverWithoutFlush) {

    void OnFlushBegin(DB* /*db*/, const FlushJobInfo& flush_job_info) override {
      count++;
-      assert(FlushReason::kWriteBufferManager == flush_job_info.flush_reason);
+      ASSERT_EQ(FlushReason::kWriteBufferManager, flush_job_info.flush_reason);
    }
  };
  std::shared_ptr<TestFlushListener> test_listener =
@@ -1684,7 +1694,9 @@ TEST_F(DBWALTest, RestoreTotalLogSizeAfterRecoverWithoutFlush) {
            1 * kMB);
  // Write one more key to trigger flush.
  ASSERT_OK(Put(0, "foo", "v2"));
-  dbfull()->TEST_WaitForFlushMemTable();
+  for (auto* h : handles_) {
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(h));
+  }
  // Flushed two column families.
  ASSERT_EQ(2, test_listener->count.load());
 }
@@ -1758,9 +1770,9 @@ TEST_F(DBWALTest, WalTermTest) {
  wo.disableWAL = false;

  WriteBatch batch;
-  batch.Put("foo", "bar");
+  ASSERT_OK(batch.Put("foo", "bar"));
  batch.MarkWalTerminationPoint();
-  batch.Put("foo2", "bar2");
+  ASSERT_OK(batch.Put("foo2", "bar2"));

  ASSERT_OK(dbfull()->Write(wo, &batch));

@@ -32,6 +32,22 @@ class DBBasicTestWithTimestampBase : public DBTestBase {
    return ret;
  }

+  static std::string KeyWithPrefix(std::string prefix, uint64_t k) {
+    std::string ret;
+    PutFixed64(&ret, k);
+    std::reverse(ret.begin(), ret.end());
+    return prefix + ret;
+  }
+
+  static std::vector<Slice> ConvertStrToSlice(
+      std::vector<std::string>& strings) {
+    std::vector<Slice> ret;
+    for (const auto& s : strings) {
+      ret.emplace_back(s);
+    }
+    return ret;
+  }
+
  class TestComparator : public Comparator {
   private:
    const Comparator* cmp_without_ts_;
@@ -179,6 +195,113 @@ class DBBasicTestWithTimestamp : public DBBasicTestWithTimestampBase {
      : DBBasicTestWithTimestampBase("db_basic_test_with_timestamp") {}
 };

+TEST_F(DBBasicTestWithTimestamp, CompactRangeWithSpecifiedRange) {
+  Options options = CurrentOptions();
+  options.env = env_;
+  options.create_if_missing = true;
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+
+  WriteOptions write_opts;
+  std::string ts_str = Timestamp(1, 0);
+  Slice ts = ts_str;
+  write_opts.timestamp = &ts;
+
+  ASSERT_OK(db_->Put(write_opts, "foo1", "bar"));
+  ASSERT_OK(Flush());
+
+  ASSERT_OK(db_->Put(write_opts, "foo2", "bar"));
+  ASSERT_OK(Flush());
+
+  std::string start_str = "foo";
+  std::string end_str = "foo2";
+  Slice start(start_str), end(end_str);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), &start, &end));
+
+  Close();
+}
+
+TEST_F(DBBasicTestWithTimestamp, GetApproximateSizes) {
+  Options options = CurrentOptions();
+  options.write_buffer_size = 100000000;  // Large write buffer
+  options.compression = kNoCompression;
+  options.create_if_missing = true;
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+  auto default_cf = db_->DefaultColumnFamily();
+
+  WriteOptions write_opts;
+  std::string ts_str = Timestamp(1, 0);
+  Slice ts = ts_str;
+  write_opts.timestamp = &ts;
+
+  const int N = 128;
+  Random rnd(301);
+  for (int i = 0; i < N; i++) {
+    ASSERT_OK(db_->Put(write_opts, Key(i), rnd.RandomString(1024)));
+  }
+
+  uint64_t size;
+  std::string start = Key(50);
+  std::string end = Key(60);
+  Range r(start, end);
+  SizeApproximationOptions size_approx_options;
+  size_approx_options.include_memtabtles = true;
+  size_approx_options.include_files = true;
+  ASSERT_OK(
+      db_->GetApproximateSizes(size_approx_options, default_cf, &r, 1, &size));
+  ASSERT_GT(size, 6000);
+  ASSERT_LT(size, 204800);
+
+  // test multiple ranges
+  std::vector<Range> ranges;
+  std::string start_tmp = Key(10);
+  std::string end_tmp = Key(20);
+  ranges.emplace_back(Range(start_tmp, end_tmp));
+  ranges.emplace_back(Range(start, end));
+  uint64_t range_sizes[2];
+  ASSERT_OK(db_->GetApproximateSizes(size_approx_options, default_cf,
+                                     ranges.data(), 2, range_sizes));
+
+  ASSERT_EQ(range_sizes[1], size);
+
+  // Zero if not including mem table
+  db_->GetApproximateSizes(&r, 1, &size);
+  ASSERT_EQ(size, 0);
+
+  start = Key(500);
+  end = Key(600);
+  r = Range(start, end);
+  ASSERT_OK(
+      db_->GetApproximateSizes(size_approx_options, default_cf, &r, 1, &size));
+  ASSERT_EQ(size, 0);
+
+  // Test range boundaries
+  ASSERT_OK(db_->Put(write_opts, Key(1000), rnd.RandomString(1024)));
+  // Should include start key
+  start = Key(1000);
+  end = Key(1100);
+  r = Range(start, end);
+  ASSERT_OK(
+      db_->GetApproximateSizes(size_approx_options, default_cf, &r, 1, &size));
+  ASSERT_GT(size, 0);
+
+  // Should exclude end key
+  start = Key(900);
+  end = Key(1000);
+  r = Range(start, end);
+  ASSERT_OK(
+      db_->GetApproximateSizes(size_approx_options, default_cf, &r, 1, &size));
+  ASSERT_EQ(size, 0);
+  std::cout << size << std::endl;
+
+  Close();
+}
+
 TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterate) {
  const int kNumKeysPerFile = 128;
  const uint64_t kMaxKey = 1024;
@@ -249,6 +372,8 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLessThanKey) {
  options.env = env_;
  options.create_if_missing = true;
  options.prefix_extractor.reset(NewFixedPrefixTransform(3));
+  options.memtable_whole_key_filtering = true;
+  options.memtable_prefix_bloom_size_ratio = 0.1;
  BlockBasedTableOptions bbto;
  bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
  bbto.cache_index_and_filter_blocks = true;
@@ -265,16 +390,16 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLessThanKey) {
  write_opts.timestamp = &ts;

  ASSERT_OK(db_->Put(write_opts, "foo1", "bar"));
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(db_->Put(write_opts, "foo2", "bar"));
-  Flush();
+  ASSERT_OK(Flush());

  // Move sst file to next level
  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  ASSERT_OK(db_->Put(write_opts, "foo3", "bar"));
-  Flush();
+  ASSERT_OK(Flush());

  ReadOptions read_opts;
  std::string read_ts = Timestamp(1, 0);
@@ -285,6 +410,10 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLessThanKey) {
    iter->Seek("foo");
    ASSERT_TRUE(iter->Valid());
    ASSERT_OK(iter->status());
+    iter->Next();
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_OK(iter->status());
+
    iter->Seek("bbb");
    ASSERT_FALSE(iter->Valid());
    ASSERT_OK(iter->status());
@@ -298,6 +427,8 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLargerThanKey) {
  options.env = env_;
  options.create_if_missing = true;
  options.prefix_extractor.reset(NewFixedPrefixTransform(20));
+  options.memtable_whole_key_filtering = true;
+  options.memtable_prefix_bloom_size_ratio = 0.1;
  BlockBasedTableOptions bbto;
  bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
  bbto.cache_index_and_filter_blocks = true;
@@ -314,21 +445,76 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLargerThanKey) {
  write_opts.timestamp = &ts;

  ASSERT_OK(db_->Put(write_opts, "foo1", "bar"));
-  Flush();
+  ASSERT_OK(Flush());

  ASSERT_OK(db_->Put(write_opts, "foo2", "bar"));
-  Flush();
+  ASSERT_OK(Flush());

  // Move sst file to next level
  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  ASSERT_OK(db_->Put(write_opts, "foo3", "bar"));
-  Flush();
+  ASSERT_OK(Flush());

  ReadOptions read_opts;
  std::string read_ts = Timestamp(2, 0);
  ts = read_ts;
  read_opts.timestamp = &ts;
+  {
+    std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts));
+    // Make sure the prefix extractor doesn't include timestamp, otherwise it
+    // may return invalid result.
+    iter->Seek("foo");
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_OK(iter->status());
+    iter->Next();
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_OK(iter->status());
+  }
+
+  Close();
+}
+
+TEST_F(DBBasicTestWithTimestamp, SeekWithBound) {
+  Options options = CurrentOptions();
+  options.env = env_;
+  options.create_if_missing = true;
+  options.prefix_extractor.reset(NewFixedPrefixTransform(2));
+  BlockBasedTableOptions bbto;
+  bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
+  bbto.cache_index_and_filter_blocks = true;
+  bbto.whole_key_filtering = true;
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+
+  WriteOptions write_opts;
+  std::string ts_str = Timestamp(1, 0);
+  Slice ts = ts_str;
+  write_opts.timestamp = &ts;
+
+  ASSERT_OK(db_->Put(write_opts, "foo1", "bar"));
+  ASSERT_OK(Flush());
+
+  ASSERT_OK(db_->Put(write_opts, "foo2", "bar"));
+  ASSERT_OK(Flush());
+
+  // Move sst file to next level
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
+
+  ASSERT_OK(db_->Put(write_opts, "foo3", "bar"));
+  ASSERT_OK(Flush());
+
+  ReadOptions read_opts;
+  std::string read_ts = Timestamp(2, 0);
+  ts = read_ts;
+  read_opts.timestamp = &ts;
+  std::string up_bound = "foo5";
+  Slice up_bound_slice = up_bound;
+  read_opts.iterate_upper_bound = &up_bound_slice;
+  read_opts.auto_prefix_mode = true;
  {
    std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts));
    // Make sure the prefix extractor doesn't include timestamp, otherwise it
@@ -611,7 +797,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetWithFastLocalBloom) {

  ASSERT_OK(db_->Put(write_opts, "foo", "bar"));

-  Flush();
+  ASSERT_OK(Flush());

  // Read with MultiGet
  ReadOptions read_opts;
@@ -652,7 +838,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetWithPrefix) {

  ASSERT_OK(db_->Put(write_opts, "foo", "bar"));

-  Flush();
+  ASSERT_OK(Flush());

  // Read with MultiGet
  ReadOptions read_opts;
@@ -740,7 +926,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetRangeFiltering) {
    Slice key_slice = key;
    Slice value_slice = value;
    ASSERT_OK(db_->Put(write_opts, key_slice, value_slice));
-    Flush();
+    ASSERT_OK(Flush());
  }

  // Make num_levels to 2 to do key range filtering of sst files
@@ -748,7 +934,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetRangeFiltering) {

  ASSERT_OK(db_->Put(write_opts, "foo", "bar"));

-  Flush();
+  ASSERT_OK(Flush());

  // Read with MultiGet
  ts_str = Timestamp(2, 0);
@@ -791,7 +977,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetPrefixFilter) {

  ASSERT_OK(db_->Put(write_opts, "foo", "bar"));

-  Flush();
+  ASSERT_OK(Flush());
  // Read with MultiGet
  ts_str = Timestamp(2, 0);
  ts = ts_str;
@@ -909,6 +1095,131 @@ TEST_F(DBBasicTestWithTimestamp, CompactDeletionWithTimestampMarkerToBottom) {
  Close();
 }

+class DBBasicTestWithTimestampFilterPrefixSettings
+    : public DBBasicTestWithTimestampBase,
+      public testing::WithParamInterface<
+          std::tuple<std::shared_ptr<const FilterPolicy>, bool, bool,
+                     std::shared_ptr<const SliceTransform>, bool, double>> {
+ public:
+  DBBasicTestWithTimestampFilterPrefixSettings()
+      : DBBasicTestWithTimestampBase(
+            "db_basic_test_with_timestamp_filter_prefix") {}
+};
+
+TEST_P(DBBasicTestWithTimestampFilterPrefixSettings, GetAndMultiGet) {
+  Options options = CurrentOptions();
+  options.env = env_;
+  options.create_if_missing = true;
+  BlockBasedTableOptions bbto;
+  bbto.filter_policy = std::get<0>(GetParam());
+  bbto.whole_key_filtering = std::get<1>(GetParam());
+  bbto.cache_index_and_filter_blocks = std::get<2>(GetParam());
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  options.prefix_extractor = std::get<3>(GetParam());
+  options.memtable_whole_key_filtering = std::get<4>(GetParam());
+  options.memtable_prefix_bloom_size_ratio = std::get<5>(GetParam());
+
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+  const int kMaxKey = 1000;
+
+  // Write any value
+  WriteOptions write_opts;
+  std::string ts_str = Timestamp(1, 0);
+  Slice ts = ts_str;
+  write_opts.timestamp = &ts;
+
+  int idx = 0;
+  for (; idx < kMaxKey / 4; idx++) {
+    ASSERT_OK(db_->Put(write_opts, Key1(idx), "bar"));
+    ASSERT_OK(db_->Put(write_opts, KeyWithPrefix("foo", idx), "bar"));
+  }
+
+  ASSERT_OK(Flush());
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
+
+  for (; idx < kMaxKey / 2; idx++) {
+    ASSERT_OK(db_->Put(write_opts, Key1(idx), "bar"));
+    ASSERT_OK(db_->Put(write_opts, KeyWithPrefix("foo", idx), "bar"));
+  }
+
+  ASSERT_OK(Flush());
+
+  for (; idx < kMaxKey; idx++) {
+    ASSERT_OK(db_->Put(write_opts, Key1(idx), "bar"));
+    ASSERT_OK(db_->Put(write_opts, KeyWithPrefix("foo", idx), "bar"));
+  }
+
+  // Read with MultiGet
+  ReadOptions read_opts;
+  read_opts.timestamp = &ts;
+
+  ReadOptions read_opts_total_order;
+  read_opts_total_order.timestamp = &ts;
+  read_opts_total_order.total_order_seek = true;
+
+  for (idx = 0; idx < kMaxKey; idx++) {
+    size_t batch_size = 4;
+    std::vector<std::string> keys_str(batch_size);
+    std::vector<PinnableSlice> values(batch_size);
+    std::vector<Status> statuses(batch_size);
+    ColumnFamilyHandle* cfh = db_->DefaultColumnFamily();
+
+    keys_str[0] = Key1(idx);
+    keys_str[1] = KeyWithPrefix("foo", idx);
+    keys_str[2] = Key1(kMaxKey + idx);
+    keys_str[3] = KeyWithPrefix("foo", kMaxKey + idx);
+
+    auto keys = ConvertStrToSlice(keys_str);
+
+    db_->MultiGet(read_opts, cfh, batch_size, keys.data(), values.data(),
+                  statuses.data());
+
+    for (int i = 0; i < 2; i++) {
+      ASSERT_OK(statuses[i]);
+    }
+    for (int i = 2; i < 4; i++) {
+      ASSERT_TRUE(statuses[i].IsNotFound());
+    }
+
+    for (int i = 0; i < 2; i++) {
+      std::string value;
+      ASSERT_OK(db_->Get(read_opts, keys[i], &value));
+      std::unique_ptr<Iterator> it1(db_->NewIterator(read_opts));
+      ASSERT_NE(nullptr, it1);
+      ASSERT_OK(it1->status());
+      // TODO(zjay) Fix seek with prefix
+      // it1->Seek(keys[i]);
+      // ASSERT_TRUE(it1->Valid());
+    }
+
+    for (int i = 2; i < 4; i++) {
+      std::string value;
+      Status s = db_->Get(read_opts, keys[i], &value);
+      ASSERT_TRUE(s.IsNotFound());
+    }
+  }
+  Close();
+}
+
+INSTANTIATE_TEST_CASE_P(
+    Timestamp, DBBasicTestWithTimestampFilterPrefixSettings,
+    ::testing::Combine(
+        ::testing::Values(
+            std::shared_ptr<const FilterPolicy>(nullptr),
+            std::shared_ptr<const FilterPolicy>(NewBloomFilterPolicy(10, true)),
+            std::shared_ptr<const FilterPolicy>(NewBloomFilterPolicy(10,
+                                                                     false))),
+        ::testing::Bool(), ::testing::Bool(),
+        ::testing::Values(
+            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(1)),
+            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(4)),
+            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(7)),
+            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(8))),
+        ::testing::Bool(), ::testing::Values(0, 0.1)));
+
 class DataVisibilityTest : public DBBasicTestWithTimestampBase {
 public:
  DataVisibilityTest() : DBBasicTestWithTimestampBase("data_visibility_test") {
@@ -1372,7 +1683,7 @@ TEST_F(DataVisibilityTest, MultiGetWithTimestamp) {
  VerifyDefaultCF(snap0);
  VerifyDefaultCF(snap1);

-  Flush();
+  ASSERT_OK(Flush());

  const Snapshot* snap2 = db_->GetSnapshot();
  PutTestData(2);
@@ -1458,7 +1769,7 @@ TEST_F(DataVisibilityTest, MultiGetCrossCF) {
  VerifyDefaultCF(snap0);
  VerifyDefaultCF(snap1);

-  Flush();
+  ASSERT_OK(Flush());

  const Snapshot* snap2 = db_->GetSnapshot();
  PutTestData(2);
@@ -1839,6 +2150,8 @@ TEST_F(DBBasicTestWithTimestamp, BatchWriteAndMultiGet) {
  options.create_if_missing = true;
  options.env = env_;
  options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeysPerFile));
+  options.memtable_prefix_bloom_size_ratio = 0.1;
+  options.memtable_whole_key_filtering = true;

  size_t ts_sz = Timestamp(0, 0).size();
  TestComparator test_cmp(ts_sz);
@@ -2093,6 +2406,7 @@ INSTANTIATE_TEST_CASE_P(
    Timestamp, DBBasicTestWithTimestampPrefixSeek,
    ::testing::Combine(
        ::testing::Values(
+            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(1)),
            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(4)),
            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(7)),
            std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(8))),
@@ -60,14 +60,15 @@ TEST_P(DBWriteTest, WriteStallRemoveNoSlowdownWrite) {
    std::string key = "foo" + std::to_string(a);
    WriteOptions wo;
    wo.no_slowdown = false;
-    dbfull()->Put(wo, key, "bar");
+    ASSERT_OK(dbfull()->Put(wo, key, "bar"));
  };
  std::function<void()> write_no_slowdown_func = [&]() {
    int a = thread_num.fetch_add(1);
    std::string key = "foo" + std::to_string(a);
    WriteOptions wo;
    wo.no_slowdown = true;
-    dbfull()->Put(wo, key, "bar");
+    Status s = dbfull()->Put(wo, key, "bar");
+    ASSERT_TRUE(s.ok() || s.IsIncomplete());
  };
  std::function<void(void*)> unblock_main_thread_func = [&](void*) {
    mutex.Lock();
@@ -77,13 +78,13 @@ TEST_P(DBWriteTest, WriteStallRemoveNoSlowdownWrite) {
  };

  // Create 3 L0 files and schedule 4th without waiting
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
-  Flush();
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
-  Flush();
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
-  Flush();
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "WriteThread::JoinBatchGroup:Start", unblock_main_thread_func);
@@ -104,7 +105,7 @@ TEST_P(DBWriteTest, WriteStallRemoveNoSlowdownWrite) {
  // write_thread
  FlushOptions fopt;
  fopt.wait = false;
-  dbfull()->Flush(fopt);
+  ASSERT_OK(dbfull()->Flush(fopt));

  // Create a mix of slowdown/no_slowdown write threads
  mutex.Lock();
@@ -145,7 +146,7 @@ TEST_P(DBWriteTest, WriteStallRemoveNoSlowdownWrite) {
  mutex.Unlock();

  TEST_SYNC_POINT("DBWriteTest::WriteStallRemoveNoSlowdownWrite:1");
-  dbfull()->TEST_WaitForFlushMemTable(nullptr);
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(nullptr));
  // This would have triggered a write stall. Unblock the write group leader
  TEST_SYNC_POINT("DBWriteTest::WriteStallRemoveNoSlowdownWrite:2");
  // The leader is going to create missing newer links. When the leader
@@ -178,14 +179,15 @@ TEST_P(DBWriteTest, WriteThreadHangOnWriteStall) {
    std::string key = "foo" + std::to_string(a);
    WriteOptions wo;
    wo.no_slowdown = false;
-    dbfull()->Put(wo, key, "bar");
+    ASSERT_OK(dbfull()->Put(wo, key, "bar"));
  };
  std::function<void()> write_no_slowdown_func = [&]() {
    int a = thread_num.fetch_add(1);
    std::string key = "foo" + std::to_string(a);
    WriteOptions wo;
    wo.no_slowdown = true;
-    dbfull()->Put(wo, key, "bar");
+    Status s = dbfull()->Put(wo, key, "bar");
+    ASSERT_TRUE(s.ok() || s.IsIncomplete());
  };
  std::function<void(void *)> unblock_main_thread_func = [&](void *) {
    mutex.Lock();
@@ -195,13 +197,13 @@ TEST_P(DBWriteTest, WriteThreadHangOnWriteStall) {
  };

  // Create 3 L0 files and schedule 4th without waiting
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
-  Flush();
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
-  Flush();
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
-  Flush();
-  Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar");
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put("foo" + std::to_string(thread_num.fetch_add(1)), "bar"));

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "WriteThread::JoinBatchGroup:Start", unblock_main_thread_func);
@@ -222,7 +224,7 @@ TEST_P(DBWriteTest, WriteThreadHangOnWriteStall) {
  // write_thread
  FlushOptions fopt;
  fopt.wait = false;
-  dbfull()->Flush(fopt);
+  ASSERT_OK(dbfull()->Flush(fopt));

  // Create a mix of slowdown/no_slowdown write threads
  mutex.Lock();
@@ -243,7 +245,7 @@ TEST_P(DBWriteTest, WriteThreadHangOnWriteStall) {
  mutex.Unlock();

  TEST_SYNC_POINT("DBWriteTest::WriteThreadHangOnWriteStall:1");
-  dbfull()->TEST_WaitForFlushMemTable(nullptr);
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable(nullptr));
  // This would have triggered a write stall. Unblock the write group leader
  TEST_SYNC_POINT("DBWriteTest::WriteThreadHangOnWriteStall:2");
  // The leader is going to create missing newer links. When the leader finishes,
@@ -307,6 +309,11 @@ TEST_P(DBWriteTest, IOErrorOnWALWritePropagateToWriteThreadFollower) {
    threads[i].join();
  }
  ASSERT_EQ(1, leader_count);
+
+  // The Failed PUT operations can cause a BG error to be set.
+  // Mark it as Checked for the ASSERT_STATUS_CHECKED
+  dbfull()->Resume().PermitUncheckedError();
+
  // Close before mock_env destruct.
  Close();
 }
@@ -320,7 +327,7 @@ TEST_P(DBWriteTest, ManualWalFlushInEffect) {
  ASSERT_TRUE(dbfull()->FlushWAL(false).ok());
  ASSERT_TRUE(dbfull()->TEST_WALBufferIsEmpty());
  // try the 2nd wal created during SwitchWAL
-  dbfull()->TEST_SwitchWAL();
+  ASSERT_OK(dbfull()->TEST_SwitchWAL());
  ASSERT_TRUE(Put("key" + ToString(0), "value").ok());
  ASSERT_TRUE(options.manual_wal_flush != dbfull()->TEST_WALBufferIsEmpty());
  ASSERT_TRUE(dbfull()->FlushWAL(false).ok());
@@ -351,7 +358,9 @@ TEST_P(DBWriteTest, IOErrorOnWALWriteTriggersReadOnlyMode) {
    }
    */
    if (!options.manual_wal_flush) {
-      ASSERT_FALSE(res.ok());
+      ASSERT_NOK(res);
+    } else {
+      ASSERT_OK(res);
    }
  }
  // Close before mock_env destruct.
@@ -395,7 +404,7 @@ TEST_P(DBWriteTest, LockWalInEffect) {
  ASSERT_TRUE(dbfull()->TEST_WALBufferIsEmpty(false));
  ASSERT_OK(dbfull()->UnlockWAL());
  // try the 2nd wal created during SwitchWAL
-  dbfull()->TEST_SwitchWAL();
+  ASSERT_OK(dbfull()->TEST_SwitchWAL());
  ASSERT_OK(Put("key" + ToString(0), "value"));
  ASSERT_TRUE(options.manual_wal_flush != dbfull()->TEST_WALBufferIsEmpty());
  ASSERT_OK(dbfull()->LockWAL());
@@ -423,13 +432,14 @@ TEST_P(DBWriteTest, ConcurrentlyDisabledWAL) {
              ROCKSDB_NAMESPACE::WriteOptions write_option_default;
              std::string no_wal_key = no_wal_key_prefix + std::to_string(t) +
                                       "_" + std::to_string(i);
-              this->Put(no_wal_key, no_wal_value, write_option_disable);
+              ASSERT_OK(
+                  this->Put(no_wal_key, no_wal_value, write_option_disable));
              std::string wal_key =
                  wal_key_prefix + std::to_string(i) + "_" + std::to_string(i);
-              this->Put(wal_key, wal_value, write_option_default);
-              dbfull()->SyncWAL();
+              ASSERT_OK(this->Put(wal_key, wal_value, write_option_default));
+              ASSERT_OK(dbfull()->SyncWAL());
            }
-            return 0;
+            return;
        });
    }
    for (auto& t: threads) {
@@ -78,6 +78,22 @@ void AppendInternalKeyFooter(std::string* result, SequenceNumber s,
  PutFixed64(result, PackSequenceAndType(s, t));
 }

+void AppendKeyWithMinTimestamp(std::string* result, const Slice& key,
+                               size_t ts_sz) {
+  assert(ts_sz > 0);
+  const std::string kTsMin(ts_sz, static_cast<unsigned char>(0));
+  result->append(key.data(), key.size());
+  result->append(kTsMin.data(), ts_sz);
+}
+
+void AppendKeyWithMaxTimestamp(std::string* result, const Slice& key,
+                               size_t ts_sz) {
+  assert(ts_sz > 0);
+  const std::string kTsMax(ts_sz, static_cast<unsigned char>(0xff));
+  result->append(key.data(), key.size());
+  result->append(kTsMax.data(), ts_sz);
+}
+
 std::string ParsedInternalKey::DebugString(bool log_err_key, bool hex) const {
  std::string result = "'";
  if (log_err_key) {
@@ -167,6 +167,14 @@ extern void AppendInternalKeyWithDifferentTimestamp(
 extern void AppendInternalKeyFooter(std::string* result, SequenceNumber s,
                                    ValueType t);

+// Append the key and a minimal timestamp to *result
+extern void AppendKeyWithMinTimestamp(std::string* result, const Slice& key,
+                                      size_t ts_sz);
+
+// Append the key and a maximal timestamp to *result
+extern void AppendKeyWithMaxTimestamp(std::string* result, const Slice& key,
+                                      size_t ts_sz);
+
 // Attempt to parse an internal key from "internal_key".  On success,
 // stores the parsed data in "*result", and returns true.
 //
@@ -40,7 +40,7 @@ class DeleteFileTest : public DBTestBase {
        wal_dir_(dbname_ + "/wal_files") {}

  void SetOptions(Options* options) {
-    assert(options);
+    ASSERT_NE(options, nullptr);
    options->delete_obsolete_files_period_micros = 0;  // always do full purge
    options->enable_thread_tracking = true;
    options->write_buffer_size = 1024 * 1024 * 1000;
@@ -105,7 +105,7 @@ class DeleteFileTest : public DBTestBase {
  void CheckFileTypeCounts(const std::string& dir, int required_log,
                           int required_sst, int required_manifest) {
    std::vector<std::string> filenames;
-    env_->GetChildren(dir, &filenames);
+    ASSERT_OK(env_->GetChildren(dir, &filenames));

    int log_cnt = 0, sst_cnt = 0, manifest_cnt = 0;
    for (auto file : filenames) {
@@ -180,7 +180,8 @@ TEST_F(DeleteFileTest, AddKeysAndQueryLevels) {
  ASSERT_TRUE(status.IsInvalidArgument());

  // Lowest level file deletion should succeed.
-  ASSERT_OK(db_->DeleteFile(level2file));
+  status = db_->DeleteFile(level2file);
+  ASSERT_OK(status);
 }

 TEST_F(DeleteFileTest, PurgeObsoleteFilesTest) {
@@ -201,7 +202,7 @@ TEST_F(DeleteFileTest, PurgeObsoleteFilesTest) {
  compact_options.change_level = true;
  compact_options.target_level = 2;
  Slice first_slice(first), last_slice(last);
-  db_->CompactRange(compact_options, &first_slice, &last_slice);
+  ASSERT_OK(db_->CompactRange(compact_options, &first_slice, &last_slice));
  // 1 sst after compaction
  CheckFileTypeCounts(dbname_, 0, 1, 1);

@@ -210,7 +211,9 @@ TEST_F(DeleteFileTest, PurgeObsoleteFilesTest) {
  Iterator *itr = nullptr;
  CreateTwoLevels();
  itr = db_->NewIterator(ReadOptions());
-  db_->CompactRange(compact_options, &first_slice, &last_slice);
+  ASSERT_OK(itr->status());
+  ASSERT_OK(db_->CompactRange(compact_options, &first_slice, &last_slice));
+  ASSERT_OK(itr->status());
  // 3 sst after compaction with live iterator
  CheckFileTypeCounts(dbname_, 0, 3, 1);
  delete itr;
@@ -237,7 +240,8 @@ TEST_F(DeleteFileTest, BackgroundPurgeIteratorTest) {
  ReadOptions read_options;
  read_options.background_purge_on_iterator_cleanup = true;
  itr = db_->NewIterator(read_options);
-  db_->CompactRange(compact_options, &first_slice, &last_slice);
+  ASSERT_OK(itr->status());
+  ASSERT_OK(db_->CompactRange(compact_options, &first_slice, &last_slice));
  // 3 sst after compaction with live iterator
  CheckFileTypeCounts(dbname_, 0, 3, 1);
  test::SleepingBackgroundTask sleeping_task_before;
@@ -344,11 +348,12 @@ TEST_F(DeleteFileTest, BackgroundPurgeCopyOptions) {
    ReadOptions read_options;
    read_options.background_purge_on_iterator_cleanup = true;
    itr = db_->NewIterator(read_options);
+    ASSERT_OK(itr->status());
    // ReadOptions is deleted, but iterator cleanup function should not be
    // affected
  }

-  db_->CompactRange(compact_options, &first_slice, &last_slice);
+  ASSERT_OK(db_->CompactRange(compact_options, &first_slice, &last_slice));
  // 3 sst after compaction with live iterator
  CheckFileTypeCounts(dbname_, 0, 3, 1);
  delete itr;
@@ -382,9 +387,11 @@ TEST_F(DeleteFileTest, BackgroundPurgeTestMultipleJobs) {
  ReadOptions read_options;
  read_options.background_purge_on_iterator_cleanup = true;
  Iterator* itr1 = db_->NewIterator(read_options);
+  ASSERT_OK(itr1->status());
  CreateTwoLevels();
  Iterator* itr2 = db_->NewIterator(read_options);
-  db_->CompactRange(compact_options, &first_slice, &last_slice);
+  ASSERT_OK(itr2->status());
+  ASSERT_OK(db_->CompactRange(compact_options, &first_slice, &last_slice));
  // 5 sst files after 2 compactions with 2 live iterators
  CheckFileTypeCounts(dbname_, 0, 5, 1);

@@ -417,6 +424,7 @@ TEST_F(DeleteFileTest, DeleteFileWithIterator) {
  CreateTwoLevels();
  ReadOptions read_options;
  Iterator* it = db_->NewIterator(read_options);
+  ASSERT_OK(it->status());
  std::vector<LiveFileMetaData> metadata;
  db_->GetLiveFilesMetaData(&metadata);

@@ -432,7 +440,7 @@ TEST_F(DeleteFileTest, DeleteFileWithIterator) {
  Status status = db_->DeleteFile(level2file);
  fprintf(stdout, "Deletion status %s: %s\n",
          level2file.c_str(), status.ToString().c_str());
-  ASSERT_TRUE(status.ok());
+  ASSERT_OK(status);
  it->SeekToFirst();
  int numKeysIterated = 0;
  while(it->Valid()) {
@@ -452,7 +460,7 @@ TEST_F(DeleteFileTest, DeleteLogFiles) {

  AddKeys(10, 0);
  VectorLogPtr logfiles;
-  db_->GetSortedWalFiles(logfiles);
+  ASSERT_OK(db_->GetSortedWalFiles(logfiles));
  ASSERT_GT(logfiles.size(), 0UL);
  // Take the last log file which is expected to be alive and try to delete it
  // Should not succeed because live logs are not allowed to be deleted
@@ -461,7 +469,7 @@ TEST_F(DeleteFileTest, DeleteLogFiles) {
  ASSERT_OK(env_->FileExists(wal_dir_ + "/" + alive_log->PathName()));
  fprintf(stdout, "Deleting alive log file %s\n",
          alive_log->PathName().c_str());
-  ASSERT_TRUE(!db_->DeleteFile(alive_log->PathName()).ok());
+  ASSERT_NOK(db_->DeleteFile(alive_log->PathName()));
  ASSERT_OK(env_->FileExists(wal_dir_ + "/" + alive_log->PathName()));
  logfiles.clear();

@@ -469,10 +477,10 @@ TEST_F(DeleteFileTest, DeleteLogFiles) {
  // Call Flush again to flush out memtable and move alive log to archived log
  // and try to delete the archived log file
  FlushOptions fopts;
-  db_->Flush(fopts);
+  ASSERT_OK(db_->Flush(fopts));
  AddKeys(10, 0);
-  db_->Flush(fopts);
-  db_->GetSortedWalFiles(logfiles);
+  ASSERT_OK(db_->Flush(fopts));
+  ASSERT_OK(db_->GetSortedWalFiles(logfiles));
  ASSERT_GT(logfiles.size(), 0UL);
  std::unique_ptr<LogFile> archived_log = std::move(logfiles.front());
  ASSERT_EQ(archived_log->Type(), kArchivedLogFile);
@@ -480,8 +488,8 @@ TEST_F(DeleteFileTest, DeleteLogFiles) {
  fprintf(stdout, "Deleting archived log file %s\n",
          archived_log->PathName().c_str());
  ASSERT_OK(db_->DeleteFile(archived_log->PathName()));
-  ASSERT_EQ(Status::NotFound(),
-            env_->FileExists(wal_dir_ + "/" + archived_log->PathName()));
+  ASSERT_TRUE(
+      env_->FileExists(wal_dir_ + "/" + archived_log->PathName()).IsNotFound());
 }

 TEST_F(DeleteFileTest, DeleteNonDefaultColumnFamily) {
@@ -520,6 +528,7 @@ TEST_F(DeleteFileTest, DeleteNonDefaultColumnFamily) {

  {
    std::unique_ptr<Iterator> itr(db_->NewIterator(ReadOptions(), handles_[1]));
+    ASSERT_OK(itr->status());
    int count = 0;
    for (itr->SeekToFirst(); itr->Valid(); itr->Next()) {
      ASSERT_OK(itr->status());
@@ -111,6 +111,23 @@ std::map<std::tuple<BackgroundErrorReason, Status::Code, Status::SubCode, bool>,
                         Status::Code::kIOError, Status::SubCode::kIOFenced,
                         false),
         Status::Severity::kFatalError},
+        // Errors during MANIFEST write when WAL is disabled
+        {std::make_tuple(BackgroundErrorReason::kManifestWriteNoWAL,
+                         Status::Code::kIOError, Status::SubCode::kNoSpace,
+                         true),
+         Status::Severity::kHardError},
+        {std::make_tuple(BackgroundErrorReason::kManifestWriteNoWAL,
+                         Status::Code::kIOError, Status::SubCode::kNoSpace,
+                         false),
+         Status::Severity::kHardError},
+        {std::make_tuple(BackgroundErrorReason::kManifestWriteNoWAL,
+                         Status::Code::kIOError, Status::SubCode::kIOFenced,
+                         true),
+         Status::Severity::kFatalError},
+        {std::make_tuple(BackgroundErrorReason::kManifestWriteNoWAL,
+                         Status::Code::kIOError, Status::SubCode::kIOFenced,
+                         false),
+         Status::Severity::kFatalError},

 };

@@ -175,6 +192,12 @@ std::map<std::tuple<BackgroundErrorReason, Status::Code, bool>,
        {std::make_tuple(BackgroundErrorReason::kFlushNoWAL,
                         Status::Code::kIOError, false),
         Status::Severity::kNoError},
+        {std::make_tuple(BackgroundErrorReason::kManifestWriteNoWAL,
+                         Status::Code::kIOError, true),
+         Status::Severity::kFatalError},
+        {std::make_tuple(BackgroundErrorReason::kManifestWriteNoWAL,
+                         Status::Code::kIOError, false),
+         Status::Severity::kFatalError},
 };

 std::map<std::tuple<BackgroundErrorReason, bool>, Status::Severity>
@@ -244,10 +267,11 @@ void ErrorHandler::CancelErrorRecovery() {
 // This can also get called as part of a recovery operation. In that case, we
 // also track the error separately in recovery_error_ so we can tell in the
 // end whether recovery succeeded or not
-Status ErrorHandler::SetBGError(const Status& bg_err, BackgroundErrorReason reason) {
+const Status& ErrorHandler::SetBGError(const Status& bg_err,
+                                       BackgroundErrorReason reason) {
  db_mutex_->AssertHeld();
  if (bg_err.ok()) {
-    return Status::OK();
+    return bg_err;
  }

  bool paranoid = db_options_.paranoid_checks;
@@ -324,11 +348,27 @@ Status ErrorHandler::SetBGError(const Status& bg_err, BackgroundErrorReason reas
  return bg_error_;
 }

-Status ErrorHandler::SetBGError(const IOStatus& bg_io_err,
-                                BackgroundErrorReason reason) {
+// This is the main function for looking at IO related error during the
+// background operations. The main logic is:
+// 1) if the error is caused by data loss, the error is mapped to
+//    unrecoverable error. Application/user must take action to handle
+//    this situation.
+// 2) if the error is a Retryable IO error, auto resume will be called and the
+//    auto resume can be controlled by resume count and resume interval
+//    options. There are three sub-cases:
+//    a) if the error happens during compaction, it is mapped to a soft error.
+//       the compaction thread will reschedule a new compaction.
+//    b) if the error happens during flush and also WAL is empty, it is mapped
+//       to a soft error. Note that, it includes the case that IO error happens
+//       in SST or manifest write during flush.
+//    c) all other errors are mapped to hard error.
+// 3) for other cases, SetBGError(const Status& bg_err, BackgroundErrorReason
+//    reason) will be called to handle other error cases.
+const Status& ErrorHandler::SetBGError(const IOStatus& bg_io_err,
+                                       BackgroundErrorReason reason) {
  db_mutex_->AssertHeld();
  if (bg_io_err.ok()) {
-    return Status::OK();
+    return bg_io_err;
  }
  ROCKS_LOG_WARN(db_options_.info_log, "Background IO error %s",
                 bg_io_err.ToString().c_str());
@@ -336,13 +376,13 @@ Status ErrorHandler::SetBGError(const IOStatus& bg_io_err,
  if (recovery_in_prog_ && recovery_io_error_.ok()) {
    recovery_io_error_ = bg_io_err;
  }
-  if (BackgroundErrorReason::kManifestWrite == reason) {
+  if (BackgroundErrorReason::kManifestWrite == reason ||
+      BackgroundErrorReason::kManifestWriteNoWAL == reason) {
    // Always returns ok
    db_->DisableFileDeletionsWithLock().PermitUncheckedError();
  }

  Status new_bg_io_err = bg_io_err;
-  Status s;
  DBRecoverContext context;
  if (bg_io_err.GetDataLoss()) {
    // First, data loss is treated as unrecoverable error. So it can directly
@@ -353,8 +393,8 @@ Status ErrorHandler::SetBGError(const IOStatus& bg_io_err,
    if (recovery_in_prog_ && recovery_error_.ok()) {
      recovery_error_ = bg_err;
    }
-    EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason, &s,
-                                          db_mutex_, &auto_recovery);
+    EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason,
+                                          &bg_err, db_mutex_, &auto_recovery);
    recover_context_ = context;
    return bg_error_;
  } else if (bg_io_err.GetRetryable()) {
@@ -365,8 +405,9 @@ Status ErrorHandler::SetBGError(const IOStatus& bg_io_err,
    // soft error. In other cases, treat the retryable IO error as hard
    // error.
    bool auto_recovery = false;
-    EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason, &s,
-                                          db_mutex_, &auto_recovery);
+    EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason,
+                                          &new_bg_io_err, db_mutex_,
+                                          &auto_recovery);
    if (BackgroundErrorReason::kCompaction == reason) {
      Status bg_err(new_bg_io_err, Status::Severity::kSoftError);
      if (bg_err.severity() > bg_error_.severity()) {
@@ -374,7 +415,8 @@ Status ErrorHandler::SetBGError(const IOStatus& bg_io_err,
      }
      recover_context_ = context;
      return bg_error_;
-    } else if (BackgroundErrorReason::kFlushNoWAL == reason) {
+    } else if (BackgroundErrorReason::kFlushNoWAL == reason ||
+               BackgroundErrorReason::kManifestWriteNoWAL == reason) {
      // When the BG Retryable IO error reason is flush without WAL,
      // We map it to a soft error. At the same time, all the background work
      // should be stopped except the BG work from recovery. Therefore, we
@@ -405,12 +447,11 @@ Status ErrorHandler::SetBGError(const IOStatus& bg_io_err,
      return StartRecoverFromRetryableBGIOError(bg_io_err);
    }
  } else {
-    s = SetBGError(new_bg_io_err, reason);
+    return SetBGError(new_bg_io_err, reason);
  }
-  return s;
 }

-Status ErrorHandler::OverrideNoSpaceError(Status bg_error,
+Status ErrorHandler::OverrideNoSpaceError(const Status& bg_error,
                                          bool* auto_recovery) {
 #ifndef ROCKSDB_LITE
  if (bg_error.severity() >= Status::Severity::kFatalError) {
@@ -466,7 +507,11 @@ Status ErrorHandler::ClearBGError() {
  // Signal that recovery succeeded
  if (recovery_error_.ok()) {
    Status old_bg_error = bg_error_;
+    // Clear and check the recovery IO and BG error
    bg_error_ = Status::OK();
+    recovery_io_error_ = IOStatus::OK();
+    bg_error_.PermitUncheckedError();
+    recovery_io_error_.PermitUncheckedError();
    recovery_in_prog_ = false;
    soft_error_no_bg_work_ = false;
    EventHelpers::NotifyOnErrorRecoveryCompleted(db_options_.listeners,
@@ -516,6 +561,7 @@ Status ErrorHandler::RecoverFromBGError(bool is_manual) {
  // during the recovery process. While recovering, the only operations that
  // can generate background errors should be the flush operations
  recovery_error_ = Status::OK();
+  recovery_error_.PermitUncheckedError();
  Status s = db_->ResumeImpl(recover_context_);
  if (s.ok()) {
    soft_error_no_bg_work_ = false;
@@ -537,24 +583,34 @@ Status ErrorHandler::RecoverFromBGError(bool is_manual) {
 #endif
 }

-Status ErrorHandler::StartRecoverFromRetryableBGIOError(IOStatus io_error) {
+const Status& ErrorHandler::StartRecoverFromRetryableBGIOError(
+    const IOStatus& io_error) {
 #ifndef ROCKSDB_LITE
  db_mutex_->AssertHeld();
-  if (bg_error_.ok() || io_error.ok()) {
-    return Status::OK();
-  }
-  if (db_options_.max_bgerror_resume_count <= 0 || recovery_in_prog_ ||
-      recovery_thread_) {
+  if (bg_error_.ok()) {
+    return bg_error_;
+  } else if (io_error.ok()) {
+    return io_error;
+  } else if (db_options_.max_bgerror_resume_count <= 0 || recovery_in_prog_) {
    // Auto resume BG error is not enabled, directly return bg_error_.
    return bg_error_;
  }

+  if (recovery_thread_) {
+    // In this case, if recovery_in_prog_ is false, current thread should
+    // wait the previous recover thread to finish and create a new thread
+    // to recover from the bg error.
+    db_mutex_->Unlock();
+    recovery_thread_->join();
+    db_mutex_->Lock();
+  }
+
  recovery_in_prog_ = true;
  recovery_thread_.reset(
      new port::Thread(&ErrorHandler::RecoverFromRetryableBGIOError, this));

  if (recovery_io_error_.ok() && recovery_error_.ok()) {
-    return Status::OK();
+    return recovery_error_;
  } else {
    TEST_SYNC_POINT("StartRecoverRetryableBGIOError:RecoverFail");
    return bg_error_;
@@ -619,6 +675,7 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
        TEST_SYNC_POINT("RecoverFromRetryableBGIOError:RecoverSuccess");
        Status old_bg_error = bg_error_;
        bg_error_ = Status::OK();
+        bg_error_.PermitUncheckedError();
        EventHelpers::NotifyOnErrorRecoveryCompleted(db_options_.listeners,
                                                     old_bg_error, db_mutex_);
        recovery_in_prog_ = false;
@@ -31,17 +31,14 @@ class ErrorHandler {
                InstrumentedMutex* db_mutex)
       : db_(db),
         db_options_(db_options),
-         bg_error_(Status::OK()),
-         recovery_error_(Status::OK()),
-         recovery_io_error_(IOStatus::OK()),
         cv_(db_mutex),
         end_recovery_(false),
         recovery_thread_(nullptr),
         db_mutex_(db_mutex),
         auto_recovery_(false),
         recovery_in_prog_(false),
-         soft_error_no_bg_work_(false) {}
-   ~ErrorHandler() {
+         soft_error_no_bg_work_(false) {
+     // Clear the checked flag for uninitialized errors
     bg_error_.PermitUncheckedError();
     recovery_error_.PermitUncheckedError();
     recovery_io_error_.PermitUncheckedError();
@@ -53,13 +50,14 @@ class ErrorHandler {
                                     Status::Code code,
                                     Status::SubCode subcode);

-   Status SetBGError(const Status& bg_err, BackgroundErrorReason reason);
+   const Status& SetBGError(const Status& bg_err, BackgroundErrorReason reason);

-   Status SetBGError(const IOStatus& bg_io_err, BackgroundErrorReason reason);
+   const Status& SetBGError(const IOStatus& bg_io_err,
+                            BackgroundErrorReason reason);

-   Status GetBGError() { return bg_error_; }
+   Status GetBGError() const { return bg_error_; }

-   Status GetRecoveryError() { return recovery_error_; }
+   Status GetRecoveryError() const { return recovery_error_; }

   Status ClearBGError();

@@ -110,9 +108,9 @@ class ErrorHandler {
    // Used to store the context for recover, such as flush reason.
    DBRecoverContext recover_context_;

-    Status OverrideNoSpaceError(Status bg_error, bool* auto_recovery);
+    Status OverrideNoSpaceError(const Status& bg_error, bool* auto_recovery);
    void RecoverFromNoSpace();
-    Status StartRecoverFromRetryableBGIOError(IOStatus io_error);
+    const Status& StartRecoverFromRetryableBGIOError(const IOStatus& io_error);
    void RecoverFromRetryableBGIOError();
 };

@@ -173,7 +173,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWriteError) {
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  Reopen(options);
  ASSERT_EQ("val", Get(Key(0)));
@@ -271,7 +271,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError1) {
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_EQ("val2", Get(Key(2)));
  ASSERT_OK(Put(Key(3), "val3", wo));
@@ -314,7 +314,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError2) {
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_EQ("val2", Get(Key(2)));
  ASSERT_OK(Put(Key(3), "val3", wo));
@@ -357,7 +357,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError3) {
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_EQ("val2", Get(Key(2)));
  ASSERT_OK(Put(Key(3), "val3", wo));
@@ -399,7 +399,7 @@ TEST_F(DBErrorHandlingFSTest, ManifestWriteError) {
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  new_manifest = GetManifestNameFromLiveFiles();
  ASSERT_NE(new_manifest, old_manifest);
@@ -442,7 +442,52 @@ TEST_F(DBErrorHandlingFSTest, ManifestWriteRetryableError) {
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
+
+  new_manifest = GetManifestNameFromLiveFiles();
+  ASSERT_NE(new_manifest, old_manifest);
+
+  Reopen(options);
+  ASSERT_EQ("val", Get(Key(0)));
+  ASSERT_EQ("val", Get(Key(1)));
+  Close();
+}
+
+TEST_F(DBErrorHandlingFSTest, ManifestWriteNoWALRetryableError) {
+  std::shared_ptr<ErrorHandlerFSListener> listener(
+      new ErrorHandlerFSListener());
+  Options options = GetDefaultOptions();
+  options.env = fault_env_.get();
+  options.create_if_missing = true;
+  options.listeners.emplace_back(listener);
+  options.max_bgerror_resume_count = 0;
+  Status s;
+  std::string old_manifest;
+  std::string new_manifest;
+
+  listener->EnableAutoRecovery(false);
+  DestroyAndReopen(options);
+  old_manifest = GetManifestNameFromLiveFiles();
+
+  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
+  error_msg.SetRetryable(true);
+
+  WriteOptions wo = WriteOptions();
+  wo.disableWAL = true;
+  ASSERT_OK(Put(Key(0), "val", wo));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put(Key(1), "val", wo));
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionSet::LogAndApply:WriteManifest",
+      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
+  SyncPoint::GetInstance()->EnableProcessing();
+  s = Flush();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+  SyncPoint::GetInstance()->DisableProcessing();
+  fault_fs_->SetFilesystemActive(true);
+  s = dbfull()->Resume();
+  ASSERT_OK(s);

  new_manifest = GetManifestNameFromLiveFiles();
  ASSERT_NE(new_manifest, old_manifest);
@@ -490,7 +535,7 @@ TEST_F(DBErrorHandlingFSTest, DoubleManifestWriteError) {

  // A successful Resume() will create a new manifest file
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  new_manifest = GetManifestNameFromLiveFiles();
  ASSERT_NE(new_manifest, old_manifest);
@@ -523,7 +568,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionManifestWriteError) {
  ASSERT_OK(Put(Key(0), "val"));
  ASSERT_OK(Put(Key(2), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      // Wait for flush of 2nd L0 file before starting compaction
@@ -553,7 +598,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionManifestWriteError) {
  // This Flush will trigger a compaction, which will fail when appending to
  // the manifest
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  TEST_SYNC_POINT("CompactionManifestWriteError:0");
  // Clear all errors so when the compaction is retried, it will succeed
@@ -564,7 +609,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionManifestWriteError) {

  s = dbfull()->TEST_WaitForCompact();
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  new_manifest = GetManifestNameFromLiveFiles();
  ASSERT_NE(new_manifest, old_manifest);
@@ -597,7 +642,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionManifestWriteRetryableError) {
  ASSERT_OK(Put(Key(0), "val"));
  ASSERT_OK(Put(Key(2), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  listener->OverrideBGError(Status(error_msg, Status::Severity::kHardError));
  listener->EnableAutoRecovery(false);
@@ -623,7 +668,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionManifestWriteRetryableError) {

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  TEST_SYNC_POINT("CompactionManifestWriteError:0");
  TEST_SYNC_POINT("CompactionManifestWriteError:1");
@@ -635,7 +680,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionManifestWriteRetryableError) {
  SyncPoint::GetInstance()->ClearAllCallBacks();
  SyncPoint::GetInstance()->DisableProcessing();
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  new_manifest = GetManifestNameFromLiveFiles();
  ASSERT_NE(new_manifest, old_manifest);
@@ -661,7 +706,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteError) {
  ASSERT_OK(Put(Key(0), "va;"));
  ASSERT_OK(Put(Key(2), "va;"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  listener->OverrideBGError(
      Status(Status::NoSpace(), Status::Severity::kHardError));
@@ -678,14 +723,14 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteError) {

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  s = dbfull()->TEST_WaitForCompact();
  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);

  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  Destroy(options);
 }

@@ -707,7 +752,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableError) {
  ASSERT_OK(Put(Key(0), "va;"));
  ASSERT_OK(Put(Key(2), "va;"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  listener->OverrideBGError(Status(error_msg, Status::Severity::kHardError));
  listener->EnableAutoRecovery(false);
@@ -721,7 +766,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableError) {

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  s = dbfull()->TEST_WaitForCompact();
  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
@@ -730,7 +775,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableError) {
  SyncPoint::GetInstance()->ClearAllCallBacks();
  SyncPoint::GetInstance()->DisableProcessing();
  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  Destroy(options);
 }

@@ -745,7 +790,7 @@ TEST_F(DBErrorHandlingFSTest, CorruptionError) {
  ASSERT_OK(Put(Key(0), "va;"));
  ASSERT_OK(Put(Key(2), "va;"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
@@ -759,7 +804,7 @@ TEST_F(DBErrorHandlingFSTest, CorruptionError) {

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  s = dbfull()->TEST_WaitForCompact();
  ASSERT_EQ(s.severity(),
@@ -767,7 +812,7 @@ TEST_F(DBErrorHandlingFSTest, CorruptionError) {

  fault_fs_->SetFilesystemActive(true);
  s = dbfull()->Resume();
-  ASSERT_NE(s, Status::OK());
+  ASSERT_NOK(s);
  Destroy(options);
 }

@@ -799,7 +844,7 @@ TEST_F(DBErrorHandlingFSTest, AutoRecoverFlushError) {
  ASSERT_EQ(listener->WaitForRecovery(5000000), true);

  s = Put(Key(1), "val");
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  Reopen(options);
  ASSERT_EQ("val", Get(Key(0)));
@@ -829,7 +874,7 @@ TEST_F(DBErrorHandlingFSTest, FailRecoverFlushError) {
  // We should be able to shutdown the database while auto recovery is going
  // on in the background
  Close();
-  DestroyDB(dbname_, options);
+  DestroyDB(dbname_, options).PermitUncheckedError();
 }

 TEST_F(DBErrorHandlingFSTest, WALWriteError) {
@@ -859,7 +904,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteError) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  {
@@ -915,7 +960,6 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableError) {
  options.listeners.emplace_back(listener);
  options.paranoid_checks = true;
  options.max_bgerror_resume_count = 0;
-  Status s;
  Random rnd(301);

  DestroyAndReopen(options);
@@ -933,7 +977,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableError) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  // For the second batch, the first 2 file Append are successful, then the
@@ -956,8 +1000,8 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableError) {
    SyncPoint::GetInstance()->EnableProcessing();
    WriteOptions wopts;
    wopts.sync = true;
-    s = dbfull()->Write(wopts, &batch);
-    ASSERT_EQ(true, s.IsIOError());
+    Status s = dbfull()->Write(wopts, &batch);
+    ASSERT_TRUE(s.IsIOError());
  }
  fault_fs_->SetFilesystemActive(true);
  SyncPoint::GetInstance()->ClearAllCallBacks();
@@ -973,8 +1017,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableError) {
  }

  // Resume and write a new batch, should be in the WAL
-  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(dbfull()->Resume());
  {
    WriteBatch batch;

@@ -984,7 +1027,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableError) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  Reopen(options);
@@ -1010,7 +1053,6 @@ TEST_F(DBErrorHandlingFSTest, MultiCFWALWriteError) {
  options.create_if_missing = true;
  options.writable_file_max_buffer_size = 32768;
  options.listeners.emplace_back(listener);
-  Status s;
  Random rnd(301);

  listener->EnableAutoRecovery();
@@ -1027,7 +1069,7 @@ TEST_F(DBErrorHandlingFSTest, MultiCFWALWriteError) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  {
@@ -1050,8 +1092,8 @@ TEST_F(DBErrorHandlingFSTest, MultiCFWALWriteError) {
    SyncPoint::GetInstance()->EnableProcessing();
    WriteOptions wopts;
    wopts.sync = true;
-    s = dbfull()->Write(wopts, &batch);
-    ASSERT_EQ(s, s.NoSpace());
+    Status s = dbfull()->Write(wopts, &batch);
+    ASSERT_TRUE(s.IsNoSpace());
  }
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
@@ -1119,9 +1161,8 @@ TEST_F(DBErrorHandlingFSTest, MultiDBCompactionError) {
    listener[i]->InjectFileCreationError(fault_fs[i], 3,
                                         IOStatus::NoSpace("Out of space"));
    snprintf(buf, sizeof(buf), "_%d", i);
-    DestroyDB(dbname_ + std::string(buf), options[i]);
-    ASSERT_EQ(DB::Open(options[i], dbname_ + std::string(buf), &dbptr),
-              Status::OK());
+    ASSERT_OK(DestroyDB(dbname_ + std::string(buf), options[i]));
+    ASSERT_OK(DB::Open(options[i], dbname_ + std::string(buf), &dbptr));
    db.emplace_back(dbptr);
  }

@@ -1134,8 +1175,8 @@ TEST_F(DBErrorHandlingFSTest, MultiDBCompactionError) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
+    ASSERT_OK(db[i]->Write(wopts, &batch));
+    ASSERT_OK(db[i]->Flush(FlushOptions()));
  }

  def_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
@@ -1149,8 +1190,8 @@ TEST_F(DBErrorHandlingFSTest, MultiDBCompactionError) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
+    ASSERT_OK(db[i]->Write(wopts, &batch));
+    ASSERT_OK(db[i]->Flush(FlushOptions()));
  }

  for (auto i = 0; i < kNumDbInstances; ++i) {
@@ -1163,8 +1204,7 @@ TEST_F(DBErrorHandlingFSTest, MultiDBCompactionError) {
  for (auto i = 0; i < kNumDbInstances; ++i) {
    std::string prop;
    ASSERT_EQ(listener[i]->WaitForRecovery(5000000), true);
-    ASSERT_EQ(static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true),
-              Status::OK());
+    ASSERT_OK(static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true));
    EXPECT_TRUE(db[i]->GetProperty(
        "rocksdb.num-files-at-level" + NumberToString(0), &prop));
    EXPECT_EQ(atoi(prop.c_str()), 0);
@@ -1181,7 +1221,7 @@ TEST_F(DBErrorHandlingFSTest, MultiDBCompactionError) {
    if (getenv("KEEP_DB")) {
      printf("DB is still at %s%s\n", dbname_.c_str(), buf);
    } else {
-      Status s = DestroyDB(dbname_ + std::string(buf), options[i]);
+      ASSERT_OK(DestroyDB(dbname_ + std::string(buf), options[i]));
    }
  }
  options.clear();
@@ -1236,9 +1276,8 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {
        break;
    }
    snprintf(buf, sizeof(buf), "_%d", i);
-    DestroyDB(dbname_ + std::string(buf), options[i]);
-    ASSERT_EQ(DB::Open(options[i], dbname_ + std::string(buf), &dbptr),
-              Status::OK());
+    ASSERT_OK(DestroyDB(dbname_ + std::string(buf), options[i]));
+    ASSERT_OK(DB::Open(options[i], dbname_ + std::string(buf), &dbptr));
    db.emplace_back(dbptr);
  }

@@ -1251,8 +1290,8 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
+    ASSERT_OK(db[i]->Write(wopts, &batch));
+    ASSERT_OK(db[i]->Flush(FlushOptions()));
  }

  def_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
@@ -1266,11 +1305,11 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(db[i]->Write(wopts, &batch));
    if (i != 1) {
-      ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
+      ASSERT_OK(db[i]->Flush(FlushOptions()));
    } else {
-      ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::NoSpace());
+      ASSERT_TRUE(db[i]->Flush(FlushOptions()).IsNoSpace());
    }
  }

@@ -1284,7 +1323,7 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {
        ASSERT_EQ(s.severity(), Status::Severity::kHardError);
        break;
      case 2:
-        ASSERT_EQ(s, Status::OK());
+        ASSERT_OK(s);
        break;
    }
    fault_fs[i]->SetFilesystemActive(true);
@@ -1297,8 +1336,7 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {
      ASSERT_EQ(listener[i]->WaitForRecovery(5000000), true);
    }
    if (i == 1) {
-      ASSERT_EQ(static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true),
-                Status::OK());
+      ASSERT_OK(static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true));
    }
    EXPECT_TRUE(db[i]->GetProperty(
        "rocksdb.num-files-at-level" + NumberToString(0), &prop));
@@ -1316,7 +1354,7 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {
    if (getenv("KEEP_DB")) {
      printf("DB is still at %s%s\n", dbname_.c_str(), buf);
    } else {
-      DestroyDB(dbname_ + std::string(buf), options[i]);
+      EXPECT_OK(DestroyDB(dbname_ + std::string(buf), options[i]));
    }
  }
  options.clear();
@@ -1373,12 +1411,10 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover1) {
  ASSERT_EQ("val2", Get(Key(2)));

  // call auto resume
-  s = dbfull()->Resume();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(dbfull()->Resume());
  ASSERT_OK(Put(Key(3), "val3", wo));
-  s = Flush();
  // After resume is successful, the flush should be ok.
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Flush());
  ASSERT_EQ("val3", Get(Key(3)));
  Destroy(options);
 }
@@ -1468,8 +1504,7 @@ TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover1) {
  Reopen(options);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Flush());
  ASSERT_EQ("val2", Get(Key(2)));

  Destroy(options);
@@ -1509,8 +1544,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover2) {
  Reopen(options);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Flush());
  ASSERT_EQ("val2", Get(Key(2)));

  Destroy(options);
@@ -1557,10 +1591,9 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover3) {
  // resume manually here.
  s = dbfull()->Resume();
  ASSERT_EQ("val1", Get(Key(1)));
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Flush());
  ASSERT_EQ("val2", Get(Key(2)));

  Destroy(options);
@@ -1611,19 +1644,16 @@ TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover4) {
  // Even the FS is recoverd, due to the Fatal Error in bg_error_ the resume
  // and flush will all fail.
  ASSERT_EQ("val1", Get(Key(1)));
-  s = dbfull()->Resume();
-  ASSERT_NE(s, Status::OK());
+  ASSERT_NOK(dbfull()->Resume());
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_NE(s, Status::OK());
+  ASSERT_NOK(Flush());
  ASSERT_EQ("NOT_FOUND", Get(Key(2)));

  Reopen(options);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Flush());
  ASSERT_EQ("val2", Get(Key(2)));

  Destroy(options);
@@ -1663,7 +1693,7 @@ TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover5) {
  // The first resume will cause recovery_error and its severity is the
  // Fatal error
  s = dbfull()->Close();
-  ASSERT_NE(s, Status::OK());
+  ASSERT_NOK(s);
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
  SyncPoint::GetInstance()->DisableProcessing();
  fault_fs_->SetFilesystemActive(true);
@@ -1672,7 +1702,7 @@ TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover5) {
  ASSERT_NE("val1", Get(Key(1)));
  ASSERT_OK(Put(Key(2), "val2"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  ASSERT_EQ("val2", Get(Key(2)));

  Destroy(options);
@@ -1723,14 +1753,14 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover6) {
  // The first resume will cause recovery_error and its severity is the
  // Fatal error
  s = dbfull()->Close();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  SyncPoint::GetInstance()->DisableProcessing();

  Reopen(options);
  ASSERT_EQ("val1", Get(Key(1)));
  ASSERT_OK(Put(Key(2), "val2"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);
  ASSERT_EQ("val2", Get(Key(2)));

  Destroy(options);
@@ -1789,6 +1819,61 @@ TEST_F(DBErrorHandlingFSTest, ManifestWriteRetryableErrorAutoRecover) {
  Close();
 }

+TEST_F(DBErrorHandlingFSTest, ManifestWriteNoWALRetryableErrorAutoRecover) {
+  // Fail the first resume and let the second resume be successful
+  std::shared_ptr<ErrorHandlerFSListener> listener(
+      new ErrorHandlerFSListener());
+  Options options = GetDefaultOptions();
+  options.env = fault_env_.get();
+  options.create_if_missing = true;
+  options.listeners.emplace_back(listener);
+  options.max_bgerror_resume_count = 2;
+  options.bgerror_resume_retry_interval = 100000;  // 0.1 second
+  Status s;
+  std::string old_manifest;
+  std::string new_manifest;
+
+  listener->EnableAutoRecovery(false);
+  DestroyAndReopen(options);
+  old_manifest = GetManifestNameFromLiveFiles();
+
+  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
+  error_msg.SetRetryable(true);
+
+  WriteOptions wo = WriteOptions();
+  wo.disableWAL = true;
+  ASSERT_OK(Put(Key(0), "val", wo));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put(Key(1), "val", wo));
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"RecoverFromRetryableBGIOError:BeforeStart",
+        "ManifestWriteNoWALRetryableErrorAutoRecover:0"},
+       {"ManifestWriteNoWALRetryableErrorAutoRecover:1",
+        "RecoverFromRetryableBGIOError:BeforeWait1"},
+       {"RecoverFromRetryableBGIOError:RecoverSuccess",
+        "ManifestWriteNoWALRetryableErrorAutoRecover:2"}});
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionSet::LogAndApply:WriteManifest",
+      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
+  SyncPoint::GetInstance()->EnableProcessing();
+  s = Flush();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
+  TEST_SYNC_POINT("ManifestWriteNoWALRetryableErrorAutoRecover:0");
+  fault_fs_->SetFilesystemActive(true);
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
+  TEST_SYNC_POINT("ManifestWriteNoWALRetryableErrorAutoRecover:1");
+  TEST_SYNC_POINT("ManifestWriteNoWALRetryableErrorAutoRecover:2");
+  SyncPoint::GetInstance()->DisableProcessing();
+
+  new_manifest = GetManifestNameFromLiveFiles();
+  ASSERT_NE(new_manifest, old_manifest);
+
+  Reopen(options);
+  ASSERT_EQ("val", Get(Key(0)));
+  ASSERT_EQ("val", Get(Key(1)));
+  Close();
+}
+
 TEST_F(DBErrorHandlingFSTest,
       CompactionManifestWriteRetryableErrorAutoRecover) {
  std::shared_ptr<ErrorHandlerFSListener> listener(
@@ -1812,8 +1897,7 @@ TEST_F(DBErrorHandlingFSTest,

  ASSERT_OK(Put(Key(0), "val"));
  ASSERT_OK(Put(Key(2), "val"));
-  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Flush());

  listener->OverrideBGError(Status(error_msg, Status::Severity::kHardError));
  listener->EnableAutoRecovery(false);
@@ -1850,7 +1934,7 @@ TEST_F(DBErrorHandlingFSTest,

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  TEST_SYNC_POINT("CompactionManifestWriteErrorAR:0");
  TEST_SYNC_POINT("CompactionManifestWriteErrorAR:1");
@@ -1900,7 +1984,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableErrorAutoRecover) {
  ASSERT_OK(Put(Key(0), "va;"));
  ASSERT_OK(Put(Key(2), "va;"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  listener->OverrideBGError(Status(error_msg, Status::Severity::kHardError));
  listener->EnableAutoRecovery(false);
@@ -1925,7 +2009,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableErrorAutoRecover) {

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  s = dbfull()->TEST_WaitForCompact();
  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
@@ -1965,7 +2049,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableErrorAutoRecover1) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  // For the second batch, the first 2 file Append are successful, then the
@@ -2022,7 +2106,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableErrorAutoRecover1) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  Reopen(options);
@@ -2066,7 +2150,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableErrorAutoRecover2) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  // For the second batch, the first 2 file Append are successful, then the
@@ -2123,7 +2207,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableErrorAutoRecover2) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  Reopen(options);
@@ -2185,7 +2269,7 @@ TEST_P(DBErrorHandlingFencingTest, ManifestWriteFenced) {
  old_manifest = GetManifestNameFromLiveFiles();

  ASSERT_OK(Put(Key(0), "val"));
-  Flush();
+  ASSERT_OK(Flush());
  ASSERT_OK(Put(Key(1), "val"));
  SyncPoint::GetInstance()->SetCallBack(
      "VersionSet::LogAndApply:WriteManifest", [&](void*) {
@@ -2218,7 +2302,7 @@ TEST_P(DBErrorHandlingFencingTest, CompactionWriteFenced) {
  ASSERT_OK(Put(Key(0), "va;"));
  ASSERT_OK(Put(Key(2), "va;"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  listener->EnableAutoRecovery(true);
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
@@ -2232,7 +2316,7 @@ TEST_P(DBErrorHandlingFencingTest, CompactionWriteFenced) {

  ASSERT_OK(Put(Key(1), "val"));
  s = Flush();
-  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(s);

  s = dbfull()->TEST_WaitForCompact();
  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kFatalError);
@@ -2268,7 +2352,7 @@ TEST_P(DBErrorHandlingFencingTest, WALWriteFenced) {

    WriteOptions wopts;
    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
+    ASSERT_OK(dbfull()->Write(wopts, &batch));
  };

  {
@@ -29,8 +29,8 @@ class ExternalSSTFileBasicTest
  }

  void DestroyAndRecreateExternalSSTFilesDir() {
-    DestroyDir(env_, sst_files_dir_);
-    env_->CreateDir(sst_files_dir_);
+    ASSERT_OK(DestroyDir(env_, sst_files_dir_));
+    ASSERT_OK(env_->CreateDir(sst_files_dir_));
  }

  Status DeprecatedAddFile(const std::vector<std::string>& files,
@@ -162,7 +162,9 @@ class ExternalSSTFileBasicTest
        write_global_seqno, verify_checksums_before_ingest, true_data);
  }

-  ~ExternalSSTFileBasicTest() override { DestroyDir(env_, sst_files_dir_); }
+  ~ExternalSSTFileBasicTest() override {
+    DestroyDir(env_, sst_files_dir_).PermitUncheckedError();
+  }

 protected:
  std::string sst_files_dir_;
@@ -186,7 +188,7 @@ TEST_F(ExternalSSTFileBasicTest, Basic) {
  }
  ExternalSstFileInfo file1_info;
  Status s = sst_file_writer.Finish(&file1_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();

  // Current file size should be non-zero after success write.
  ASSERT_GT(sst_file_writer.FileSize(), 0);
@@ -202,14 +204,14 @@ TEST_F(ExternalSSTFileBasicTest, Basic) {
  ASSERT_EQ(file1_info.file_checksum_func_name, kUnknownFileChecksumFuncName);
  // sst_file_writer already finished, cannot add this value
  s = sst_file_writer.Put(Key(100), "bad_val");
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();
  s = sst_file_writer.DeleteRange(Key(100), Key(200));
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();

  DestroyAndReopen(options);
  // Add file using file path
  s = DeprecatedAddFile({file1});
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(db_->GetLatestSequenceNumber(), 0U);
  for (int k = 0; k < 100; k++) {
    ASSERT_EQ(Get(Key(k)), Key(k) + "_val");
@@ -286,7 +288,7 @@ TEST_F(ExternalSSTFileBasicTest, BasicWithFileChecksumCrc32c) {
  }
  ExternalSstFileInfo file1_info;
  Status s = sst_file_writer.Finish(&file1_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  std::string file_checksum, file_checksum_func_name;
  ASSERT_OK(checksum_helper.GetSingleFileChecksumAndFuncName(
      file1, &file_checksum, &file_checksum_func_name));
@@ -305,14 +307,14 @@ TEST_F(ExternalSSTFileBasicTest, BasicWithFileChecksumCrc32c) {
  ASSERT_EQ(file1_info.file_checksum_func_name, file_checksum_func_name);
  // sst_file_writer already finished, cannot add this value
  s = sst_file_writer.Put(Key(100), "bad_val");
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();
  s = sst_file_writer.DeleteRange(Key(100), Key(200));
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();

  DestroyAndReopen(options);
  // Add file using file path
  s = DeprecatedAddFile({file1});
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(db_->GetLatestSequenceNumber(), 0U);
  for (int k = 0; k < 100; k++) {
    ASSERT_EQ(Get(Key(k)), Key(k) + "_val");
@@ -338,7 +340,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  }
  ExternalSstFileInfo file1_info;
  Status s = sst_file_writer.Finish(&file1_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file1_info.file_path, file1);
  ASSERT_EQ(file1_info.num_entries, 100);
  ASSERT_EQ(file1_info.smallest_key, Key(1000));
@@ -357,7 +359,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  }
  ExternalSstFileInfo file2_info;
  s = sst_file_writer.Finish(&file2_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file2_info.file_path, file2);
  ASSERT_EQ(file2_info.num_entries, 200);
  ASSERT_EQ(file2_info.smallest_key, Key(1100));
@@ -376,7 +378,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  }
  ExternalSstFileInfo file3_info;
  s = sst_file_writer.Finish(&file3_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file3_info.file_path, file3);
  ASSERT_EQ(file3_info.num_entries, 200);
  ASSERT_EQ(file3_info.smallest_key, Key(1300));
@@ -395,7 +397,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  }
  ExternalSstFileInfo file4_info;
  s = sst_file_writer.Finish(&file4_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file4_info.file_path, file4);
  ASSERT_EQ(file4_info.num_entries, 300);
  ASSERT_EQ(file4_info.smallest_key, Key(1500));
@@ -414,7 +416,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  }
  ExternalSstFileInfo file5_info;
  s = sst_file_writer.Finish(&file5_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file5_info.file_path, file5);
  ASSERT_EQ(file5_info.num_entries, 200);
  ASSERT_EQ(file5_info.smallest_key, Key(1800));
@@ -433,7 +435,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  }
  ExternalSstFileInfo file6_info;
  s = sst_file_writer.Finish(&file6_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file6_info.file_path, file6);
  ASSERT_EQ(file6_info.num_entries, 200);
  ASSERT_EQ(file6_info.smallest_key, Key(2000));
@@ -447,7 +449,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  s = AddFileWithFileChecksum({file1}, {file_checksum1, "xyz"},
                              {file_checksum1}, true, false, false, false);
  // does not care the checksum input since db does not enable file checksum
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file1));
  std::vector<LiveFileMetaData> live_files;
  dbfull()->GetLiveFilesMetaData(&live_files);
@@ -465,26 +467,26 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  s = AddFileWithFileChecksum({file2}, {file_checksum2, "xyz"},
                              {file_checksum_func_name2}, true, false, false,
                              false);
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();

  // Enable verify_file_checksum option
  // The checksum name does not match, fail the ingestion
  s = AddFileWithFileChecksum({file2}, {file_checksum2}, {"xyz"}, true, false,
                              false, false);
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();

  // Enable verify_file_checksum option
  // The checksum itself does not match, fail the ingestion
  s = AddFileWithFileChecksum({file2}, {"xyz"}, {file_checksum_func_name2},
                              true, false, false, false);
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();

  // Enable verify_file_checksum option
  // All matches, ingestion is successful
  s = AddFileWithFileChecksum({file2}, {file_checksum2},
                              {file_checksum_func_name2}, true, false, false,
                              false);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  std::vector<LiveFileMetaData> live_files1;
  dbfull()->GetLiveFilesMetaData(&live_files1);
  for (auto f : live_files1) {
@@ -501,7 +503,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  std::vector<std::string> checksum, checksum_func;
  s = AddFileWithFileChecksum({file3}, checksum, checksum_func, true, false,
                              false, false);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  std::vector<LiveFileMetaData> live_files2;
  dbfull()->GetLiveFilesMetaData(&live_files2);
  for (auto f : live_files2) {
@@ -511,20 +513,20 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
      set1.insert(f.name);
    }
  }
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file3));

  // Does not enable verify_file_checksum options
  // The checksum name does not match, fail the ingestion
  s = AddFileWithFileChecksum({file4}, {file_checksum4}, {"xyz"}, false, false,
                              false, false);
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();

  // Does not enable verify_file_checksum options
  // Checksum function name matches, store the checksum being ingested.
  s = AddFileWithFileChecksum({file4}, {"asd"}, {file_checksum_func_name4},
                              false, false, false, false);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  std::vector<LiveFileMetaData> live_files3;
  dbfull()->GetLiveFilesMetaData(&live_files3);
  for (auto f : live_files3) {
@@ -535,7 +537,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
      set1.insert(f.name);
    }
  }
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file4));

  // enable verify_file_checksum options, DB enable checksum, and enable
@@ -544,8 +546,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  s = AddFileWithFileChecksum({file5}, {file_checksum5},
                              {file_checksum_func_name5}, true, false, false,
                              true);
-  ASSERT_OK(s);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  std::vector<LiveFileMetaData> live_files4;
  dbfull()->GetLiveFilesMetaData(&live_files4);
  for (auto f : live_files4) {
@@ -558,7 +559,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
      set1.insert(f.name);
    }
  }
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file5));

  // Does not enable verify_file_checksum options and also the ingested file
@@ -567,7 +568,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
  std::vector<std::string> files_c6, files_name6;
  s = AddFileWithFileChecksum({file6}, files_c6, files_name6, false, false,
                              false, false);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  std::vector<LiveFileMetaData> live_files6;
  dbfull()->GetLiveFilesMetaData(&live_files6);
  for (auto f : live_files6) {
@@ -577,7 +578,7 @@ TEST_F(ExternalSSTFileBasicTest, IngestFileWithFileChecksum) {
      set1.insert(f.name);
    }
  }
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file6));
 }

@@ -595,7 +596,7 @@ TEST_F(ExternalSSTFileBasicTest, NoCopy) {
  }
  ExternalSstFileInfo file1_info;
  Status s = sst_file_writer.Finish(&file1_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file1_info.file_path, file1);
  ASSERT_EQ(file1_info.num_entries, 100);
  ASSERT_EQ(file1_info.smallest_key, Key(0));
@@ -609,7 +610,7 @@ TEST_F(ExternalSSTFileBasicTest, NoCopy) {
  }
  ExternalSstFileInfo file2_info;
  s = sst_file_writer.Finish(&file2_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file2_info.file_path, file2);
  ASSERT_EQ(file2_info.num_entries, 200);
  ASSERT_EQ(file2_info.smallest_key, Key(100));
@@ -623,23 +624,23 @@ TEST_F(ExternalSSTFileBasicTest, NoCopy) {
  }
  ExternalSstFileInfo file3_info;
  s = sst_file_writer.Finish(&file3_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file3_info.file_path, file3);
  ASSERT_EQ(file3_info.num_entries, 15);
  ASSERT_EQ(file3_info.smallest_key, Key(110));
  ASSERT_EQ(file3_info.largest_key, Key(124));

  s = DeprecatedAddFile({file1}, true /* move file */);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(Status::NotFound(), env_->FileExists(file1));

  s = DeprecatedAddFile({file2}, false /* copy file */);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file2));

  // This file has overlapping values with the existing data
  s = DeprecatedAddFile({file3}, true /* move file */);
-  ASSERT_FALSE(s.ok()) << s.ToString();
+  ASSERT_NOK(s) << s.ToString();
  ASSERT_OK(env_->FileExists(file3));

  for (int k = 0; k < 300; k++) {
@@ -1126,7 +1127,7 @@ TEST_F(ExternalSSTFileBasicTest, SyncFailure) {
    if (i == 2) {
      ingest_opt.write_global_seqno = true;
    }
-    ASSERT_FALSE(db_->IngestExternalFile({file_name}, ingest_opt).ok());
+    ASSERT_NOK(db_->IngestExternalFile({file_name}, ingest_opt));
    db_->ReleaseSnapshot(snapshot);

    SyncPoint::GetInstance()->DisableProcessing();
@@ -1326,7 +1327,7 @@ TEST_F(ExternalSSTFileBasicTest, AdjacentRangeDeletionTombstones) {
  ASSERT_OK(sst_file_writer.DeleteRange(Key(300), Key(400)));
  ExternalSstFileInfo file8_info;
  Status s = sst_file_writer.Finish(&file8_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file8_info.file_path, file8);
  ASSERT_EQ(file8_info.num_entries, 0);
  ASSERT_EQ(file8_info.smallest_key, "");
@@ -1341,7 +1342,7 @@ TEST_F(ExternalSSTFileBasicTest, AdjacentRangeDeletionTombstones) {
  ASSERT_OK(sst_file_writer.DeleteRange(Key(400), Key(500)));
  ExternalSstFileInfo file9_info;
  s = sst_file_writer.Finish(&file9_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(file9_info.file_path, file9);
  ASSERT_EQ(file9_info.num_entries, 0);
  ASSERT_EQ(file9_info.smallest_key, "");
@@ -1353,7 +1354,7 @@ TEST_F(ExternalSSTFileBasicTest, AdjacentRangeDeletionTombstones) {
  // Range deletion tombstones are exclusive on their end key, so these SSTs
  // should not be considered as overlapping.
  s = DeprecatedAddFile({file8, file9});
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(s) << s.ToString();
  ASSERT_EQ(db_->GetLatestSequenceNumber(), 0U);
  DestroyAndRecreateExternalSSTFilesDir();
 }
@@ -201,7 +201,8 @@ Status ExternalSstFileIngestionJob::Prepare(
            requested_checksum_func_name, &generated_checksum,
            &generated_checksum_func_name,
            ingestion_options_.verify_checksums_readahead_size,
-            db_options_.allow_mmap_reads, io_tracer_);
+            db_options_.allow_mmap_reads, io_tracer_,
+            db_options_.rate_limiter.get());
        if (!io_s.ok()) {
          status = io_s;
          ROCKS_LOG_WARN(db_options_.info_log,
@@ -335,6 +336,12 @@ Status ExternalSstFileIngestionJob::Run() {
  // with the files we are ingesting
  bool need_flush = false;
  status = NeedsFlush(&need_flush, super_version);
+  if (!status.ok()) {
+    return status;
+  }
+  if (need_flush) {
+    return Status::TryAgain();
+  }
  assert(status.ok() && need_flush == false);
 #endif

@@ -850,7 +857,7 @@ IOStatus ExternalSstFileIngestionJob::GenerateChecksumForIngestedFile(
      db_options_.file_checksum_gen_factory.get(), requested_checksum_func_name,
      &file_checksum, &file_checksum_func_name,
      ingestion_options_.verify_checksums_readahead_size,
-      db_options_.allow_mmap_reads, io_tracer_);
+      db_options_.allow_mmap_reads, io_tracer_, db_options_.rate_limiter.get());
  if (!io_s.ok()) {
    return io_s;
  }
@@ -47,8 +47,8 @@ class ExternSSTFileLinkFailFallbackTest
      : DBTestBase("/external_sst_file_test", /*env_do_fsync=*/true),
        test_env_(new ExternalSSTTestEnv(env_, true)) {
    sst_files_dir_ = dbname_ + "/sst_files/";
-    DestroyDir(env_, sst_files_dir_);
-    env_->CreateDir(sst_files_dir_);
+    EXPECT_EQ(DestroyDir(env_, sst_files_dir_), Status::OK());
+    EXPECT_EQ(env_->CreateDir(sst_files_dir_), Status::OK());
    options_ = CurrentOptions();
    options_.disable_auto_compactions = true;
    options_.env = test_env_;
@@ -79,8 +79,8 @@ class ExternalSSTFileTest
  }

  void DestroyAndRecreateExternalSSTFilesDir() {
-    DestroyDir(env_, sst_files_dir_);
-    env_->CreateDir(sst_files_dir_);
+    ASSERT_OK(DestroyDir(env_, sst_files_dir_));
+    ASSERT_OK(env_->CreateDir(sst_files_dir_));
  }

  Status GenerateOneExternalFile(
@@ -116,7 +116,7 @@ class ExternalSSTFileTest
    for (const auto& entry : data) {
      s = sst_file_writer.Put(entry.first, entry.second);
      if (!s.ok()) {
-        sst_file_writer.Finish();
+        sst_file_writer.Finish().PermitUncheckedError();
        return s;
      }
    }
@@ -171,7 +171,7 @@ class ExternalSSTFileTest
    for (auto& entry : data) {
      s = sst_file_writer.Put(entry.first, entry.second);
      if (!s.ok()) {
-        sst_file_writer.Finish();
+        sst_file_writer.Finish().PermitUncheckedError();
        return s;
      }
    }
@@ -213,11 +213,10 @@ class ExternalSSTFileTest
    size_t num_cfs = column_families.size();
    assert(ifos.size() == num_cfs);
    assert(data.size() == num_cfs);
-    Status s;
    std::vector<IngestExternalFileArg> args(num_cfs);
    for (size_t i = 0; i != num_cfs; ++i) {
      std::string external_file_path;
-      s = GenerateOneExternalFile(
+      Status s = GenerateOneExternalFile(
          options, column_families[i], data[i], file_id, sort_data,
          &external_file_path,
          true_data.size() == num_cfs ? &true_data[i] : nullptr);
@@ -230,8 +229,7 @@ class ExternalSSTFileTest
      args[i].external_files.push_back(external_file_path);
      args[i].options = ifos[i];
    }
-    s = db_->IngestExternalFiles(args);
-    return s;
+    return db_->IngestExternalFiles(args);
  }

  Status GenerateAndAddExternalFile(
@@ -282,7 +280,9 @@ class ExternalSSTFileTest
    return db_->IngestExternalFile(files, opts);
  }

-  ~ExternalSSTFileTest() override { DestroyDir(env_, sst_files_dir_); }
+  ~ExternalSSTFileTest() override {
+    DestroyDir(env_, sst_files_dir_).PermitUncheckedError();
+  }

 protected:
  int last_file_id_ = 0;
@@ -305,8 +305,7 @@ TEST_F(ExternalSSTFileTest, Basic) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    ExternalSstFileInfo file1_info;
-    Status s = sst_file_writer.Finish(&file1_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file1_info));

    // Current file size should be non-zero after success write.
    ASSERT_GT(sst_file_writer.FileSize(), 0);
@@ -319,8 +318,7 @@ TEST_F(ExternalSSTFileTest, Basic) {
    ASSERT_EQ(file1_info.smallest_range_del_key, "");
    ASSERT_EQ(file1_info.largest_range_del_key, "");
    // sst_file_writer already finished, cannot add this value
-    s = sst_file_writer.Put(Key(100), "bad_val");
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(sst_file_writer.Put(Key(100), "bad_val"));

    // file2.sst (100 => 199)
    std::string file2 = sst_files_dir_ + "file2.sst";
@@ -329,11 +327,9 @@ TEST_F(ExternalSSTFileTest, Basic) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    // Cannot add this key because it's not after last added key
-    s = sst_file_writer.Put(Key(99), "bad_val");
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(sst_file_writer.Put(Key(99), "bad_val"));
    ExternalSstFileInfo file2_info;
-    s = sst_file_writer.Finish(&file2_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file2_info));
    ASSERT_EQ(file2_info.file_path, file2);
    ASSERT_EQ(file2_info.num_entries, 100);
    ASSERT_EQ(file2_info.smallest_key, Key(100));
@@ -347,9 +343,8 @@ TEST_F(ExternalSSTFileTest, Basic) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val_overlap"));
    }
    ExternalSstFileInfo file3_info;
-    s = sst_file_writer.Finish(&file3_info);
+    ASSERT_OK(sst_file_writer.Finish(&file3_info));

-    ASSERT_TRUE(s.ok()) << s.ToString();
    // Current file size should be non-zero after success finish.
    ASSERT_GT(sst_file_writer.FileSize(), 0);
    ASSERT_EQ(file3_info.file_path, file3);
@@ -365,8 +360,7 @@ TEST_F(ExternalSSTFileTest, Basic) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val_overlap"));
    }
    ExternalSstFileInfo file4_info;
-    s = sst_file_writer.Finish(&file4_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file4_info));
    ASSERT_EQ(file4_info.file_path, file4);
    ASSERT_EQ(file4_info.num_entries, 10);
    ASSERT_EQ(file4_info.smallest_key, Key(30));
@@ -379,8 +373,7 @@ TEST_F(ExternalSSTFileTest, Basic) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    ExternalSstFileInfo file5_info;
-    s = sst_file_writer.Finish(&file5_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file5_info));
    ASSERT_EQ(file5_info.file_path, file5);
    ASSERT_EQ(file5_info.num_entries, 100);
    ASSERT_EQ(file5_info.smallest_key, Key(400));
@@ -389,10 +382,9 @@ TEST_F(ExternalSSTFileTest, Basic) {
    // file6.sst (delete 400 => 500)
    std::string file6 = sst_files_dir_ + "file6.sst";
    ASSERT_OK(sst_file_writer.Open(file6));
-    sst_file_writer.DeleteRange(Key(400), Key(500));
+    ASSERT_OK(sst_file_writer.DeleteRange(Key(400), Key(500)));
    ExternalSstFileInfo file6_info;
-    s = sst_file_writer.Finish(&file6_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file6_info));
    ASSERT_EQ(file6_info.file_path, file6);
    ASSERT_EQ(file6_info.num_entries, 0);
    ASSERT_EQ(file6_info.smallest_key, "");
@@ -404,17 +396,16 @@ TEST_F(ExternalSSTFileTest, Basic) {
    // file7.sst (delete 500 => 570, put 520 => 599 divisible by 2)
    std::string file7 = sst_files_dir_ + "file7.sst";
    ASSERT_OK(sst_file_writer.Open(file7));
-    sst_file_writer.DeleteRange(Key(500), Key(550));
+    ASSERT_OK(sst_file_writer.DeleteRange(Key(500), Key(550)));
    for (int k = 520; k < 560; k += 2) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
-    sst_file_writer.DeleteRange(Key(525), Key(575));
+    ASSERT_OK(sst_file_writer.DeleteRange(Key(525), Key(575)));
    for (int k = 560; k < 600; k += 2) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    ExternalSstFileInfo file7_info;
-    s = sst_file_writer.Finish(&file7_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file7_info));
    ASSERT_EQ(file7_info.file_path, file7);
    ASSERT_EQ(file7_info.num_entries, 40);
    ASSERT_EQ(file7_info.smallest_key, Key(520));
@@ -426,10 +417,9 @@ TEST_F(ExternalSSTFileTest, Basic) {
    // file8.sst (delete 600 => 700)
    std::string file8 = sst_files_dir_ + "file8.sst";
    ASSERT_OK(sst_file_writer.Open(file8));
-    sst_file_writer.DeleteRange(Key(600), Key(700));
+    ASSERT_OK(sst_file_writer.DeleteRange(Key(600), Key(700)));
    ExternalSstFileInfo file8_info;
-    s = sst_file_writer.Finish(&file8_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file8_info));
    ASSERT_EQ(file8_info.file_path, file8);
    ASSERT_EQ(file8_info.num_entries, 0);
    ASSERT_EQ(file8_info.smallest_key, "");
@@ -441,13 +431,11 @@ TEST_F(ExternalSSTFileTest, Basic) {
    // Cannot create an empty sst file
    std::string file_empty = sst_files_dir_ + "file_empty.sst";
    ExternalSstFileInfo file_empty_info;
-    s = sst_file_writer.Finish(&file_empty_info);
-    ASSERT_NOK(s);
+    ASSERT_NOK(sst_file_writer.Finish(&file_empty_info));

    DestroyAndReopen(options);
    // Add file using file path
-    s = DeprecatedAddFile({file1});
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(DeprecatedAddFile({file1}));
    ASSERT_EQ(db_->GetLatestSequenceNumber(), 0U);
    for (int k = 0; k < 100; k++) {
      ASSERT_EQ(Get(Key(k)), Key(k) + "_val");
@@ -468,12 +456,10 @@ TEST_F(ExternalSSTFileTest, Basic) {
    }

    // This file has overlapping values with the existing data
-    s = DeprecatedAddFile({file3});
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile({file3}));

    // This file has overlapping values with the existing data
-    s = DeprecatedAddFile({file4});
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile({file4}));

    // Overwrite values of keys divisible by 5
    for (int k = 0; k < 200; k += 5) {
@@ -485,8 +471,7 @@ TEST_F(ExternalSSTFileTest, Basic) {
    ASSERT_OK(DeprecatedAddFile({file5}));

    // This file has overlapping values with the existing data
-    s = DeprecatedAddFile({file6});
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile({file6}));

    // Key range of file7 (500 => 598) don't overlap with any keys in DB
    ASSERT_OK(DeprecatedAddFile({file7}));
@@ -614,15 +599,13 @@ TEST_F(ExternalSSTFileTest, AddList) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    ExternalSstFileInfo file1_info;
-    Status s = sst_file_writer.Finish(&file1_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file1_info));
    ASSERT_EQ(file1_info.file_path, file1);
    ASSERT_EQ(file1_info.num_entries, 100);
    ASSERT_EQ(file1_info.smallest_key, Key(0));
    ASSERT_EQ(file1_info.largest_key, Key(99));
    // sst_file_writer already finished, cannot add this value
-    s = sst_file_writer.Put(Key(100), "bad_val");
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(sst_file_writer.Put(Key(100), "bad_val"));

    // file2.sst (100 => 199)
    std::string file2 = sst_files_dir_ + "file2.sst";
@@ -631,11 +614,9 @@ TEST_F(ExternalSSTFileTest, AddList) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    // Cannot add this key because it's not after last added key
-    s = sst_file_writer.Put(Key(99), "bad_val");
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(sst_file_writer.Put(Key(99), "bad_val"));
    ExternalSstFileInfo file2_info;
-    s = sst_file_writer.Finish(&file2_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file2_info));
    ASSERT_EQ(file2_info.file_path, file2);
    ASSERT_EQ(file2_info.num_entries, 100);
    ASSERT_EQ(file2_info.smallest_key, Key(100));
@@ -649,8 +630,7 @@ TEST_F(ExternalSSTFileTest, AddList) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val_overlap"));
    }
    ExternalSstFileInfo file3_info;
-    s = sst_file_writer.Finish(&file3_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file3_info));
    ASSERT_EQ(file3_info.file_path, file3);
    ASSERT_EQ(file3_info.num_entries, 5);
    ASSERT_EQ(file3_info.smallest_key, Key(195));
@@ -664,8 +644,7 @@ TEST_F(ExternalSSTFileTest, AddList) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val_overlap"));
    }
    ExternalSstFileInfo file4_info;
-    s = sst_file_writer.Finish(&file4_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file4_info));
    ASSERT_EQ(file4_info.file_path, file4);
    ASSERT_EQ(file4_info.num_entries, 10);
    ASSERT_EQ(file4_info.smallest_key, Key(30));
@@ -678,8 +657,7 @@ TEST_F(ExternalSSTFileTest, AddList) {
      ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
    }
    ExternalSstFileInfo file5_info;
-    s = sst_file_writer.Finish(&file5_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file5_info));
    ASSERT_EQ(file5_info.file_path, file5);
    ASSERT_EQ(file5_info.num_entries, 100);
    ASSERT_EQ(file5_info.smallest_key, Key(200));
@@ -691,8 +669,7 @@ TEST_F(ExternalSSTFileTest, AddList) {
    ASSERT_OK(sst_file_writer.DeleteRange(Key(0), Key(75)));
    ASSERT_OK(sst_file_writer.DeleteRange(Key(25), Key(100)));
    ExternalSstFileInfo file6_info;
-    s = sst_file_writer.Finish(&file6_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file6_info));
    ASSERT_EQ(file6_info.file_path, file6);
    ASSERT_EQ(file6_info.num_entries, 0);
    ASSERT_EQ(file6_info.smallest_key, "");
@@ -706,8 +683,7 @@ TEST_F(ExternalSSTFileTest, AddList) {
    ASSERT_OK(sst_file_writer.Open(file7));
    ASSERT_OK(sst_file_writer.DeleteRange(Key(99), Key(201)));
    ExternalSstFileInfo file7_info;
-    s = sst_file_writer.Finish(&file7_info);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(sst_file_writer.Finish(&file7_info));
    ASSERT_EQ(file7_info.file_path, file7);
    ASSERT_EQ(file7_info.num_entries, 0);
    ASSERT_EQ(file7_info.smallest_key, "");
@@ -727,17 +703,13 @@ TEST_F(ExternalSSTFileTest, AddList) {
    DestroyAndReopen(options);

    // These lists of files have key ranges that overlap with each other
-    s = DeprecatedAddFile(file_list1);
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile(file_list1));
    // Both of the following overlap on the range deletion tombstone.
-    s = DeprecatedAddFile(file_list4);
-    ASSERT_FALSE(s.ok()) << s.ToString();
-    s = DeprecatedAddFile(file_list5);
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile(file_list4));
+    ASSERT_NOK(DeprecatedAddFile(file_list5));

    // Add files using file path list
-    s = DeprecatedAddFile(file_list0);
-    ASSERT_TRUE(s.ok()) << s.ToString();
+    ASSERT_OK(DeprecatedAddFile(file_list0));
    ASSERT_EQ(db_->GetLatestSequenceNumber(), 0U);
    for (int k = 0; k < 200; k++) {
      ASSERT_EQ(Get(Key(k)), Key(k) + "_val");
@@ -778,8 +750,7 @@ TEST_F(ExternalSSTFileTest, AddList) {
    }

    // This file list has overlapping values with the existing data
-    s = DeprecatedAddFile(file_list3);
-    ASSERT_FALSE(s.ok()) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile(file_list3));

    // Overwrite values of keys divisible by 5
    for (int k = 0; k < 200; k += 5) {
@@ -847,16 +818,14 @@ TEST_F(ExternalSSTFileTest, AddListAtomicity) {
      for (int k = i * 100; k < (i + 1) * 100; k++) {
        ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
      }
-      Status s = sst_file_writer.Finish(&files_info[i]);
-      ASSERT_TRUE(s.ok()) << s.ToString();
+      ASSERT_OK(sst_file_writer.Finish(&files_info[i]));
      ASSERT_EQ(files_info[i].file_path, files[i]);
      ASSERT_EQ(files_info[i].num_entries, 100);
      ASSERT_EQ(files_info[i].smallest_key, Key(i * 100));
      ASSERT_EQ(files_info[i].largest_key, Key((i + 1) * 100 - 1));
    }
    files.push_back(sst_files_dir_ + "file" + std::to_string(n) + ".sst");
-    auto s = DeprecatedAddFile(files);
-    ASSERT_NOK(s) << s.ToString();
+    ASSERT_NOK(DeprecatedAddFile(files));
    for (int k = 0; k < n * 100; k++) {
      ASSERT_EQ("NOT_FOUND", Get(Key(k)));
    }
@@ -878,17 +847,14 @@ TEST_F(ExternalSSTFileTest, PurgeObsoleteFilesBug) {

  // file1.sst (0 => 500)
  std::string sst_file_path = sst_files_dir_ + "file1.sst";
-  Status s = sst_file_writer.Open(sst_file_path);
-  ASSERT_OK(s);
+  ASSERT_OK(sst_file_writer.Open(sst_file_path));
  for (int i = 0; i < 500; i++) {
    std::string k = Key(i);
-    s = sst_file_writer.Put(k, k + "_val");
-    ASSERT_OK(s);
+    ASSERT_OK(sst_file_writer.Put(k, k + "_val"));
  }

  ExternalSstFileInfo sst_file_info;
-  s = sst_file_writer.Finish(&sst_file_info);
-  ASSERT_OK(s);
+  ASSERT_OK(sst_file_writer.Finish(&sst_file_info));

  options.delete_obsolete_files_period_micros = 0;
  options.disable_auto_compactions = true;
@@ -900,12 +866,11 @@ TEST_F(ExternalSSTFileTest, PurgeObsoleteFilesBug) {
        ASSERT_OK(Flush());
        ASSERT_OK(Put("aaa", "xxx"));
        ASSERT_OK(Flush());
-        db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+        ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
      });
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

-  s = DeprecatedAddFile({sst_file_path});
-  ASSERT_OK(s);
+  ASSERT_OK(DeprecatedAddFile({sst_file_path}));

  for (int i = 0; i < 500; i++) {
    std::string k = Key(i);
@@ -928,8 +893,7 @@ TEST_F(ExternalSSTFileTest, SkipSnapshot) {
    ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
  }
  ExternalSstFileInfo file1_info;
-  Status s = sst_file_writer.Finish(&file1_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(sst_file_writer.Finish(&file1_info));
  ASSERT_EQ(file1_info.file_path, file1);
  ASSERT_EQ(file1_info.num_entries, 100);
  ASSERT_EQ(file1_info.smallest_key, Key(0));
@@ -942,8 +906,7 @@ TEST_F(ExternalSSTFileTest, SkipSnapshot) {
    ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
  }
  ExternalSstFileInfo file2_info;
-  s = sst_file_writer.Finish(&file2_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(sst_file_writer.Finish(&file2_info));
  ASSERT_EQ(file2_info.file_path, file2);
  ASSERT_EQ(file2_info.num_entries, 200);
  ASSERT_EQ(file2_info.smallest_key, Key(100));
@@ -972,8 +935,7 @@ TEST_F(ExternalSSTFileTest, SkipSnapshot) {
    ASSERT_OK(sst_file_writer.Put(Key(k), Key(k) + "_val"));
  }
  ExternalSstFileInfo file3_info;
-  s = sst_file_writer.Finish(&file3_info);
-  ASSERT_TRUE(s.ok()) << s.ToString();
+  ASSERT_OK(sst_file_writer.Finish(&file3_info));
  ASSERT_EQ(file3_info.file_path, file3);
  ASSERT_EQ(file3_info.num_entries, 100);
  ASSERT_EQ(file3_info.smallest_key, Key(300));
@@ -1019,8 +981,7 @@ TEST_F(ExternalSSTFileTest, MultiThreaded) {
        ASSERT_OK(sst_file_writer.Put(Key(k), Key(k)));
      }

-      Status s = sst_file_writer.Finish();
-      ASSERT_TRUE(s.ok()) << s.ToString();
+      ASSERT_OK(sst_file_writer.Finish());
    };
    // Write num_files files in parallel
    std::vector<port::Thread> sst_writer_threads;
@@ -1082,8 +1043,7 @@ TEST_F(ExternalSSTFileTest, MultiThreaded) {
    // Overwrite values of keys divisible by 100
    for (int k = 0; k < num_files * keys_per_file; k += 100) {
      std::string key = Key(k);
-      Status s = Put(key, key + "_new");
-      ASSERT_TRUE(s.ok());
+      ASSERT_OK(Put(key, key + "_new"));
    }

    for (int i = 0; i < 2; i++) {
@@ -1167,7 +1127,8 @@ TEST_F(ExternalSSTFileTest, OverlappingRanges) {

        // Generate the file containing the range
        std::string file_name = sst_files_dir_ + env_->GenerateUniqueId();
-        ASSERT_OK(sst_file_writer.Open(file_name));
+        s = sst_file_writer.Open(file_name);
+        ASSERT_OK(s);
        for (int k = range_start; k <= range_end; k++) {
          s = sst_file_writer.Put(Key(k), range_val);
          ASSERT_OK(s);
@@ -1212,10 +1173,10 @@ TEST_F(ExternalSSTFileTest, OverlappingRanges) {

      // Flush / Compact the DB
      if (i && i % 50 == 0) {
-        Flush();
+        ASSERT_OK(Flush());
      }
      if (i && i % 75 == 0) {
-        db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+        ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
      }
    }

@@ -1293,7 +1254,7 @@ TEST_P(ExternalSSTFileTest, PickedLevel) {
  // Hold compaction from finishing
  TEST_SYNC_POINT("ExternalSSTFileTest::PickedLevel:2");

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  EXPECT_EQ(FilesPerLevel(), "1,1,1,2");

  size_t kcnt = 0;
@@ -1345,15 +1306,18 @@ TEST_F(ExternalSSTFileTest, PickedLevelBug) {
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  // While writing the MANIFEST start a thread that will ask for compaction
+  Status bg_compact_status;
  ROCKSDB_NAMESPACE::port::Thread bg_compact([&]() {
-    ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
+    bg_compact_status =
+        db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
  });
  TEST_SYNC_POINT("ExternalSSTFileTest::PickedLevelBug:2");

  // Start a thread that will ingest a new file
+  Status bg_addfile_status;
  ROCKSDB_NAMESPACE::port::Thread bg_addfile([&]() {
    file_keys = {1, 2, 3};
-    ASSERT_OK(GenerateAndAddExternalFile(options, file_keys, 1));
+    bg_addfile_status = GenerateAndAddExternalFile(options, file_keys, 1);
  });

  // Wait for AddFile to start picking levels and writing MANIFEST
@@ -1374,7 +1338,10 @@ TEST_F(ExternalSSTFileTest, PickedLevelBug) {
  bg_addfile.join();
  bg_compact.join();

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(bg_addfile_status);
+  ASSERT_OK(bg_compact_status);
+
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  int total_keys = 0;
  Iterator* iter = db_->NewIterator(ReadOptions());
@@ -1411,7 +1378,7 @@ TEST_F(ExternalSSTFileTest, IngestNonExistingFile) {

  // After full compaction, there should be only 1 file.
  std::vector<std::string> files;
-  env_->GetChildren(dbname_, &files);
+  ASSERT_OK(env_->GetChildren(dbname_, &files));
  int num_sst_files = 0;
  for (auto& f : files) {
    uint64_t number;
@@ -1533,7 +1500,7 @@ TEST_F(ExternalSSTFileTest, PickedLevelDynamic) {
  TEST_SYNC_POINT("ExternalSSTFileTest::PickedLevelDynamic:2");

  // Output of the compaction will go to L3
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  EXPECT_EQ(FilesPerLevel(), "1,0,0,2");

  Close();
@@ -1675,7 +1642,7 @@ TEST_F(ExternalSSTFileTest, AddFileTrivialMoveBug) {
  cro.exclusive_manual_compaction = false;
  ASSERT_OK(db_->CompactRange(cro, nullptr, nullptr));

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
 }
@@ -1727,9 +1694,9 @@ TEST_F(ExternalSSTFileTest, WithUnorderedWrite) {
  Options options = CurrentOptions();
  options.unordered_write = true;
  DestroyAndReopen(options);
-  Put("foo", "v1");
+  ASSERT_OK(Put("foo", "v1"));
  SyncPoint::GetInstance()->EnableProcessing();
-  port::Thread writer([&]() { Put("bar", "v2"); });
+  port::Thread writer([&]() { ASSERT_OK(Put("bar", "v2")); });

  TEST_SYNC_POINT("ExternalSSTFileTest::WithUnorderedWrite:WaitWriteWAL");
  ASSERT_OK(GenerateAndAddExternalFile(options, {{"bar", "v3"}}, -1,
@@ -1778,7 +1745,7 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoRandomized) {
    }
    size_t kcnt = 0;
    VerifyDBFromMap(true_data, &kcnt, false);
-    db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+    ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
    VerifyDBFromMap(true_data, &kcnt, false);
  }
 }
@@ -1862,8 +1829,8 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {
    ASSERT_OK(Put(Key(k), "memtable"));
    true_data[Key(k)] = "memtable";
  }
-  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
-                      &entries_in_memtable);
+  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
+                                  &entries_in_memtable));
  ASSERT_GE(entries_in_memtable, 1);

  bool write_global_seqno = std::get<0>(GetParam());
@@ -1872,40 +1839,40 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {90, 100, 110}, -1, true, write_global_seqno,
      verify_checksums_before_ingest, false, false, &true_data));
-  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
-                      &entries_in_memtable);
+  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
+                                  &entries_in_memtable));
  ASSERT_GE(entries_in_memtable, 1);

  // This file will flush the memtable
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {19, 20, 21}, -1, true, write_global_seqno,
      verify_checksums_before_ingest, false, false, &true_data));
-  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
-                      &entries_in_memtable);
+  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
+                                  &entries_in_memtable));
  ASSERT_EQ(entries_in_memtable, 0);

  for (int k : {200, 201, 205, 206}) {
    ASSERT_OK(Put(Key(k), "memtable"));
    true_data[Key(k)] = "memtable";
  }
-  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
-                      &entries_in_memtable);
+  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
+                                  &entries_in_memtable));
  ASSERT_GE(entries_in_memtable, 1);

  // No need for flush, this file keys fit between the memtable keys
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {202, 203, 204}, -1, true, write_global_seqno,
      verify_checksums_before_ingest, false, false, &true_data));
-  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
-                      &entries_in_memtable);
+  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
+                                  &entries_in_memtable));
  ASSERT_GE(entries_in_memtable, 1);

  // This file will flush the memtable
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {206, 207}, -1, true, write_global_seqno,
      verify_checksums_before_ingest, false, false, &true_data));
-  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
-                      &entries_in_memtable);
+  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
+                                  &entries_in_memtable));
  ASSERT_EQ(entries_in_memtable, 0);

  size_t kcnt = 0;
@@ -2303,7 +2270,7 @@ TEST_P(ExternalSSTFileTest, IngestBehind) {
    ASSERT_OK(Put(Key(i), "memtable"));
    true_data[Key(i)] = "memtable";
  }
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  // Universal picker should go at second from the bottom level
  ASSERT_EQ("0,1", FilesPerLevel());
  ASSERT_OK(GenerateAndAddExternalFile(
@@ -2317,7 +2284,7 @@ TEST_P(ExternalSSTFileTest, IngestBehind) {
      verify_checksums_before_ingest, true /*ingest_behind*/,
      false /*sort_data*/, &true_data));
  ASSERT_EQ("0,1,1", FilesPerLevel());
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  // bottom level should be empty
  ASSERT_EQ("0,1", FilesPerLevel());

@@ -2465,9 +2432,8 @@ TEST_P(ExternalSSTFileTest, IngestFilesIntoMultipleColumnFamilies_Success) {
  // Resize the true_data vector upon construction to avoid re-alloc
  std::vector<std::map<std::string, std::string>> true_data(
      column_families.size());
-  Status s = GenerateAndAddExternalFiles(options, column_families, ifos, data,
-                                         -1, true, true_data);
-  ASSERT_OK(s);
+  ASSERT_OK(GenerateAndAddExternalFiles(options, column_families, ifos, data,
+                                        -1, true, true_data));
  Close();
  ReopenWithColumnFamilies({kDefaultColumnFamilyName, "pikachu", "eevee"},
                           options);
@@ -2648,9 +2614,8 @@ TEST_P(ExternalSSTFileTest, IngestFilesIntoMultipleColumnFamilies_PrepareFail) {
  std::vector<std::map<std::string, std::string>> true_data(
      column_families.size());
  port::Thread ingest_thread([&]() {
-    Status s = GenerateAndAddExternalFiles(options, column_families, ifos, data,
-                                           -1, true, true_data);
-    ASSERT_NOK(s);
+    ASSERT_NOK(GenerateAndAddExternalFiles(options, column_families, ifos, data,
+                                           -1, true, true_data));
  });
  TEST_SYNC_POINT(
      "ExternalSSTFileTest::IngestFilesIntoMultipleColumnFamilies_PrepareFail:"
@@ -2718,9 +2683,8 @@ TEST_P(ExternalSSTFileTest, IngestFilesIntoMultipleColumnFamilies_CommitFail) {
  std::vector<std::map<std::string, std::string>> true_data(
      column_families.size());
  port::Thread ingest_thread([&]() {
-    Status s = GenerateAndAddExternalFiles(options, column_families, ifos, data,
-                                           -1, true, true_data);
-    ASSERT_NOK(s);
+    ASSERT_NOK(GenerateAndAddExternalFiles(options, column_families, ifos, data,
+                                           -1, true, true_data));
  });
  TEST_SYNC_POINT(
      "ExternalSSTFileTest::IngestFilesIntoMultipleColumnFamilies_CommitFail:"
@@ -2793,9 +2757,8 @@ TEST_P(ExternalSSTFileTest,
  std::vector<std::map<std::string, std::string>> true_data(
      column_families.size());
  port::Thread ingest_thread([&]() {
-    Status s = GenerateAndAddExternalFiles(options, column_families, ifos, data,
-                                           -1, true, true_data);
-    ASSERT_NOK(s);
+    ASSERT_NOK(GenerateAndAddExternalFiles(options, column_families, ifos, data,
+                                           -1, true, true_data));
  });
  TEST_SYNC_POINT(
      "ExternalSSTFileTest::IngestFilesIntoMultipleColumnFamilies_"
@@ -2806,7 +2769,7 @@ TEST_P(ExternalSSTFileTest,
      "PartialManifestWriteFail:1");
  ingest_thread.join();

-  fault_injection_env->DropUnsyncedFileData();
+  ASSERT_OK(fault_injection_env->DropUnsyncedFileData());
  fault_injection_env->SetFilesystemActive(true);
  Close();
  ReopenWithColumnFamilies({kDefaultColumnFamilyName, "pikachu", "eevee"},
@@ -2841,7 +2804,7 @@ TEST_P(ExternalSSTFileTest, IngestFilesTriggerFlushingWithTwoWriteQueue) {
  // sure that it won't enter the 2nd writer queue for the second time.
  std::vector<std::pair<std::string, std::string>> data;
  data.push_back(std::make_pair("1001", "v2"));
-  GenerateAndAddExternalFile(options, data);
+  ASSERT_OK(GenerateAndAddExternalFile(options, data, -1, true));
 }

 TEST_P(ExternalSSTFileTest, DeltaEncodingWhileGlobalSeqnoPresent) {
@@ -192,7 +192,7 @@ class FaultInjectionTest
    for (int i = start_idx; i < start_idx + num_vals; i++) {
      Slice key = Key(i, &key_space);
      batch.Clear();
-      batch.Put(key, Value(i, &value_space));
+      ASSERT_OK(batch.Put(key, Value(i, &value_space)));
      ASSERT_OK(db_->Write(write_options, &batch));
    }
  }
@@ -272,12 +272,12 @@ class FaultInjectionTest
    for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
      ASSERT_OK(db_->Delete(WriteOptions(), iter->key()));
    }
-
+    ASSERT_OK(iter->status());
    delete iter;

    FlushOptions flush_options;
    flush_options.wait = true;
-    db_->Flush(flush_options);
+    ASSERT_OK(db_->Flush(flush_options));
  }

  // rnd cannot be null for kResetDropRandomUnsyncedData
@@ -310,7 +310,7 @@ class FaultInjectionTest

    Build(write_options, 0, num_pre_sync);
    if (sync_use_compact_) {
-      db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+      ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
    }
    write_options.sync = false;
    Build(write_options, num_pre_sync, num_post_sync);
@@ -342,7 +342,7 @@ class FaultInjectionTest
  }

  void WaitCompactionFinish() {
-    static_cast<DBImpl*>(db_->GetRootDB())->TEST_WaitForCompact();
+    ASSERT_OK(static_cast<DBImpl*>(db_->GetRootDB())->TEST_WaitForCompact());
    ASSERT_OK(db_->Put(WriteOptions(), "", ""));
  }
 };
@@ -409,7 +409,7 @@ TEST_P(FaultInjectionTest, WriteOptionSyncTest) {
  write_options.sync = true;
  ASSERT_OK(
      db_->Put(write_options, Key(2, &key_space), Value(2, &value_space)));
-  db_->FlushWAL(false);
+  ASSERT_OK(db_->FlushWAL(false));

  env_->SetFilesystemActive(false);
  NoWriteTestReopenWithFault(kResetDropAndDeleteUnsynced);
@@ -450,7 +450,7 @@ TEST_P(FaultInjectionTest, UninstalledCompaction) {
  Build(WriteOptions(), 0, kNumKeys);
  FlushOptions flush_options;
  flush_options.wait = true;
-  db_->Flush(flush_options);
+  ASSERT_OK(db_->Flush(flush_options));
  ASSERT_OK(db_->Put(WriteOptions(), "", ""));
  TEST_SYNC_POINT("FaultInjectionTest::FaultTest:0");
  TEST_SYNC_POINT("FaultInjectionTest::FaultTest:1");
@@ -521,9 +521,9 @@ TEST_P(FaultInjectionTest, WriteBatchWalTerminationTest) {
  wo.sync = true;
  wo.disableWAL = false;
  WriteBatch batch;
-  batch.Put("cats", "dogs");
+  ASSERT_OK(batch.Put("cats", "dogs"));
  batch.MarkWalTerminationPoint();
-  batch.Put("boys", "girls");
+  ASSERT_OK(batch.Put("boys", "girls"));
  ASSERT_OK(db_->Write(wo, &batch));

  env_->SetFilesystemActive(false);
@@ -80,22 +80,24 @@ const char* GetFlushReasonString (FlushReason flush_reason) {
  }
 }

-FlushJob::FlushJob(
-    const std::string& dbname, ColumnFamilyData* cfd,
-    const ImmutableDBOptions& db_options,
-    const MutableCFOptions& mutable_cf_options, const uint64_t* max_memtable_id,
-    const FileOptions& file_options, VersionSet* versions,
-    InstrumentedMutex* db_mutex, std::atomic<bool>* shutting_down,
-    std::vector<SequenceNumber> existing_snapshots,
-    SequenceNumber earliest_write_conflict_snapshot,
-    SnapshotChecker* snapshot_checker, JobContext* job_context,
-    LogBuffer* log_buffer, FSDirectory* db_directory,
-    FSDirectory* output_file_directory, CompressionType output_compression,
-    Statistics* stats, EventLogger* event_logger, bool measure_io_stats,
-    const bool sync_output_directory, const bool write_manifest,
-    Env::Priority thread_pri, const std::shared_ptr<IOTracer>& io_tracer,
-    const std::string& db_id, const std::string& db_session_id,
-    std::string full_history_ts_low)
+FlushJob::FlushJob(const std::string& dbname, ColumnFamilyData* cfd,
+                   const ImmutableDBOptions& db_options,
+                   const MutableCFOptions& mutable_cf_options,
+                   uint64_t max_memtable_id, const FileOptions& file_options,
+                   VersionSet* versions, InstrumentedMutex* db_mutex,
+                   std::atomic<bool>* shutting_down,
+                   std::vector<SequenceNumber> existing_snapshots,
+                   SequenceNumber earliest_write_conflict_snapshot,
+                   SnapshotChecker* snapshot_checker, JobContext* job_context,
+                   LogBuffer* log_buffer, FSDirectory* db_directory,
+                   FSDirectory* output_file_directory,
+                   CompressionType output_compression, Statistics* stats,
+                   EventLogger* event_logger, bool measure_io_stats,
+                   const bool sync_output_directory, const bool write_manifest,
+                   Env::Priority thread_pri,
+                   const std::shared_ptr<IOTracer>& io_tracer,
+                   const std::string& db_id, const std::string& db_session_id,
+                   std::string full_history_ts_low)
    : dbname_(dbname),
      db_id_(db_id),
      db_session_id_(db_session_id),
@@ -60,10 +60,9 @@ class FlushJob {
  // IMPORTANT: mutable_cf_options needs to be alive while FlushJob is alive
  FlushJob(const std::string& dbname, ColumnFamilyData* cfd,
           const ImmutableDBOptions& db_options,
-           const MutableCFOptions& mutable_cf_options,
-           const uint64_t* max_memtable_id, const FileOptions& file_options,
-           VersionSet* versions, InstrumentedMutex* db_mutex,
-           std::atomic<bool>* shutting_down,
+           const MutableCFOptions& mutable_cf_options, uint64_t max_memtable_id,
+           const FileOptions& file_options, VersionSet* versions,
+           InstrumentedMutex* db_mutex, std::atomic<bool>* shutting_down,
           std::vector<SequenceNumber> existing_snapshots,
           SequenceNumber earliest_write_conflict_snapshot,
           SnapshotChecker* snapshot_checker, JobContext* job_context,
@@ -110,12 +109,11 @@ class FlushJob {
  ColumnFamilyData* cfd_;
  const ImmutableDBOptions& db_options_;
  const MutableCFOptions& mutable_cf_options_;
-  // Pointer to a variable storing the largest memtable id to flush in this
+  // A variable storing the largest memtable id to flush in this
  // flush job. RocksDB uses this variable to select the memtables to flush in
  // this job. All memtables in this column family with an ID smaller than or
-  // equal to *max_memtable_id_ will be selected for flush. If null, then all
-  // memtables in the column family will be selected.
-  const uint64_t* max_memtable_id_;
+  // equal to max_memtable_id_ will be selected for flush.
+  uint64_t max_memtable_id_;
  const FileOptions file_options_;
  VersionSet* versions_;
  InstrumentedMutex* db_mutex_;
@@ -52,7 +52,7 @@ class FlushJobTestBase : public testing::Test {
  }

  void NewDB() {
-    SetIdentityFile(env_, dbname_);
+    ASSERT_OK(SetIdentityFile(env_, dbname_));
    VersionEdit new_db;

    new_db.SetLogNumber(0);
@@ -85,6 +85,7 @@ class FlushJobTestBase : public testing::Test {
      std::string record;
      new_db.EncodeTo(&record);
      s = log.AddRecord(record);
+      ASSERT_OK(s);

      for (const auto& e : new_cfs) {
        record.clear();
@@ -158,7 +159,7 @@ TEST_F(FlushJobTest, Empty) {
  SnapshotChecker* snapshot_checker = nullptr;  // not relavant
  FlushJob flush_job(
      dbname_, versions_->GetColumnFamilySet()->GetDefault(), db_options_,
-      *cfd->GetLatestMutableCFOptions(), nullptr /* memtable_id */,
+      *cfd->GetLatestMutableCFOptions(), port::kMaxUint64 /* memtable_id */,
      env_options_, versions_.get(), &mutex_, &shutting_down_, {},
      kMaxSequenceNumber, snapshot_checker, &job_context, nullptr, nullptr,
      nullptr, kNoCompression, nullptr, &event_logger, false,
@@ -187,7 +188,7 @@ TEST_F(FlushJobTest, NonEmpty) {
  for (int i = 1; i < 10000; ++i) {
    std::string key(ToString((i + 1000) % 10000));
    std::string value("value" + key);
-    new_mem->Add(SequenceNumber(i), kTypeValue, key, value);
+    ASSERT_OK(new_mem->Add(SequenceNumber(i), kTypeValue, key, value));
    if ((i + 1000) % 10000 < 9995) {
      InternalKey internal_key(key, SequenceNumber(i), kTypeValue);
      inserted_keys.push_back({internal_key.Encode().ToString(), value});
@@ -195,7 +196,8 @@ TEST_F(FlushJobTest, NonEmpty) {
  }

  {
-    new_mem->Add(SequenceNumber(10000), kTypeRangeDeletion, "9995", "9999a");
+    ASSERT_OK(new_mem->Add(SequenceNumber(10000), kTypeRangeDeletion, "9995",
+                           "9999a"));
    InternalKey internal_key("9995", SequenceNumber(10000), kTypeRangeDeletion);
    inserted_keys.push_back({internal_key.Encode().ToString(), "9999a"});
  }
@@ -222,7 +224,7 @@ TEST_F(FlushJobTest, NonEmpty) {
    }

    const SequenceNumber seq(i + 10001);
-    new_mem->Add(seq, kTypeBlobIndex, key, blob_index);
+    ASSERT_OK(new_mem->Add(seq, kTypeBlobIndex, key, blob_index));

    InternalKey internal_key(key, seq, kTypeBlobIndex);
    inserted_keys.push_back({internal_key.Encode().ToString(), blob_index});
@@ -239,7 +241,7 @@ TEST_F(FlushJobTest, NonEmpty) {
  SnapshotChecker* snapshot_checker = nullptr;  // not relavant
  FlushJob flush_job(
      dbname_, versions_->GetColumnFamilySet()->GetDefault(), db_options_,
-      *cfd->GetLatestMutableCFOptions(), nullptr /* memtable_id */,
+      *cfd->GetLatestMutableCFOptions(), port::kMaxUint64 /* memtable_id */,
      env_options_, versions_.get(), &mutex_, &shutting_down_, {},
      kMaxSequenceNumber, snapshot_checker, &job_context, nullptr, nullptr,
      nullptr, kNoCompression, db_options_.statistics.get(), &event_logger,
@@ -283,8 +285,8 @@ TEST_F(FlushJobTest, FlushMemTablesSingleColumnFamily) {
    for (size_t j = 0; j < num_keys_per_table; ++j) {
      std::string key(ToString(j + i * num_keys_per_table));
      std::string value("value" + key);
-      mem->Add(SequenceNumber(j + i * num_keys_per_table), kTypeValue, key,
-               value);
+      ASSERT_OK(mem->Add(SequenceNumber(j + i * num_keys_per_table), kTypeValue,
+                         key, value));
    }
  }

@@ -301,7 +303,7 @@ TEST_F(FlushJobTest, FlushMemTablesSingleColumnFamily) {
  uint64_t flush_memtable_id = smallest_memtable_id + num_mems_to_flush - 1;
  FlushJob flush_job(
      dbname_, versions_->GetColumnFamilySet()->GetDefault(), db_options_,
-      *cfd->GetLatestMutableCFOptions(), &flush_memtable_id, env_options_,
+      *cfd->GetLatestMutableCFOptions(), flush_memtable_id, env_options_,
      versions_.get(), &mutex_, &shutting_down_, {}, kMaxSequenceNumber,
      snapshot_checker, &job_context, nullptr, nullptr, nullptr, kNoCompression,
      db_options_.statistics.get(), &event_logger, true,
@@ -356,7 +358,7 @@ TEST_F(FlushJobTest, FlushMemtablesMultipleColumnFamilies) {
      for (size_t j = 0; j != num_keys_per_memtable; ++j) {
        std::string key(ToString(j + i * num_keys_per_memtable));
        std::string value("value" + key);
-        mem->Add(curr_seqno++, kTypeValue, key, value);
+        ASSERT_OK(mem->Add(curr_seqno++, kTypeValue, key, value));
      }

      cfd->imm()->Add(mem, &to_delete);
@@ -373,7 +375,7 @@ TEST_F(FlushJobTest, FlushMemtablesMultipleColumnFamilies) {
    std::vector<SequenceNumber> snapshot_seqs;
    flush_jobs.emplace_back(new FlushJob(
        dbname_, cfd, db_options_, *cfd->GetLatestMutableCFOptions(),
-        &memtable_ids[k], env_options_, versions_.get(), &mutex_,
+        memtable_ids[k], env_options_, versions_.get(), &mutex_,
        &shutting_down_, snapshot_seqs, kMaxSequenceNumber, snapshot_checker,
        &job_context, nullptr, nullptr, nullptr, kNoCompression,
        db_options_.statistics.get(), &event_logger, true,
@@ -411,8 +413,9 @@ TEST_F(FlushJobTest, FlushMemtablesMultipleColumnFamilies) {

  Status s = InstallMemtableAtomicFlushResults(
      nullptr /* imm_lists */, all_cfds, mutable_cf_options_list, mems_list,
-      versions_.get(), &mutex_, file_meta_ptrs, &job_context.memtables_to_free,
-      nullptr /* db_directory */, nullptr /* log_buffer */);
+      versions_.get(), nullptr /* prep_tracker */, &mutex_, file_meta_ptrs,
+      &job_context.memtables_to_free, nullptr /* db_directory */,
+      nullptr /* log_buffer */);
  ASSERT_OK(s);

  mutex_.Unlock();
@@ -466,7 +469,7 @@ TEST_F(FlushJobTest, Snapshots) {
    for (int j = 0; j < insertions; ++j) {
      std::string value(rnd.HumanReadableString(10));
      auto seqno = ++current_seqno;
-      new_mem->Add(SequenceNumber(seqno), kTypeValue, key, value);
+      ASSERT_OK(new_mem->Add(SequenceNumber(seqno), kTypeValue, key, value));
      // a key is visible only if:
      // 1. it's the last one written (j == insertions - 1)
      // 2. there's a snapshot pointing at it
@@ -490,7 +493,7 @@ TEST_F(FlushJobTest, Snapshots) {
  SnapshotChecker* snapshot_checker = nullptr;  // not relavant
  FlushJob flush_job(
      dbname_, versions_->GetColumnFamilySet()->GetDefault(), db_options_,
-      *cfd->GetLatestMutableCFOptions(), nullptr /* memtable_id */,
+      *cfd->GetLatestMutableCFOptions(), port::kMaxUint64 /* memtable_id */,
      env_options_, versions_.get(), &mutex_, &shutting_down_, snapshots,
      kMaxSequenceNumber, snapshot_checker, &job_context, nullptr, nullptr,
      nullptr, kNoCompression, db_options_.statistics.get(), &event_logger,
@@ -518,7 +521,7 @@ class FlushJobTimestampTest : public FlushJobTestBase {
                             Slice value) {
    std::string key_str(std::move(key));
    PutFixed64(&key_str, ts);
-    memtable->Add(seq, value_type, key_str, value);
+    ASSERT_OK(memtable->Add(seq, value_type, key_str, value));
  }

 protected:
@@ -557,8 +560,8 @@ TEST_F(FlushJobTimestampTest, AllKeysExpired) {
  PutFixed64(&full_history_ts_low, std::numeric_limits<uint64_t>::max());
  FlushJob flush_job(
      dbname_, cfd, db_options_, *cfd->GetLatestMutableCFOptions(),
-      nullptr /* memtable_id */, env_options_, versions_.get(), &mutex_,
-      &shutting_down_, snapshots, kMaxSequenceNumber, snapshot_checker,
+      port::kMaxUint64 /* memtable_id */, env_options_, versions_.get(),
+      &mutex_, &shutting_down_, snapshots, kMaxSequenceNumber, snapshot_checker,
      &job_context, nullptr, nullptr, nullptr, kNoCompression,
      db_options_.statistics.get(), &event_logger, true,
      true /* sync_output_directory */, true /* write_manifest */,
@@ -608,8 +611,8 @@ TEST_F(FlushJobTimestampTest, NoKeyExpired) {
  PutFixed64(&full_history_ts_low, 0);
  FlushJob flush_job(
      dbname_, cfd, db_options_, *cfd->GetLatestMutableCFOptions(),
-      nullptr /* memtable_id */, env_options_, versions_.get(), &mutex_,
-      &shutting_down_, snapshots, kMaxSequenceNumber, snapshot_checker,
+      port::kMaxUint64 /* memtable_id */, env_options_, versions_.get(),
+      &mutex_, &shutting_down_, snapshots, kMaxSequenceNumber, snapshot_checker,
      &job_context, nullptr, nullptr, nullptr, kNoCompression,
      db_options_.statistics.get(), &event_logger, true,
      true /* sync_output_directory */, true /* write_manifest */,
@@ -46,7 +46,9 @@ class ForwardLevelIterator : public InternalIterator {
        file_iter_(nullptr),
        pinned_iters_mgr_(nullptr),
        prefix_extractor_(prefix_extractor),
-        allow_unprepared_value_(allow_unprepared_value) {}
+        allow_unprepared_value_(allow_unprepared_value) {
+    status_.PermitUncheckedError();  // Allow uninitialized status through
+  }

  ~ForwardLevelIterator() override {
    // Reset current pointer
@@ -238,6 +240,12 @@ ForwardIterator::ForwardIterator(DBImpl* db, const ReadOptions& read_options,
  if (sv_) {
    RebuildIterators(false);
  }
+
+  // immutable_status_ is a local aggregation of the
+  // status of the immutable Iterators.
+  // We have to PermitUncheckedError in case it is never
+  // used, otherwise it will fail ASSERT_STATUS_CHECKED.
+  immutable_status_.PermitUncheckedError();
 }

 ForwardIterator::~ForwardIterator() {
@@ -104,9 +104,9 @@ TEST_F(ImportColumnFamilyTest, ImportSSTFileWriterFiles) {
    ASSERT_NE(import_cfh_, nullptr);

    std::string value;
-    db_->Get(ReadOptions(), import_cfh_, "K1", &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, "K1", &value));
    ASSERT_EQ(value, "V1");
-    db_->Get(ReadOptions(), import_cfh_, "K2", &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, "K2", &value));
    ASSERT_EQ(value, "V2");
    ASSERT_OK(db_->DropColumnFamily(import_cfh_));
    ASSERT_OK(db_->DestroyColumnFamilyHandle(import_cfh_));
@@ -125,9 +125,9 @@ TEST_F(ImportColumnFamilyTest, ImportSSTFileWriterFiles) {
    ASSERT_NE(import_cfh_, nullptr);

    std::string value;
-    db_->Get(ReadOptions(), import_cfh_, "K3", &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, "K3", &value));
    ASSERT_EQ(value, "V1");
-    db_->Get(ReadOptions(), import_cfh_, "K4", &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, "K4", &value));
    ASSERT_EQ(value, "V2");
  }
 }
@@ -214,7 +214,7 @@ TEST_F(ImportColumnFamilyTest, ImportSSTFileWriterFilesWithOverlap) {

  for (int i = 0; i < 100; i++) {
    std::string value;
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value));
    if (i % 16 == 0) {
      ASSERT_EQ(value, Key(i) + "_overwrite4");
    } else if (i % 4 == 0) {
@@ -235,7 +235,7 @@ TEST_F(ImportColumnFamilyTest, ImportSSTFileWriterFilesWithOverlap) {
  ASSERT_OK(db_->Flush(FlushOptions(), import_cfh_));
  for (int i = 0; i < 100; i++) {
    std::string value;
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value));
    if (i % 5 == 0) {
      ASSERT_EQ(value, Key(i) + "_overwrite5");
    } else if (i % 16 == 0) {
@@ -254,7 +254,7 @@ TEST_F(ImportColumnFamilyTest, ImportSSTFileWriterFilesWithOverlap) {
      db_->CompactRange(CompactRangeOptions(), import_cfh_, nullptr, nullptr));
  for (int i = 0; i < 100; i++) {
    std::string value;
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value));
    if (i % 5 == 0) {
      ASSERT_EQ(value, Key(i) + "_overwrite5");
    } else if (i % 16 == 0) {
@@ -318,12 +318,12 @@ TEST_F(ImportColumnFamilyTest, ImportExportedSSTFromAnotherCF) {
  std::string value1, value2;

  for (int i = 0; i < 100; ++i) {
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value1);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value1));
    ASSERT_EQ(Get(1, Key(i)), value1);
  }

  for (int i = 0; i < 100; ++i) {
-    db_->Get(ReadOptions(), import_cfh2_, Key(i), &value2);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh2_, Key(i), &value2));
    ASSERT_EQ(Get(1, Key(i)), value2);
  }

@@ -340,16 +340,16 @@ TEST_F(ImportColumnFamilyTest, ImportExportedSSTFromAnotherCF) {
        db_->Get(ReadOptions(), import_cfh_, Key(i), &value1).IsNotFound());
  }
  for (int i = 25; i < 50; ++i) {
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value1);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value1));
    ASSERT_EQ(Key(i) + "_overwrite3", value1);
  }
  for (int i = 50; i < 100; ++i) {
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value1);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value1));
    ASSERT_EQ(Key(i) + "_overwrite2", value1);
  }

  for (int i = 0; i < 100; ++i) {
-    db_->Get(ReadOptions(), import_cfh2_, Key(i), &value2);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh2_, Key(i), &value2));
    ASSERT_EQ(Get(1, Key(i)), value2);
  }

@@ -363,16 +363,16 @@ TEST_F(ImportColumnFamilyTest, ImportExportedSSTFromAnotherCF) {
        db_->Get(ReadOptions(), import_cfh_, Key(i), &value1).IsNotFound());
  }
  for (int i = 25; i < 50; ++i) {
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value1);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value1));
    ASSERT_EQ(Key(i) + "_overwrite3", value1);
  }
  for (int i = 50; i < 100; ++i) {
-    db_->Get(ReadOptions(), import_cfh_, Key(i), &value1);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh_, Key(i), &value1));
    ASSERT_EQ(Key(i) + "_overwrite2", value1);
  }

  for (int i = 0; i < 100; ++i) {
-    db_->Get(ReadOptions(), import_cfh2_, Key(i), &value2);
+    ASSERT_OK(db_->Get(ReadOptions(), import_cfh2_, Key(i), &value2));
    ASSERT_EQ(Get(1, Key(i)), value2);
  }
 }
@@ -424,7 +424,7 @@ TEST_F(ImportColumnFamilyTest, ImportExportedSSTFromAnotherDB) {

  for (int i = 0; i < 100; ++i) {
    std::string value;
-    db_copy->Get(ReadOptions(), cfh, Key(i), &value);
+    ASSERT_OK(db_copy->Get(ReadOptions(), cfh, Key(i), &value));
    ASSERT_EQ(Get(1, Key(i)), value);
  }
  ASSERT_OK(db_copy->DropColumnFamily(cfh));
@@ -61,6 +61,7 @@ void PrintLevelStatsHeader(char* buf, size_t len, const std::string& cf_name,
                           const std::string& group_by) {
  int written_size =
      snprintf(buf, len, "\n** Compaction Stats [%s] **\n", cf_name.c_str());
+  written_size = std::min(written_size, static_cast<int>(len));
  auto hdr = [](LevelStatType t) {
    return InternalStats::compaction_level_stats.at(t).header_name.c_str();
  };
@@ -80,6 +81,7 @@ void PrintLevelStatsHeader(char* buf, size_t len, const std::string& cf_name,
      hdr(LevelStatType::KEY_DROP));

  written_size += line_size;
+  written_size = std::min(written_size, static_cast<int>(len));
  snprintf(buf + written_size, len - written_size, "%s\n",
           std::string(line_size, '-').c_str());
 }
@@ -376,10 +378,11 @@ const std::unordered_map<std::string, DBPropertyInfo>
         {false, &InternalStats::HandleSsTables, nullptr, nullptr, nullptr}},
        {DB::Properties::kAggregatedTableProperties,
         {false, &InternalStats::HandleAggregatedTableProperties, nullptr,
-          nullptr, nullptr}},
+          &InternalStats::HandleAggregatedTablePropertiesMap, nullptr}},
        {DB::Properties::kAggregatedTablePropertiesAtLevel,
         {false, &InternalStats::HandleAggregatedTablePropertiesAtLevel,
-          nullptr, nullptr, nullptr}},
+          nullptr, &InternalStats::HandleAggregatedTablePropertiesAtLevelMap,
+          nullptr}},
        {DB::Properties::kNumImmutableMemTable,
         {false, nullptr, &InternalStats::HandleNumImmutableMemTable, nullptr,
          nullptr}},
@@ -508,11 +511,12 @@ bool InternalStats::GetStringProperty(const DBPropertyInfo& property_info,
 }

 bool InternalStats::GetMapProperty(const DBPropertyInfo& property_info,
-                                   const Slice& /*property*/,
+                                   const Slice& property,
                                   std::map<std::string, std::string>* value) {
  assert(value != nullptr);
  assert(property_info.handle_map != nullptr);
-  return (this->*(property_info.handle_map))(value);
+  Slice arg = GetPropertyNameAndArg(property).second;
+  return (this->*(property_info.handle_map))(value, arg);
 }

 bool InternalStats::GetIntProperty(const DBPropertyInfo& property_info,
@@ -588,7 +592,7 @@ bool InternalStats::HandleStats(std::string* value, Slice suffix) {
 }

 bool InternalStats::HandleCFMapStats(
-    std::map<std::string, std::string>* cf_stats) {
+    std::map<std::string, std::string>* cf_stats, Slice /*suffix*/) {
  DumpCFMapStats(cf_stats);
  return true;
 }
@@ -632,7 +636,27 @@ bool InternalStats::HandleAggregatedTableProperties(std::string* value,
  return true;
 }

-bool InternalStats::HandleAggregatedTablePropertiesAtLevel(std::string* value,
+static std::map<std::string, std::string> MapUint64ValuesToString(
+    const std::map<std::string, uint64_t>& from) {
+  std::map<std::string, std::string> to;
+  for (const auto& e : from) {
+    to[e.first] = ToString(e.second);
+  }
+  return to;
+}
+
+bool InternalStats::HandleAggregatedTablePropertiesMap(
+    std::map<std::string, std::string>* values, Slice /*suffix*/) {
+  std::shared_ptr<const TableProperties> tp;
+  auto s = cfd_->current()->GetAggregatedTableProperties(&tp);
+  if (!s.ok()) {
+    return false;
+  }
+  *values = MapUint64ValuesToString(tp->GetAggregatablePropertiesAsMap());
+  return true;
+}
+
+bool InternalStats::HandleAggregatedTablePropertiesAtLevel(std::string* values,
                                                           Slice suffix) {
  uint64_t level;
  bool ok = ConsumeDecimalNumber(&suffix, &level) && suffix.empty();
@@ -645,7 +669,24 @@ bool InternalStats::HandleAggregatedTablePropertiesAtLevel(std::string* value,
  if (!s.ok()) {
    return false;
  }
-  *value = tp->ToString();
+  *values = tp->ToString();
+  return true;
+}
+
+bool InternalStats::HandleAggregatedTablePropertiesAtLevelMap(
+    std::map<std::string, std::string>* values, Slice suffix) {
+  uint64_t level;
+  bool ok = ConsumeDecimalNumber(&suffix, &level) && suffix.empty();
+  if (!ok || static_cast<int>(level) >= number_levels_) {
+    return false;
+  }
+  std::shared_ptr<const TableProperties> tp;
+  auto s = cfd_->current()->GetAggregatedTableProperties(
+      &tp, static_cast<int>(level));
+  if (!s.ok()) {
+    return false;
+  }
+  *values = MapUint64ValuesToString(tp->GetAggregatablePropertiesAsMap());
  return true;
 }

@@ -44,7 +44,9 @@ struct DBPropertyInfo {
                                    Version* version);

  // @param props Map of general properties to populate
-  bool (InternalStats::*handle_map)(std::map<std::string, std::string>* props);
+  // @param suffix Argument portion of the property. (see handle_string)
+  bool (InternalStats::*handle_map)(std::map<std::string, std::string>* props,
+                                    Slice suffix);

  // handle the string type properties rely on DBImpl methods
  // @param value Value-result argument for storing the property's string value
@@ -525,7 +527,8 @@ class InternalStats {
  bool HandleCompressionRatioAtLevelPrefix(std::string* value, Slice suffix);
  bool HandleLevelStats(std::string* value, Slice suffix);
  bool HandleStats(std::string* value, Slice suffix);
-  bool HandleCFMapStats(std::map<std::string, std::string>* compaction_stats);
+  bool HandleCFMapStats(std::map<std::string, std::string>* compaction_stats,
+                        Slice suffix);
  bool HandleCFStats(std::string* value, Slice suffix);
  bool HandleCFStatsNoFileHistogram(std::string* value, Slice suffix);
  bool HandleCFFileHistogram(std::string* value, Slice suffix);
@@ -533,6 +536,10 @@ class InternalStats {
  bool HandleSsTables(std::string* value, Slice suffix);
  bool HandleAggregatedTableProperties(std::string* value, Slice suffix);
  bool HandleAggregatedTablePropertiesAtLevel(std::string* value, Slice suffix);
+  bool HandleAggregatedTablePropertiesMap(
+      std::map<std::string, std::string>* values, Slice suffix);
+  bool HandleAggregatedTablePropertiesAtLevelMap(
+      std::map<std::string, std::string>* values, Slice suffix);
  bool HandleNumImmutableMemTable(uint64_t* value, DBImpl* db,
                                  Version* version);
  bool HandleNumImmutableMemTableFlushed(uint64_t* value, DBImpl* db,
@@ -192,10 +192,10 @@ TEST_F(EventListenerTest, OnSingleDBCompactionTest) {
  ASSERT_OK(Put(7, "popovich", std::string(90000, 'p')));
  for (int i = 1; i < 8; ++i) {
    ASSERT_OK(Flush(i));
-    dbfull()->TEST_WaitForFlushMemTable();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
    ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), handles_[i],
                                     nullptr, nullptr));
-    dbfull()->TEST_WaitForCompact();
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
  }

  ASSERT_EQ(listener->compacted_dbs_.size(), cf_names.size());
@@ -211,6 +211,10 @@ class TestFlushListener : public EventListener {
      : slowdown_count(0), stop_count(0), db_closed(), env_(env), test_(test) {
    db_closed = false;
  }
+
+  virtual ~TestFlushListener() {
+    prev_fc_info_.status.PermitUncheckedError();  // Ignore the status
+  }
  void OnTableFileCreated(
      const TableFileCreationInfo& info) override {
    // remember the info for later checking the FlushJobInfo.
@@ -333,7 +337,7 @@ TEST_F(EventListenerTest, OnSingleDBFlushTest) {
  ASSERT_OK(Put(7, "popovich", std::string(90000, 'p')));
  for (int i = 1; i < 8; ++i) {
    ASSERT_OK(Flush(i));
-    dbfull()->TEST_WaitForFlushMemTable();
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
    ASSERT_EQ(listener->flushed_dbs_.size(), i);
    ASSERT_EQ(listener->flushed_column_family_names_.size(), i);
  }
@@ -417,7 +421,7 @@ TEST_F(EventListenerTest, MultiDBMultiListeners) {
    ASSERT_OK(DB::Open(options, dbname_ + ToString(d), &db));
    for (size_t c = 0; c < cf_names.size(); ++c) {
      ColumnFamilyHandle* handle;
-      db->CreateColumnFamily(cf_opts, cf_names[c], &handle);
+      ASSERT_OK(db->CreateColumnFamily(cf_opts, cf_names[c], &handle));
      handles.push_back(handle);
    }

@@ -435,7 +439,8 @@ TEST_F(EventListenerTest, MultiDBMultiListeners) {
  for (size_t c = 0; c < cf_names.size(); ++c) {
    for (int d = 0; d < kNumDBs; ++d) {
      ASSERT_OK(dbs[d]->Flush(FlushOptions(), vec_handles[d][c]));
-      static_cast_with_check<DBImpl>(dbs[d])->TEST_WaitForFlushMemTable();
+      ASSERT_OK(
+          static_cast_with_check<DBImpl>(dbs[d])->TEST_WaitForFlushMemTable());
    }
  }

@@ -494,10 +499,10 @@ TEST_F(EventListenerTest, DisableBGCompaction) {
  // keep writing until writes are forced to stop.
  for (int i = 0; static_cast<int>(cf_meta.file_count) < kSlowdownTrigger * 10;
       ++i) {
-    Put(1, ToString(i), std::string(10000, 'x'), WriteOptions());
+    ASSERT_OK(Put(1, ToString(i), std::string(10000, 'x'), WriteOptions()));
    FlushOptions fo;
    fo.allow_write_stall = true;
-    db_->Flush(fo, handles_[1]);
+    ASSERT_OK(db_->Flush(fo, handles_[1]));
    db_->GetColumnFamilyMetaData(handles_[1], &cf_meta);
  }
  ASSERT_GE(listener->slowdown_count, kSlowdownTrigger * 9);
@@ -534,7 +539,7 @@ TEST_F(EventListenerTest, CompactionReasonLevel) {
  for (int i = 0; i < 4; i++) {
    GenerateNewRandomFile(&rnd);
  }
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  ASSERT_EQ(listener->compaction_reasons_.size(), 1);
  ASSERT_EQ(listener->compaction_reasons_[0],
@@ -551,14 +556,14 @@ TEST_F(EventListenerTest, CompactionReasonLevel) {
  }

  // Do a trivial move from L0 -> L1
-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  options.max_bytes_for_level_base = 1;
  Close();
  listener->compaction_reasons_.clear();
  Reopen(options);

-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  ASSERT_GT(listener->compaction_reasons_.size(), 1);

  for (auto compaction_reason : listener->compaction_reasons_) {
@@ -570,7 +575,7 @@ TEST_F(EventListenerTest, CompactionReasonLevel) {
  listener->compaction_reasons_.clear();
  Reopen(options);

-  Put("key", "value");
+  ASSERT_OK(Put("key", "value"));
  CompactRangeOptions cro;
  cro.bottommost_level_compaction = BottommostLevelCompaction::kForceOptimized;
  ASSERT_OK(db_->CompactRange(cro, nullptr, nullptr));
@@ -604,7 +609,7 @@ TEST_F(EventListenerTest, CompactionReasonUniversal) {
  for (int i = 0; i < 8; i++) {
    GenerateNewRandomFile(&rnd);
  }
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  ASSERT_GT(listener->compaction_reasons_.size(), 0);
  for (auto compaction_reason : listener->compaction_reasons_) {
@@ -622,7 +627,7 @@ TEST_F(EventListenerTest, CompactionReasonUniversal) {
  for (int i = 0; i < 8; i++) {
    GenerateNewRandomFile(&rnd);
  }
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  ASSERT_GT(listener->compaction_reasons_.size(), 0);
  for (auto compaction_reason : listener->compaction_reasons_) {
@@ -634,7 +639,7 @@ TEST_F(EventListenerTest, CompactionReasonUniversal) {
  listener->compaction_reasons_.clear();
  Reopen(options);

-  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));

  ASSERT_GT(listener->compaction_reasons_.size(), 0);
  for (auto compaction_reason : listener->compaction_reasons_) {
@@ -663,7 +668,7 @@ TEST_F(EventListenerTest, CompactionReasonFIFO) {
  for (int i = 0; i < 4; i++) {
    GenerateNewRandomFile(&rnd);
  }
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());

  ASSERT_GT(listener->compaction_reasons_.size(), 0);
  for (auto compaction_reason : listener->compaction_reasons_) {
@@ -783,7 +788,7 @@ TEST_F(EventListenerTest, TableFileCreationListenersTest) {
  ASSERT_OK(Put("foo", "aaa"));
  ASSERT_OK(Put("bar", "bbb"));
  ASSERT_OK(Flush());
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  listener->CheckAndResetCounters(1, 1, 0, 0, 0, 0);
  ASSERT_OK(Put("foo", "aaa1"));
  ASSERT_OK(Put("bar", "bbb1"));
@@ -796,21 +801,23 @@ TEST_F(EventListenerTest, TableFileCreationListenersTest) {
  ASSERT_OK(Put("foo", "aaa2"));
  ASSERT_OK(Put("bar", "bbb2"));
  ASSERT_OK(Flush());
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  listener->CheckAndResetCounters(1, 1, 0, 0, 0, 0);

  const Slice kRangeStart = "a";
  const Slice kRangeEnd = "z";
-  dbfull()->CompactRange(CompactRangeOptions(), &kRangeStart, &kRangeEnd);
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_OK(
+      dbfull()->CompactRange(CompactRangeOptions(), &kRangeStart, &kRangeEnd));
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
  listener->CheckAndResetCounters(0, 0, 0, 1, 1, 0);

  ASSERT_OK(Put("foo", "aaa3"));
  ASSERT_OK(Put("bar", "bbb3"));
  ASSERT_OK(Flush());
  test_env->SetStatus(Status::NotSupported("not supported"));
-  dbfull()->CompactRange(CompactRangeOptions(), &kRangeStart, &kRangeEnd);
-  dbfull()->TEST_WaitForCompact();
+  ASSERT_NOK(
+      dbfull()->CompactRange(CompactRangeOptions(), &kRangeStart, &kRangeEnd));
+  ASSERT_NOK(dbfull()->TEST_WaitForCompact());
  listener->CheckAndResetCounters(1, 1, 0, 1, 1, 1);
  Close();
 }
@@ -1076,8 +1083,8 @@ TEST_F(EventListenerTest, OnFileOperationTest) {
  }
  DestroyAndReopen(options);
  ASSERT_OK(Put("foo", "aaa"));
-  dbfull()->Flush(FlushOptions());
-  dbfull()->TEST_WaitForFlushMemTable();
+  ASSERT_OK(dbfull()->Flush(FlushOptions()));
+  ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
  ASSERT_GE(listener->file_writes_.load(),
            listener->file_writes_success_.load());
  ASSERT_GT(listener->file_writes_.load(), 0);
@@ -119,16 +119,26 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        break;

      case kBadHeader:
-        if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency) {
-          // in clean shutdown we don't expect any error in the log files
+        if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency ||
+            wal_recovery_mode == WALRecoveryMode::kPointInTimeRecovery) {
+          // In clean shutdown we don't expect any error in the log files.
+          // In point-in-time recovery an incomplete record at the end could
+          // produce a hole in the recovered data. Report an error here, which
+          // higher layers can choose to ignore when it's provable there is no
+          // hole.
          ReportCorruption(drop_size, "truncated header");
        }
        FALLTHROUGH_INTENDED;

      case kEof:
        if (in_fragmented_record) {
-          if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency) {
-            // in clean shutdown we don't expect any error in the log files
+          if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency ||
+              wal_recovery_mode == WALRecoveryMode::kPointInTimeRecovery) {
+            // In clean shutdown we don't expect any error in the log files.
+            // In point-in-time recovery an incomplete record at the end could
+            // produce a hole in the recovered data. Report an error here, which
+            // higher layers can choose to ignore when it's provable there is no
+            // hole.
            ReportCorruption(scratch->size(), "error reading trailing data");
          }
          // This can be caused by the writer dying immediately after
@@ -142,8 +152,13 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        if (wal_recovery_mode != WALRecoveryMode::kSkipAnyCorruptedRecords) {
          // Treat a record from a previous instance of the log as EOF.
          if (in_fragmented_record) {
-            if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency) {
-              // in clean shutdown we don't expect any error in the log files
+            if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency ||
+                wal_recovery_mode == WALRecoveryMode::kPointInTimeRecovery) {
+              // In clean shutdown we don't expect any error in the log files.
+              // In point-in-time recovery an incomplete record at the end could
+              // produce a hole in the recovered data. Report an error here,
+              // which higher layers can choose to ignore when it's provable
+              // there is no hole.
              ReportCorruption(scratch->size(), "error reading trailing data");
            }
            // This can be caused by the writer dying immediately after
@@ -164,6 +179,20 @@ bool Reader::ReadRecord(Slice* record, std::string* scratch,
        break;

      case kBadRecordLen:
+        if (eof_) {
+          if (wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency ||
+              wal_recovery_mode == WALRecoveryMode::kPointInTimeRecovery) {
+            // In clean shutdown we don't expect any error in the log files.
+            // In point-in-time recovery an incomplete record at the end could
+            // produce a hole in the recovered data. Report an error here, which
+            // higher layers can choose to ignore when it's provable there is no
+            // hole.
+            ReportCorruption(drop_size, "truncated record body");
+          }
+          return false;
+        }
+        FALLTHROUGH_INTENDED;
+
      case kBadRecordChecksum:
        if (recycled_ &&
            wal_recovery_mode ==
@@ -355,18 +384,14 @@ unsigned int Reader::ReadPhysicalRecord(Slice* result, size_t* drop_size) {
      }
    }
    if (header_size + length > buffer_.size()) {
+      assert(buffer_.size() >= static_cast<size_t>(header_size));
      *drop_size = buffer_.size();
      buffer_.clear();
-      if (!eof_) {
-        return kBadRecordLen;
-      }
-      // If the end of the file has been reached without reading |length|
-      // bytes of payload, assume the writer died in the middle of writing the
-      // record. Don't report a corruption unless requested.
-      if (*drop_size) {
-        return kBadHeader;
-      }
-      return kEof;
+      // If the end of the read has been reached without seeing
+      // `header_size + length` bytes of payload, report a corruption. The
+      // higher layers can decide how to handle it based on the recovery mode,
+      // whether this occurred at EOF, whether this is the final WAL, etc.
+      return kBadRecordLen;
    }

    if (type == kZeroType && length == 0) {
@@ -191,7 +191,7 @@ class LogTest : public ::testing::TestWithParam<std::tuple<int, bool>> {
  Slice* get_reader_contents() { return &reader_contents_; }

  void Write(const std::string& msg) {
-    writer_.AddRecord(Slice(msg));
+    ASSERT_OK(writer_.AddRecord(Slice(msg)));
  }

  size_t WrittenBytes() const {
@@ -465,7 +465,7 @@ TEST_P(LogTest, BadLengthAtEndIsNotIgnored) {
  ShrinkSize(1);
  ASSERT_EQ("EOF", Read(WALRecoveryMode::kAbsoluteConsistency));
  ASSERT_GT(DroppedBytes(), 0U);
-  ASSERT_EQ("OK", MatchError("Corruption: truncated header"));
+  ASSERT_EQ("OK", MatchError("Corruption: truncated record body"));
 }

 TEST_P(LogTest, ChecksumMismatch) {
@@ -573,9 +573,7 @@ TEST_P(LogTest, PartialLastIsNotIgnored) {
  ShrinkSize(1);
  ASSERT_EQ("EOF", Read(WALRecoveryMode::kAbsoluteConsistency));
  ASSERT_GT(DroppedBytes(), 0U);
-  ASSERT_EQ("OK", MatchError(
-                      "Corruption: truncated headerCorruption: "
-                      "error reading trailing data"));
+  ASSERT_EQ("OK", MatchError("Corruption: truncated record body"));
 }

 TEST_P(LogTest, ErrorJoinsRecords) {
@@ -691,8 +689,8 @@ TEST_P(LogTest, Recycle) {
      new test::OverwritingStringSink(get_reader_contents()),
      "" /* don't care */));
  Writer recycle_writer(std::move(dest_holder), 123, true);
-  recycle_writer.AddRecord(Slice("foooo"));
-  recycle_writer.AddRecord(Slice("bar"));
+  ASSERT_OK(recycle_writer.AddRecord(Slice("foooo")));
+  ASSERT_OK(recycle_writer.AddRecord(Slice("bar")));
  ASSERT_GE(get_reader_contents()->size(), log::kBlockSize * 2);
  ASSERT_EQ("foooo", Read());
  ASSERT_EQ("bar", Read());
@@ -784,11 +782,13 @@ class RetriableLogTest : public ::testing::TestWithParam<int> {
    return file->contents_;
  }

-  void Encode(const std::string& msg) { log_writer_->AddRecord(Slice(msg)); }
+  void Encode(const std::string& msg) {
+    ASSERT_OK(log_writer_->AddRecord(Slice(msg)));
+  }

  void Write(const Slice& data) {
-    writer_->Append(data);
-    writer_->Sync(true);
+    ASSERT_OK(writer_->Append(data));
+    ASSERT_OK(writer_->Sync(true));
  }

  bool TryRead(std::string* result) {
@@ -33,7 +33,7 @@ Writer::Writer(std::unique_ptr<WritableFileWriter>&& dest, uint64_t log_number,

 Writer::~Writer() {
  if (dest_) {
-    WriteBuffer();
+    WriteBuffer().PermitUncheckedError();
  }
 }

@@ -100,13 +100,13 @@ TEST_F(ManualCompactionTest, CompactTouchesAllKeys) {
    options.compaction_filter = new DestroyAllCompactionFilter();
    ASSERT_OK(DB::Open(options, dbname_, &db));

-    db->Put(WriteOptions(), Slice("key1"), Slice("destroy"));
-    db->Put(WriteOptions(), Slice("key2"), Slice("destroy"));
-    db->Put(WriteOptions(), Slice("key3"), Slice("value3"));
-    db->Put(WriteOptions(), Slice("key4"), Slice("destroy"));
+    ASSERT_OK(db->Put(WriteOptions(), Slice("key1"), Slice("destroy")));
+    ASSERT_OK(db->Put(WriteOptions(), Slice("key2"), Slice("destroy")));
+    ASSERT_OK(db->Put(WriteOptions(), Slice("key3"), Slice("value3")));
+    ASSERT_OK(db->Put(WriteOptions(), Slice("key4"), Slice("destroy")));

    Slice key4("key4");
-    db->CompactRange(CompactRangeOptions(), nullptr, &key4);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), nullptr, &key4));
    Iterator* itr = db->NewIterator(ReadOptions());
    itr->SeekToFirst();
    ASSERT_TRUE(itr->Valid());
@@ -135,21 +135,21 @@ TEST_F(ManualCompactionTest, Test) {
  // create first key range
  WriteBatch batch;
  for (int i = 0; i < kNumKeys; i++) {
-    batch.Put(Key1(i), "value for range 1 key");
+    ASSERT_OK(batch.Put(Key1(i), "value for range 1 key"));
  }
  ASSERT_OK(db->Write(WriteOptions(), &batch));

  // create second key range
  batch.Clear();
  for (int i = 0; i < kNumKeys; i++) {
-    batch.Put(Key2(i), "value for range 2 key");
+    ASSERT_OK(batch.Put(Key2(i), "value for range 2 key"));
  }
  ASSERT_OK(db->Write(WriteOptions(), &batch));

  // delete second key range
  batch.Clear();
  for (int i = 0; i < kNumKeys; i++) {
-    batch.Delete(Key2(i));
+    ASSERT_OK(batch.Delete(Key2(i)));
  }
  ASSERT_OK(db->Write(WriteOptions(), &batch));

@@ -160,7 +160,7 @@ TEST_F(ManualCompactionTest, Test) {
  Slice greatest(end_key.data(), end_key.size());

  // commenting out the line below causes the example to work correctly
-  db->CompactRange(CompactRangeOptions(), &least, &greatest);
+  ASSERT_OK(db->CompactRange(CompactRangeOptions(), &least, &greatest));

  // count the keys
  Iterator* iter = db->NewIterator(ReadOptions());
@@ -205,7 +205,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    Slice start("5");
    Slice end("7");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), &start, &end);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), &start, &end));
    ASSERT_EQ(0, filter->NumKeys());
  }

@@ -215,7 +215,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    Slice start("3");
    Slice end("7");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), &start, &end);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), &start, &end));
    ASSERT_EQ(2, filter->NumKeys());
    ASSERT_EQ(0, filter->KeyLevel("4"));
    ASSERT_EQ(0, filter->KeyLevel("8"));
@@ -227,7 +227,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    // no file has keys in range (-inf, 0]
    Slice end("0");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), nullptr, &end);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), nullptr, &end));
    ASSERT_EQ(0, filter->NumKeys());
  }

@@ -237,7 +237,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    // no file has keys in range [9, inf)
    Slice start("9");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), &start, nullptr);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), &start, nullptr));
    ASSERT_EQ(0, filter->NumKeys());
  }

@@ -248,7 +248,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    Slice start("2");
    Slice end("2");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), &start, &end);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), &start, &end));
    ASSERT_EQ(1, filter->NumKeys());
    ASSERT_EQ(0, filter->KeyLevel("2"));
  }
@@ -260,7 +260,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    Slice start("2");
    Slice end("5");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), &start, &end);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), &start, &end));
    ASSERT_EQ(3, filter->NumKeys());
    ASSERT_EQ(1, filter->KeyLevel("2"));
    ASSERT_EQ(1, filter->KeyLevel("4"));
@@ -273,7 +273,7 @@ TEST_F(ManualCompactionTest, SkipLevel) {
    // [0, inf) overlaps all files
    Slice start("0");
    filter->Reset();
-    db->CompactRange(CompactRangeOptions(), &start, nullptr);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), &start, nullptr));
    ASSERT_EQ(4, filter->NumKeys());
    // 1 is first compacted to L1 and then further compacted into [2, 4, 8],
    // so finally the logged level for 1 is L1.
@@ -328,9 +328,11 @@ class MemTableIterator : public InternalIterator {
    PERF_COUNTER_ADD(seek_on_memtable_count, 1);
    if (bloom_) {
      // iterator should only use prefix bloom filter
-      Slice user_k(ExtractUserKey(k));
-      if (prefix_extractor_->InDomain(user_k) &&
-          !bloom_->MayContain(prefix_extractor_->Transform(user_k))) {
+      auto ts_sz = comparator_.comparator.user_comparator()->timestamp_size();
+      Slice user_k_without_ts(ExtractUserKeyAndStripTimestamp(k, ts_sz));
+      if (prefix_extractor_->InDomain(user_k_without_ts) &&
+          !bloom_->MayContain(
+              prefix_extractor_->Transform(user_k_without_ts))) {
        PERF_COUNTER_ADD(bloom_memtable_miss_count, 1);
        valid_ = false;
        return;
@@ -345,9 +347,11 @@ class MemTableIterator : public InternalIterator {
    PERF_TIMER_GUARD(seek_on_memtable_time);
    PERF_COUNTER_ADD(seek_on_memtable_count, 1);
    if (bloom_) {
-      Slice user_k(ExtractUserKey(k));
-      if (prefix_extractor_->InDomain(user_k) &&
-          !bloom_->MayContain(prefix_extractor_->Transform(user_k))) {
+      auto ts_sz = comparator_.comparator.user_comparator()->timestamp_size();
+      Slice user_k_without_ts(ExtractUserKeyAndStripTimestamp(k, ts_sz));
+      if (prefix_extractor_->InDomain(user_k_without_ts) &&
+          !bloom_->MayContain(
+              prefix_extractor_->Transform(user_k_without_ts))) {
        PERF_COUNTER_ADD(bloom_memtable_miss_count, 1);
        valid_ = false;
        return;
@@ -480,10 +484,10 @@ MemTable::MemTableStats MemTable::ApproximateStats(const Slice& start_ikey,
  return {entry_count * (data_size / n), entry_count};
 }

-bool MemTable::Add(SequenceNumber s, ValueType type,
-                   const Slice& key, /* user key */
-                   const Slice& value, bool allow_concurrent,
-                   MemTablePostProcessInfo* post_process_info, void** hint) {
+Status MemTable::Add(SequenceNumber s, ValueType type,
+                     const Slice& key, /* user key */
+                     const Slice& value, bool allow_concurrent,
+                     MemTablePostProcessInfo* post_process_info, void** hint) {
  // Format of an entry is concatenation of:
  //  key_size     : varint32 of internal_key.size()
  //  key bytes    : char[internal_key.size()]
@@ -511,6 +515,7 @@ bool MemTable::Add(SequenceNumber s, ValueType type,
  memcpy(p, value.data(), val_size);
  assert((unsigned)(p + val_size - buf) == (unsigned)encoded_len);
  size_t ts_sz = GetInternalKeyComparator().user_comparator()->timestamp_size();
+  Slice key_without_ts = StripTimestampFromUserKey(key, ts_sz);

  if (!allow_concurrent) {
    // Extract prefix for insert with hint.
@@ -519,12 +524,12 @@ bool MemTable::Add(SequenceNumber s, ValueType type,
      Slice prefix = insert_with_hint_prefix_extractor_->Transform(key_slice);
      bool res = table->InsertKeyWithHint(handle, &insert_hints_[prefix]);
      if (UNLIKELY(!res)) {
-        return res;
+        return Status::TryAgain("key+seq exists");
      }
    } else {
      bool res = table->InsertKey(handle);
      if (UNLIKELY(!res)) {
-        return res;
+        return Status::TryAgain("key+seq exists");
      }
    }

@@ -540,12 +545,11 @@ bool MemTable::Add(SequenceNumber s, ValueType type,
    }

    if (bloom_filter_ && prefix_extractor_ &&
-        prefix_extractor_->InDomain(key)) {
-      bloom_filter_->Add(
-          prefix_extractor_->Transform(StripTimestampFromUserKey(key, ts_sz)));
+        prefix_extractor_->InDomain(key_without_ts)) {
+      bloom_filter_->Add(prefix_extractor_->Transform(key_without_ts));
    }
    if (bloom_filter_ && moptions_.memtable_whole_key_filtering) {
-      bloom_filter_->Add(StripTimestampFromUserKey(key, ts_sz));
+      bloom_filter_->Add(key_without_ts);
    }

    // The first sequence number inserted into the memtable
@@ -566,7 +570,7 @@ bool MemTable::Add(SequenceNumber s, ValueType type,
                   ? table->InsertKeyConcurrently(handle)
                   : table->InsertKeyWithHintConcurrently(handle, hint);
    if (UNLIKELY(!res)) {
-      return res;
+      return Status::TryAgain("key+seq exists");
    }

    assert(post_process_info != nullptr);
@@ -577,11 +581,12 @@ bool MemTable::Add(SequenceNumber s, ValueType type,
    }

    if (bloom_filter_ && prefix_extractor_ &&
-        prefix_extractor_->InDomain(key)) {
-      bloom_filter_->AddConcurrently(prefix_extractor_->Transform(key));
+        prefix_extractor_->InDomain(key_without_ts)) {
+      bloom_filter_->AddConcurrently(
+          prefix_extractor_->Transform(key_without_ts));
    }
    if (bloom_filter_ && moptions_.memtable_whole_key_filtering) {
-      bloom_filter_->AddConcurrently(StripTimestampFromUserKey(key, ts_sz));
+      bloom_filter_->AddConcurrently(key_without_ts);
    }

    // atomically update first_seqno_ and earliest_seqno_.
@@ -600,7 +605,7 @@ bool MemTable::Add(SequenceNumber s, ValueType type,
    is_range_del_table_empty_.store(false, std::memory_order_relaxed);
  }
  UpdateOldestKeyTime();
-  return true;
+  return Status::OK();
 }

 // Callback from MemTable::Get()
@@ -821,22 +826,21 @@ bool MemTable::Get(const LookupKey& key, std::string* value,
                 range_del_iter->MaxCoveringTombstoneSeqnum(key.user_key()));
  }

-  Slice user_key = key.user_key();
  bool found_final_value = false;
  bool merge_in_progress = s->IsMergeInProgress();
  bool may_contain = true;
  size_t ts_sz = GetInternalKeyComparator().user_comparator()->timestamp_size();
+  Slice user_key_without_ts = StripTimestampFromUserKey(key.user_key(), ts_sz);
  if (bloom_filter_) {
    // when both memtable_whole_key_filtering and prefix_extractor_ are set,
    // only do whole key filtering for Get() to save CPU
    if (moptions_.memtable_whole_key_filtering) {
-      may_contain =
-          bloom_filter_->MayContain(StripTimestampFromUserKey(user_key, ts_sz));
+      may_contain = bloom_filter_->MayContain(user_key_without_ts);
    } else {
      assert(prefix_extractor_);
-      may_contain =
-          !prefix_extractor_->InDomain(user_key) ||
-          bloom_filter_->MayContain(prefix_extractor_->Transform(user_key));
+      may_contain = !prefix_extractor_->InDomain(user_key_without_ts) ||
+                    bloom_filter_->MayContain(
+                        prefix_extractor_->Transform(user_key_without_ts));
    }
  }

@@ -893,7 +897,7 @@ void MemTable::GetFromTable(const LookupKey& key,
 }

 void MemTable::MultiGet(const ReadOptions& read_options, MultiGetRange* range,
-                        ReadCallback* callback, bool* is_blob) {
+                        ReadCallback* callback) {
  // The sequence number is updated synchronously in version_set.h
  if (IsEmpty()) {
    // Avoiding recording stats for speed.
@@ -946,9 +950,9 @@ void MemTable::MultiGet(const ReadOptions& read_options, MultiGetRange* range,
          range_del_iter->MaxCoveringTombstoneSeqnum(iter->lkey->user_key()));
    }
    GetFromTable(*(iter->lkey), iter->max_covering_tombstone_seq, true,
-                 callback, is_blob, iter->value->GetSelf(), iter->timestamp,
-                 iter->s, &(iter->merge_context), &seq, &found_final_value,
-                 &merge_in_progress);
+                 callback, &iter->is_blob_index, iter->value->GetSelf(),
+                 iter->timestamp, iter->s, &(iter->merge_context), &seq,
+                 &found_final_value, &merge_in_progress);

    if (!found_final_value && merge_in_progress) {
      *(iter->s) = Status::MergeInProgress();
@@ -973,9 +977,8 @@ void MemTable::MultiGet(const ReadOptions& read_options, MultiGetRange* range,
  PERF_COUNTER_ADD(get_from_memtable_count, 1);
 }

-void MemTable::Update(SequenceNumber seq,
-                      const Slice& key,
-                      const Slice& value) {
+Status MemTable::Update(SequenceNumber seq, const Slice& key,
+                        const Slice& value) {
  LookupKey lkey(key, seq);
  Slice mem_key = lkey.memtable_key();

@@ -1019,22 +1022,18 @@ void MemTable::Update(SequenceNumber seq,
                 (unsigned)(VarintLength(key_length) + key_length +
                            VarintLength(value.size()) + value.size()));
          RecordTick(moptions_.statistics, NUMBER_KEYS_UPDATED);
-          return;
+          return Status::OK();
        }
      }
    }
  }

-  // key doesn't exist
-  bool add_res __attribute__((__unused__));
-  add_res = Add(seq, kTypeValue, key, value);
-  // We already checked unused != seq above. In that case, Add should not fail.
-  assert(add_res);
+  // The latest value is not `kTypeValue` or key doesn't exist
+  return Add(seq, kTypeValue, key, value);
 }

-bool MemTable::UpdateCallback(SequenceNumber seq,
-                              const Slice& key,
-                              const Slice& delta) {
+Status MemTable::UpdateCallback(SequenceNumber seq, const Slice& key,
+                                const Slice& delta) {
  LookupKey lkey(key, seq);
  Slice memkey = lkey.memtable_key();

@@ -1088,16 +1087,17 @@ bool MemTable::UpdateCallback(SequenceNumber seq,
            }
            RecordTick(moptions_.statistics, NUMBER_KEYS_UPDATED);
            UpdateFlushState();
-            return true;
+            return Status::OK();
          } else if (status == UpdateStatus::UPDATED) {
-            Add(seq, kTypeValue, key, Slice(str_value));
+            Status s = Add(seq, kTypeValue, key, Slice(str_value));
            RecordTick(moptions_.statistics, NUMBER_KEYS_WRITTEN);
            UpdateFlushState();
-            return true;
+            return s;
          } else if (status == UpdateStatus::UPDATE_FAILED) {
-            // No action required. Return.
+            // `UPDATE_FAILED` is named incorrectly. It indicates no update
+            // happened. It does not indicate a failure happened.
            UpdateFlushState();
-            return true;
+            return Status::OK();
          }
        }
        default:
@@ -1105,9 +1105,8 @@ bool MemTable::UpdateCallback(SequenceNumber seq,
      }
    }
  }
-  // If the latest value is not kTypeValue
-  // or key doesn't exist
-  return false;
+  // The latest value is not `kTypeValue` or key doesn't exist
+  return Status::NotFound();
 }

 size_t MemTable::CountSuccessiveMergeEntries(const LookupKey& key) {
@@ -182,12 +182,13 @@ class MemTable {
  // REQUIRES: if allow_concurrent = false, external synchronization to prevent
  // simultaneous operations on the same MemTable.
  //
-  // Returns false if MemTableRepFactory::CanHandleDuplicatedKey() is true and
-  // the <key, seq> already exists.
-  bool Add(SequenceNumber seq, ValueType type, const Slice& key,
-           const Slice& value, bool allow_concurrent = false,
-           MemTablePostProcessInfo* post_process_info = nullptr,
-           void** hint = nullptr);
+  // Returns `Status::TryAgain` if the `seq`, `key` combination already exists
+  // in the memtable and `MemTableRepFactory::CanHandleDuplicatedKey()` is true.
+  // The next attempt should try a larger value for `seq`.
+  Status Add(SequenceNumber seq, ValueType type, const Slice& key,
+             const Slice& value, bool allow_concurrent = false,
+             MemTablePostProcessInfo* post_process_info = nullptr,
+             void** hint = nullptr);

  // Used to Get value associated with key or Get Merge Operands associated
  // with key.
@@ -237,37 +238,36 @@ class MemTable {
  }

  void MultiGet(const ReadOptions& read_options, MultiGetRange* range,
-                ReadCallback* callback, bool* is_blob);
+                ReadCallback* callback);

-  // Attempts to update the new_value inplace, else does normal Add
-  // Pseudocode
-  //   if key exists in current memtable && prev_value is of type kTypeValue
-  //     if new sizeof(new_value) <= sizeof(prev_value)
-  //       update inplace
-  //     else add(key, new_value)
-  //   else add(key, new_value)
+  // If `key` exists in current memtable with type `kTypeValue` and the existing
+  // value is at least as large as the new value, updates it in-place. Otherwise
+  // adds the new value to the memtable out-of-place.
+  //
+  // Returns `Status::TryAgain` if the `seq`, `key` combination already exists
+  // in the memtable and `MemTableRepFactory::CanHandleDuplicatedKey()` is true.
+  // The next attempt should try a larger value for `seq`.
  //
  // REQUIRES: external synchronization to prevent simultaneous
  // operations on the same MemTable.
-  void Update(SequenceNumber seq,
-              const Slice& key,
-              const Slice& value);
+  Status Update(SequenceNumber seq, const Slice& key, const Slice& value);

-  // If prev_value for key exists, attempts to update it inplace.
-  // else returns false
-  // Pseudocode
-  //   if key exists in current memtable && prev_value is of type kTypeValue
-  //     new_value = delta(prev_value)
-  //     if sizeof(new_value) <= sizeof(prev_value)
-  //       update inplace
-  //     else add(key, new_value)
-  //   else return false
+  // If `key` exists in current memtable with type `kTypeValue` and the existing
+  // value is at least as large as the new value, updates it in-place. Otherwise
+  // if `key` exists in current memtable with type `kTypeValue`, adds the new
+  // value to the memtable out-of-place.
+  //
+  // Returns `Status::NotFound` if `key` does not exist in current memtable or
+  // the latest version of `key` does not have `kTypeValue`.
+  //
+  // Returns `Status::TryAgain` if the `seq`, `key` combination already exists
+  // in the memtable and `MemTableRepFactory::CanHandleDuplicatedKey()` is true.
+  // The next attempt should try a larger value for `seq`.
  //
  // REQUIRES: external synchronization to prevent simultaneous
  // operations on the same MemTable.
-  bool UpdateCallback(SequenceNumber seq,
-                      const Slice& key,
-                      const Slice& delta);
+  Status UpdateCallback(SequenceNumber seq, const Slice& key,
+                        const Slice& delta);

  // Returns the number of successive merge entries starting from the newest
  // entry for the key up to the last non-merge entry or last entry for the
@@ -113,10 +113,10 @@ bool MemTableListVersion::Get(const LookupKey& key, std::string* value,
 }

 void MemTableListVersion::MultiGet(const ReadOptions& read_options,
-                                   MultiGetRange* range, ReadCallback* callback,
-                                   bool* is_blob) {
+                                   MultiGetRange* range,
+                                   ReadCallback* callback) {
  for (auto memtable : memlist_) {
-    memtable->MultiGet(read_options, range, callback, is_blob);
+    memtable->MultiGet(read_options, range, callback);
    if (range->empty()) {
      return;
    }
@@ -334,7 +334,7 @@ bool MemTableList::IsFlushPending() const {
 }

 // Returns the memtables that need to be flushed.
-void MemTableList::PickMemtablesToFlush(const uint64_t* max_memtable_id,
+void MemTableList::PickMemtablesToFlush(uint64_t max_memtable_id,
                                        autovector<MemTable*>* ret) {
  AutoThreadOperationStageUpdater stage_updater(
      ThreadStatus::STAGE_PICK_MEMTABLES_TO_FLUSH);
@@ -345,7 +345,7 @@ void MemTableList::PickMemtablesToFlush(const uint64_t* max_memtable_id,
    if (!atomic_flush && m->atomic_flush_seqno_ != kMaxSequenceNumber) {
      atomic_flush = true;
    }
-    if (max_memtable_id != nullptr && m->GetID() > *max_memtable_id) {
+    if (m->GetID() > max_memtable_id) {
      break;
    }
    if (!m->flush_in_progress_) {
@@ -481,15 +481,16 @@ Status MemTableList::TryInstallMemtableFlushResults(
        // We piggyback the information of  earliest log file to keep in the
        // manifest entry for the last file flushed.
        edit_list.back()->SetMinLogNumberToKeep(min_wal_number_to_keep);
-      } else {
-        min_wal_number_to_keep =
-            PrecomputeMinLogNumberToKeepNon2PC(vset, *cfd, edit_list);
      }

      std::unique_ptr<VersionEdit> wal_deletion;
      if (vset->db_options()->track_and_verify_wals_in_manifest) {
-        const auto& wals = vset->GetWalSet().GetWals();
-        if (!wals.empty() && min_wal_number_to_keep > wals.begin()->first) {
+        if (!vset->db_options()->allow_2pc) {
+          min_wal_number_to_keep =
+              PrecomputeMinLogNumberToKeepNon2PC(vset, *cfd, edit_list);
+        }
+        if (min_wal_number_to_keep >
+            vset->GetWalSet().GetMinWalNumberToKeep()) {
          wal_deletion.reset(new VersionEdit);
          wal_deletion->DeleteWalsBefore(min_wal_number_to_keep);
          edit_list.push_back(wal_deletion.get());
@@ -674,20 +675,11 @@ void MemTableList::RemoveMemTablesOrRestoreFlags(
 }

 uint64_t MemTableList::PrecomputeMinLogContainingPrepSection(
-    const autovector<MemTable*>& memtables_to_flush) {
+    const std::unordered_set<MemTable*>* memtables_to_flush) {
  uint64_t min_log = 0;

  for (auto& m : current_->memlist_) {
-    // Assume the list is very short, we can live with O(m*n). We can optimize
-    // if the performance has some problem.
-    bool should_skip = false;
-    for (MemTable* m_to_flush : memtables_to_flush) {
-      if (m == m_to_flush) {
-        should_skip = true;
-        break;
-      }
-    }
-    if (should_skip) {
+    if (memtables_to_flush && memtables_to_flush->count(m)) {
      continue;
    }

@@ -707,7 +699,8 @@ Status InstallMemtableAtomicFlushResults(
    const autovector<ColumnFamilyData*>& cfds,
    const autovector<const MutableCFOptions*>& mutable_cf_options_list,
    const autovector<const autovector<MemTable*>*>& mems_list, VersionSet* vset,
-    InstrumentedMutex* mu, const autovector<FileMetaData*>& file_metas,
+    LogsWithPrepTracker* prep_tracker, InstrumentedMutex* mu,
+    const autovector<FileMetaData*>& file_metas,
    autovector<MemTable*>* to_delete, FSDirectory* db_directory,
    LogBuffer* log_buffer) {
  AutoThreadOperationStageUpdater stage_updater(
@@ -752,19 +745,20 @@ Status InstallMemtableAtomicFlushResults(
    edit_lists.emplace_back(edits);
  }

-  // TODO(cc): after https://github.com/facebook/rocksdb/pull/7570, handle 2pc
-  // here.
+  WalNumber min_wal_number_to_keep = 0;
+  if (vset->db_options()->allow_2pc) {
+    min_wal_number_to_keep = PrecomputeMinLogNumberToKeep2PC(
+        vset, cfds, edit_lists, mems_list, prep_tracker);
+    edit_lists.back().back()->SetMinLogNumberToKeep(min_wal_number_to_keep);
+  }
+
  std::unique_ptr<VersionEdit> wal_deletion;
  if (vset->db_options()->track_and_verify_wals_in_manifest) {
-    uint64_t min_wal_number_to_keep =
-        PrecomputeMinLogNumberToKeepNon2PC(vset, *cfds[0], edit_lists[0]);
-    for (size_t i = 1; i < cfds.size(); i++) {
-      min_wal_number_to_keep = std::min(
-          min_wal_number_to_keep,
-          PrecomputeMinLogNumberToKeepNon2PC(vset, *cfds[i], edit_lists[i]));
+    if (!vset->db_options()->allow_2pc) {
+      min_wal_number_to_keep =
+          PrecomputeMinLogNumberToKeepNon2PC(vset, cfds, edit_lists);
    }
-    const auto& wals = vset->GetWalSet().GetWals();
-    if (!wals.empty() && min_wal_number_to_keep > wals.begin()->first) {
+    if (min_wal_number_to_keep > vset->GetWalSet().GetMinWalNumberToKeep()) {
      wal_deletion.reset(new VersionEdit);
      wal_deletion->DeleteWalsBefore(min_wal_number_to_keep);
      edit_lists.back().push_back(wal_deletion.get());
@@ -76,7 +76,7 @@ class MemTableListVersion {
  }

  void MultiGet(const ReadOptions& read_options, MultiGetRange* range,
-                ReadCallback* callback, bool* is_blob);
+                ReadCallback* callback);

  // Returns all the merge operands corresponding to the key by searching all
  // memtables starting from the most recent one.
@@ -138,8 +138,8 @@ class MemTableListVersion {
      const autovector<ColumnFamilyData*>& cfds,
      const autovector<const MutableCFOptions*>& mutable_cf_options_list,
      const autovector<const autovector<MemTable*>*>& mems_list,
-      VersionSet* vset, InstrumentedMutex* mu,
-      const autovector<FileMetaData*>& file_meta,
+      VersionSet* vset, LogsWithPrepTracker* prep_tracker,
+      InstrumentedMutex* mu, const autovector<FileMetaData*>& file_meta,
      autovector<MemTable*>* to_delete, FSDirectory* db_directory,
      LogBuffer* log_buffer);

@@ -251,7 +251,7 @@ class MemTableList {

  // Returns the earliest memtables that needs to be flushed. The returned
  // memtables are guaranteed to be in the ascending order of created time.
-  void PickMemtablesToFlush(const uint64_t* max_memtable_id,
+  void PickMemtablesToFlush(uint64_t max_memtable_id,
                            autovector<MemTable*>* mems);

  // Reset status of the given memtable list back to pending state so that
@@ -335,7 +335,7 @@ class MemTableList {
  // Returns the min log containing the prep section after memtables listsed in
  // `memtables_to_flush` are flushed and their status is persisted in manifest.
  uint64_t PrecomputeMinLogContainingPrepSection(
-      const autovector<MemTable*>& memtables_to_flush);
+      const std::unordered_set<MemTable*>* memtables_to_flush = nullptr);

  uint64_t GetEarliestMemTableID() const {
    auto& memlist = current_->memlist_;
@@ -381,8 +381,8 @@ class MemTableList {
      const autovector<ColumnFamilyData*>& cfds,
      const autovector<const MutableCFOptions*>& mutable_cf_options_list,
      const autovector<const autovector<MemTable*>*>& mems_list,
-      VersionSet* vset, InstrumentedMutex* mu,
-      const autovector<FileMetaData*>& file_meta,
+      VersionSet* vset, LogsWithPrepTracker* prep_tracker,
+      InstrumentedMutex* mu, const autovector<FileMetaData*>& file_meta,
      autovector<MemTable*>* to_delete, FSDirectory* db_directory,
      LogBuffer* log_buffer);

@@ -431,7 +431,8 @@ extern Status InstallMemtableAtomicFlushResults(
    const autovector<ColumnFamilyData*>& cfds,
    const autovector<const MutableCFOptions*>& mutable_cf_options_list,
    const autovector<const autovector<MemTable*>*>& mems_list, VersionSet* vset,
-    InstrumentedMutex* mu, const autovector<FileMetaData*>& file_meta,
+    LogsWithPrepTracker* prep_tracker, InstrumentedMutex* mu,
+    const autovector<FileMetaData*>& file_meta,
    autovector<MemTable*>* to_delete, FSDirectory* db_directory,
    LogBuffer* log_buffer);
 }  // namespace ROCKSDB_NAMESPACE
@@ -30,14 +30,14 @@ class MemTableListTest : public testing::Test {
  MemTableListTest() : db(nullptr), file_number(1) {
    dbname = test::PerThreadDBPath("memtable_list_test");
    options.create_if_missing = true;
-    DestroyDB(dbname, options);
+    EXPECT_OK(DestroyDB(dbname, options));
  }

  // Create a test db if not yet created
  void CreateDB() {
    if (db == nullptr) {
      options.create_if_missing = true;
-      DestroyDB(dbname, options);
+      EXPECT_OK(DestroyDB(dbname, options));
      // Open DB only with default column family
      ColumnFamilyOptions cf_options;
      std::vector<ColumnFamilyDescriptor> cf_descs;
@@ -78,7 +78,7 @@ class MemTableListTest : public testing::Test {
      handles.clear();
      delete db;
      db = nullptr;
-      DestroyDB(dbname, options, cf_descs);
+      EXPECT_OK(DestroyDB(dbname, options, cf_descs));
    }
  }

@@ -185,8 +185,9 @@ class MemTableListTest : public testing::Test {
    InstrumentedMutex mutex;
    InstrumentedMutexLock l(&mutex);
    return InstallMemtableAtomicFlushResults(
-        &lists, cfds, mutable_cf_options_list, mems_list, &versions, &mutex,
-        file_meta_ptrs, to_delete, nullptr, &log_buffer);
+        &lists, cfds, mutable_cf_options_list, mems_list, &versions,
+        nullptr /* prep_tracker */, &mutex, file_meta_ptrs, to_delete, nullptr,
+        &log_buffer);
  }
 };

@@ -199,7 +200,7 @@ TEST_F(MemTableListTest, Empty) {
  ASSERT_FALSE(list.IsFlushPending());

  autovector<MemTable*> mems;
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &mems);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &mems);
  ASSERT_EQ(0, mems.size());

  autovector<MemTable*> to_delete;
@@ -242,10 +243,10 @@ TEST_F(MemTableListTest, GetTest) {
  mem->Ref();

  // Write some keys to this memtable.
-  mem->Add(++seq, kTypeDeletion, "key1", "");
-  mem->Add(++seq, kTypeValue, "key2", "value2");
-  mem->Add(++seq, kTypeValue, "key1", "value1");
-  mem->Add(++seq, kTypeValue, "key2", "value2.2");
+  ASSERT_OK(mem->Add(++seq, kTypeDeletion, "key1", ""));
+  ASSERT_OK(mem->Add(++seq, kTypeValue, "key2", "value2"));
+  ASSERT_OK(mem->Add(++seq, kTypeValue, "key1", "value1"));
+  ASSERT_OK(mem->Add(++seq, kTypeValue, "key2", "value2.2"));

  // Fetch the newly written keys
  merge_context.Clear();
@@ -283,8 +284,8 @@ TEST_F(MemTableListTest, GetTest) {
                                kMaxSequenceNumber, 0 /* column_family_id */);
  mem2->Ref();

-  mem2->Add(++seq, kTypeDeletion, "key1", "");
-  mem2->Add(++seq, kTypeValue, "key2", "value2.3");
+  ASSERT_OK(mem2->Add(++seq, kTypeDeletion, "key1", ""));
+  ASSERT_OK(mem2->Add(++seq, kTypeValue, "key2", "value2.3"));

  // Add second memtable to list
  list.Add(mem2, &to_delete);
@@ -359,9 +360,9 @@ TEST_F(MemTableListTest, GetFromHistoryTest) {
  mem->Ref();

  // Write some keys to this memtable.
-  mem->Add(++seq, kTypeDeletion, "key1", "");
-  mem->Add(++seq, kTypeValue, "key2", "value2");
-  mem->Add(++seq, kTypeValue, "key2", "value2.2");
+  ASSERT_OK(mem->Add(++seq, kTypeDeletion, "key1", ""));
+  ASSERT_OK(mem->Add(++seq, kTypeValue, "key2", "value2"));
+  ASSERT_OK(mem->Add(++seq, kTypeValue, "key2", "value2.2"));

  // Fetch the newly written keys
  merge_context.Clear();
@@ -399,7 +400,7 @@ TEST_F(MemTableListTest, GetFromHistoryTest) {
  // Flush this memtable from the list.
  // (It will then be a part of the memtable history).
  autovector<MemTable*> to_flush;
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  ASSERT_EQ(1, to_flush.size());

  MutableCFOptions mutable_cf_options(options);
@@ -443,15 +444,15 @@ TEST_F(MemTableListTest, GetFromHistoryTest) {
                                kMaxSequenceNumber, 0 /* column_family_id */);
  mem2->Ref();

-  mem2->Add(++seq, kTypeDeletion, "key1", "");
-  mem2->Add(++seq, kTypeValue, "key3", "value3");
+  ASSERT_OK(mem2->Add(++seq, kTypeDeletion, "key1", ""));
+  ASSERT_OK(mem2->Add(++seq, kTypeValue, "key3", "value3"));

  // Add second memtable to list
  list.Add(mem2, &to_delete);
  ASSERT_EQ(0, to_delete.size());

  to_flush.clear();
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  ASSERT_EQ(1, to_flush.size());

  // Flush second memtable
@@ -554,11 +555,11 @@ TEST_F(MemTableListTest, FlushPendingTest) {
    std::string value;
    MergeContext merge_context;

-    mem->Add(++seq, kTypeValue, "key1", ToString(i));
-    mem->Add(++seq, kTypeValue, "keyN" + ToString(i), "valueN");
-    mem->Add(++seq, kTypeValue, "keyX" + ToString(i), "value");
-    mem->Add(++seq, kTypeValue, "keyM" + ToString(i), "valueM");
-    mem->Add(++seq, kTypeDeletion, "keyX" + ToString(i), "");
+    ASSERT_OK(mem->Add(++seq, kTypeValue, "key1", ToString(i)));
+    ASSERT_OK(mem->Add(++seq, kTypeValue, "keyN" + ToString(i), "valueN"));
+    ASSERT_OK(mem->Add(++seq, kTypeValue, "keyX" + ToString(i), "value"));
+    ASSERT_OK(mem->Add(++seq, kTypeValue, "keyM" + ToString(i), "valueM"));
+    ASSERT_OK(mem->Add(++seq, kTypeDeletion, "keyX" + ToString(i), ""));

    tables.push_back(mem);
  }
@@ -567,7 +568,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  ASSERT_FALSE(list.IsFlushPending());
  ASSERT_FALSE(list.imm_flush_needed.load(std::memory_order_acquire));
  autovector<MemTable*> to_flush;
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  ASSERT_EQ(0, to_flush.size());

  // Request a flush even though there is nothing to flush
@@ -576,7 +577,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  ASSERT_FALSE(list.imm_flush_needed.load(std::memory_order_acquire));

  // Attempt to 'flush' to clear request for flush
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  ASSERT_EQ(0, to_flush.size());
  ASSERT_FALSE(list.IsFlushPending());
  ASSERT_FALSE(list.imm_flush_needed.load(std::memory_order_acquire));
@@ -600,7 +601,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  ASSERT_TRUE(list.imm_flush_needed.load(std::memory_order_acquire));

  // Pick tables to flush
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  ASSERT_EQ(2, to_flush.size());
  ASSERT_EQ(2, list.NumNotFlushed());
  ASSERT_FALSE(list.IsFlushPending());
@@ -621,7 +622,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  ASSERT_EQ(0, to_delete.size());

  // Pick tables to flush
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  ASSERT_EQ(3, to_flush.size());
  ASSERT_EQ(3, list.NumNotFlushed());
  ASSERT_FALSE(list.IsFlushPending());
@@ -629,7 +630,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {

  // Pick tables to flush again
  autovector<MemTable*> to_flush2;
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush2);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush2);
  ASSERT_EQ(0, to_flush2.size());
  ASSERT_EQ(3, list.NumNotFlushed());
  ASSERT_FALSE(list.IsFlushPending());
@@ -647,7 +648,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  ASSERT_TRUE(list.imm_flush_needed.load(std::memory_order_acquire));

  // Pick tables to flush again
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush2);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush2);
  ASSERT_EQ(1, to_flush2.size());
  ASSERT_EQ(4, list.NumNotFlushed());
  ASSERT_FALSE(list.IsFlushPending());
@@ -668,7 +669,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  ASSERT_EQ(0, to_delete.size());

  // Pick tables to flush
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush);
  // Should pick 4 of 5 since 1 table has been picked in to_flush2
  ASSERT_EQ(4, to_flush.size());
  ASSERT_EQ(5, list.NumNotFlushed());
@@ -677,7 +678,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {

  // Pick tables to flush again
  autovector<MemTable*> to_flush3;
-  list.PickMemtablesToFlush(nullptr /* memtable_id */, &to_flush3);
+  list.PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */, &to_flush3);
  ASSERT_EQ(0, to_flush3.size());  // nothing not in progress of being flushed
  ASSERT_EQ(5, list.NumNotFlushed());
  ASSERT_FALSE(list.IsFlushPending());
@@ -738,7 +739,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  autovector<MemTable*> to_flush4;
  list.FlushRequested();
  ASSERT_TRUE(list.HasFlushRequested());
-  list.PickMemtablesToFlush(&memtable_id, &to_flush4);
+  list.PickMemtablesToFlush(memtable_id, &to_flush4);
  ASSERT_TRUE(to_flush4.empty());
  ASSERT_EQ(1, list.NumNotFlushed());
  ASSERT_TRUE(list.imm_flush_needed.load(std::memory_order_acquire));
@@ -749,7 +750,7 @@ TEST_F(MemTableListTest, FlushPendingTest) {
  // equal to 5. Therefore, only tables[5] will be selected.
  memtable_id = 5;
  list.FlushRequested();
-  list.PickMemtablesToFlush(&memtable_id, &to_flush4);
+  list.PickMemtablesToFlush(memtable_id, &to_flush4);
  ASSERT_EQ(1, static_cast<int>(to_flush4.size()));
  ASSERT_EQ(1, list.NumNotFlushed());
  ASSERT_FALSE(list.imm_flush_needed.load(std::memory_order_acquire));
@@ -823,11 +824,11 @@ TEST_F(MemTableListTest, AtomicFlusTest) {

      std::string value;

-      mem->Add(++seq, kTypeValue, "key1", ToString(i));
-      mem->Add(++seq, kTypeValue, "keyN" + ToString(i), "valueN");
-      mem->Add(++seq, kTypeValue, "keyX" + ToString(i), "value");
-      mem->Add(++seq, kTypeValue, "keyM" + ToString(i), "valueM");
-      mem->Add(++seq, kTypeDeletion, "keyX" + ToString(i), "");
+      ASSERT_OK(mem->Add(++seq, kTypeValue, "key1", ToString(i)));
+      ASSERT_OK(mem->Add(++seq, kTypeValue, "keyN" + ToString(i), "valueN"));
+      ASSERT_OK(mem->Add(++seq, kTypeValue, "keyX" + ToString(i), "value"));
+      ASSERT_OK(mem->Add(++seq, kTypeValue, "keyM" + ToString(i), "valueM"));
+      ASSERT_OK(mem->Add(++seq, kTypeDeletion, "keyX" + ToString(i), ""));

      elem.push_back(mem);
    }
@@ -841,7 +842,8 @@ TEST_F(MemTableListTest, AtomicFlusTest) {
    auto* list = lists[i];
    ASSERT_FALSE(list->IsFlushPending());
    ASSERT_FALSE(list->imm_flush_needed.load(std::memory_order_acquire));
-    list->PickMemtablesToFlush(nullptr /* memtable_id */, &flush_candidates[i]);
+    list->PickMemtablesToFlush(port::kMaxUint64 /* memtable_id */,
+                               &flush_candidates[i]);
    ASSERT_EQ(0, flush_candidates[i].size());
  }
  // Request flush even though there is nothing to flush
@@ -871,8 +873,7 @@ TEST_F(MemTableListTest, AtomicFlusTest) {
  // Pick memtables to flush
  for (auto i = 0; i != num_cfs; ++i) {
    flush_candidates[i].clear();
-    lists[i]->PickMemtablesToFlush(&flush_memtable_ids[i],
-                                   &flush_candidates[i]);
+    lists[i]->PickMemtablesToFlush(flush_memtable_ids[i], &flush_candidates[i]);
    ASSERT_EQ(flush_memtable_ids[i] - 0 + 1,
              static_cast<uint64_t>(flush_candidates[i].size()));
  }
@@ -80,8 +80,8 @@ std::shared_ptr<DB> OpenDb(const std::string& dbname, const bool ttl = false,
  options.create_if_missing = true;
  options.merge_operator = std::make_shared<CountMergeOperator>();
  options.max_successive_merges = max_successive_merges;
+  EXPECT_OK(DestroyDB(dbname, Options()));
  Status s;
-  DestroyDB(dbname, Options());
 // DBWithTTL is not supported in ROCKSDB_LITE
 #ifndef ROCKSDB_LITE
  if (ttl) {
@@ -95,10 +95,8 @@ std::shared_ptr<DB> OpenDb(const std::string& dbname, const bool ttl = false,
  assert(!ttl);
  s = DB::Open(options, dbname, &db);
 #endif  // !ROCKSDB_LITE
-  if (!s.ok()) {
-    std::cerr << s.ToString() << std::endl;
-    assert(false);
-  }
+  EXPECT_OK(s);
+  assert(s.ok());
  return std::shared_ptr<DB>(db);
 }

@@ -258,21 +256,25 @@ void testCounters(Counters& counters, DB* db, bool test_compaction) {

  counters.assert_set("a", 1);

-  if (test_compaction) db->Flush(o);
+  if (test_compaction) {
+    ASSERT_OK(db->Flush(o));
+  }

-  assert(counters.assert_get("a") == 1);
+  ASSERT_EQ(counters.assert_get("a"), 1);

  counters.assert_remove("b");

  // defaut value is 0 if non-existent
-  assert(counters.assert_get("b") == 0);
+  ASSERT_EQ(counters.assert_get("b"), 0);

  counters.assert_add("a", 2);

-  if (test_compaction) db->Flush(o);
+  if (test_compaction) {
+    ASSERT_OK(db->Flush(o));
+  }

  // 1+2 = 3
-  assert(counters.assert_get("a")== 3);
+  ASSERT_EQ(counters.assert_get("a"), 3);

  dumpDb(db);

@@ -282,19 +284,19 @@ void testCounters(Counters& counters, DB* db, bool test_compaction) {
    counters.assert_add("b", i);
    sum += i;
  }
-  assert(counters.assert_get("b") == sum);
+  ASSERT_EQ(counters.assert_get("b"), sum);

  dumpDb(db);

  if (test_compaction) {
-    db->Flush(o);
+    ASSERT_OK(db->Flush(o));

-    db->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+    ASSERT_OK(db->CompactRange(CompactRangeOptions(), nullptr, nullptr));

    dumpDb(db);

-    assert(counters.assert_get("a")== 3);
-    assert(counters.assert_get("b") == sum);
+    ASSERT_EQ(counters.assert_get("a"), 3);
+    ASSERT_EQ(counters.assert_get("b"), sum);
  }
 }

@@ -400,14 +402,14 @@ void testSuccessiveMerge(Counters& counters, size_t max_num_merges,
    sum += i;

    if (i % (max_num_merges + 1) == 0) {
-      assert(num_merge_operator_calls == max_num_merges + 1);
+      ASSERT_EQ(num_merge_operator_calls, max_num_merges + 1);
    } else {
-      assert(num_merge_operator_calls == 0);
+      ASSERT_EQ(num_merge_operator_calls, 0);
    }

    resetNumMergeOperatorCalls();
-    assert(counters.assert_get("z") == sum);
-    assert(num_merge_operator_calls == i % (max_num_merges + 1));
+    ASSERT_EQ(counters.assert_get("z"), sum);
+    ASSERT_EQ(num_merge_operator_calls, i % (max_num_merges + 1));
  }
 }

@@ -424,8 +426,8 @@ void testPartialMerge(Counters* counters, DB* db, size_t max_merge,
    counters->assert_add("b", i);
    tmp_sum += i;
  }
-  db->Flush(o);
-  db->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db->Flush(o));
+  ASSERT_OK(db->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  ASSERT_EQ(tmp_sum, counters->assert_get("b"));
  if (count > max_merge) {
    // in this case, FullMerge should be called instead.
@@ -438,20 +440,20 @@ void testPartialMerge(Counters* counters, DB* db, size_t max_merge,
  // Test case 2: partial merge should not be called when a put is found.
  resetNumPartialMergeCalls();
  tmp_sum = 0;
-  db->Put(ROCKSDB_NAMESPACE::WriteOptions(), "c", "10");
+  ASSERT_OK(db->Put(ROCKSDB_NAMESPACE::WriteOptions(), "c", "10"));
  for (size_t i = 1; i <= count; i++) {
    counters->assert_add("c", i);
    tmp_sum += i;
  }
-  db->Flush(o);
-  db->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  ASSERT_OK(db->Flush(o));
+  ASSERT_OK(db->CompactRange(CompactRangeOptions(), nullptr, nullptr));
  ASSERT_EQ(tmp_sum, counters->assert_get("c"));
  ASSERT_EQ(num_partial_merge_calls, 0U);
 }

 void testSingleBatchSuccessiveMerge(DB* db, size_t max_num_merges,
                                    size_t num_merges) {
-  assert(num_merges > max_num_merges);
+  ASSERT_GT(num_merges, max_num_merges);

  Slice key("BatchSuccessiveMerge");
  uint64_t merge_value = 1;
@@ -462,15 +464,12 @@ void testSingleBatchSuccessiveMerge(DB* db, size_t max_num_merges,
  // Create the batch
  WriteBatch batch;
  for (size_t i = 0; i < num_merges; ++i) {
-    batch.Merge(key, merge_value_slice);
+    ASSERT_OK(batch.Merge(key, merge_value_slice));
  }

  // Apply to memtable and count the number of merges
  resetNumMergeOperatorCalls();
-  {
-    Status s = db->Write(WriteOptions(), &batch);
-    assert(s.ok());
-  }
+  ASSERT_OK(db->Write(WriteOptions(), &batch));
  ASSERT_EQ(
      num_merge_operator_calls,
      static_cast<size_t>(num_merges - (num_merges % (max_num_merges + 1))));
@@ -478,10 +477,7 @@ void testSingleBatchSuccessiveMerge(DB* db, size_t max_num_merges,
  // Get the value
  resetNumMergeOperatorCalls();
  std::string get_value_str;
-  {
-    Status s = db->Get(ReadOptions(), key, &get_value_str);
-    assert(s.ok());
-  }
+  ASSERT_OK(db->Get(ReadOptions(), key, &get_value_str));
  assert(get_value_str.size() == sizeof(uint64_t));
  uint64_t get_value = DecodeFixed64(&get_value_str[0]);
  ASSERT_EQ(get_value, num_merges * merge_value);
@@ -505,7 +501,7 @@ void runTest(const std::string& dbname, const bool use_ttl = false) {
    }
  }

-  DestroyDB(dbname, Options());
+  ASSERT_OK(DestroyDB(dbname, Options()));

  {
    size_t max_merge = 5;
@@ -514,7 +510,8 @@ void runTest(const std::string& dbname, const bool use_ttl = false) {
    testCounters(counters, db.get(), use_compression);
    testSuccessiveMerge(counters, max_merge, max_merge * 2);
    testSingleBatchSuccessiveMerge(db.get(), 5, 7);
-    DestroyDB(dbname, Options());
+    ASSERT_OK(db->Close());
+    ASSERT_OK(DestroyDB(dbname, Options()));
  }

  {
@@ -525,14 +522,16 @@ void runTest(const std::string& dbname, const bool use_ttl = false) {
      auto db = OpenDb(dbname, use_ttl, max_merge);
      MergeBasedCounters counters(db, 0);
      testPartialMerge(&counters, db.get(), max_merge, min_merge, count);
-      DestroyDB(dbname, Options());
+      ASSERT_OK(db->Close());
+      ASSERT_OK(DestroyDB(dbname, Options()));
    }
    {
      auto db = OpenDb(dbname, use_ttl, max_merge);
      MergeBasedCounters counters(db, 0);
      testPartialMerge(&counters, db.get(), max_merge, min_merge,
                       min_merge * 10);
-      DestroyDB(dbname, Options());
+      ASSERT_OK(db->Close());
+      ASSERT_OK(DestroyDB(dbname, Options()));
    }
  }

@@ -543,15 +542,15 @@ void runTest(const std::string& dbname, const bool use_ttl = false) {
      counters.add("test-key", 1);
      counters.add("test-key", 1);
      counters.add("test-key", 1);
-      db->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+      ASSERT_OK(db->CompactRange(CompactRangeOptions(), nullptr, nullptr));
    }

    DB* reopen_db;
    ASSERT_OK(DB::Open(Options(), dbname, &reopen_db));
    std::string value;
-    ASSERT_TRUE(!(reopen_db->Get(ReadOptions(), "test-key", &value).ok()));
+    ASSERT_NOK(reopen_db->Get(ReadOptions(), "test-key", &value));
    delete reopen_db;
-    DestroyDB(dbname, Options());
+    ASSERT_OK(DestroyDB(dbname, Options()));
  }

  /* Temporary remove this test
@@ -591,7 +590,7 @@ TEST_F(MergeTest, MergeWithCompactionAndFlush) {
      testCountersWithFlushAndCompaction(counters, db.get());
    }
  }
-  DestroyDB(dbname, Options());
+  ASSERT_OK(DestroyDB(dbname, Options()));
 }
 #endif  // !ROCKSDB_LITE

@@ -61,7 +61,7 @@ class ObsoleteFilesTest : public DBTestBase {
  void CheckFileTypeCounts(const std::string& dir, int required_log,
                           int required_sst, int required_manifest) {
    std::vector<std::string> filenames;
-    env_->GetChildren(dir, &filenames);
+    ASSERT_OK(env_->GetChildren(dir, &filenames));

    int log_cnt = 0;
    int sst_cnt = 0;
@@ -65,7 +65,7 @@ TEST_F(OptionsFileTest, NumberOfOptionsFiles) {
  const int kReopenCount = 20;
  Options opt;
  opt.create_if_missing = true;
-  DestroyDB(dbname_, opt);
+  ASSERT_OK(DestroyDB(dbname_, opt));
  std::unordered_set<std::string> filename_history;
  DB* db;
  for (int i = 0; i < kReopenCount; ++i) {
@@ -76,12 +76,12 @@ TEST_F(PerfContextTest, SeekIntoDeletion) {
    std::string key = "k" + ToString(i);
    std::string value = "v" + ToString(i);

-    db->Put(write_options, key, value);
+    ASSERT_OK(db->Put(write_options, key, value));
  }

  for (int i = 0; i < FLAGS_total_keys -1 ; ++i) {
    std::string key = "k" + ToString(i);
-    db->Delete(write_options, key);
+    ASSERT_OK(db->Delete(write_options, key));
  }

  HistogramImpl hist_get;
@@ -116,10 +116,9 @@ TEST_F(PerfContextTest, SeekIntoDeletion) {
    auto elapsed_nanos = timer.ElapsedNanos();

    if (FLAGS_verbose) {
-      std::cout << "SeekToFirst uesr key comparison: \n"
-                << hist_seek_to_first.ToString()
-                << "ikey skipped: " << get_perf_context()->internal_key_skipped_count
-                << "\n"
+      std::cout << "SeekToFirst user key comparison: \n"
+                << hist_seek_to_first.ToString() << "ikey skipped: "
+                << get_perf_context()->internal_key_skipped_count << "\n"
                << "idelete skipped: "
                << get_perf_context()->internal_delete_skipped_count << "\n"
                << "elapsed: " << elapsed_nanos << "\n";
@@ -156,7 +155,7 @@ TEST_F(PerfContextTest, SeekIntoDeletion) {
  }

  if (FLAGS_verbose) {
-    std::cout << "Seek uesr key comparison: \n" << hist_seek.ToString();
+    std::cout << "Seek user key comparison: \n" << hist_seek.ToString();
  }
 }

@@ -270,7 +269,7 @@ void ProfileQueries(bool enabled_time = false) {
    std::vector<std::string> values;

    get_perf_context()->Reset();
-    db->Put(write_options, key, value);
+    ASSERT_OK(db->Put(write_options, key, value));
    if (++num_mutex_waited > 3) {
 #ifndef NDEBUG
      ThreadStatusUtil::TEST_SetStateDelay(ThreadStatus::STATE_MUTEX_WAIT, 0U);
@@ -314,7 +313,10 @@ void ProfileQueries(bool enabled_time = false) {
    hist_get.Add(get_perf_context()->user_key_comparison_count);

    get_perf_context()->Reset();
-    db->MultiGet(read_options, multiget_keys, &values);
+    auto statuses = db->MultiGet(read_options, multiget_keys, &values);
+    for (const auto& s : statuses) {
+      ASSERT_OK(s);
+    }
    hist_mget_snapshot.Add(get_perf_context()->get_snapshot_time);
    hist_mget_memtable.Add(get_perf_context()->get_from_memtable_time);
    hist_mget_files.Add(get_perf_context()->get_from_output_files_time);
@@ -324,9 +326,10 @@ void ProfileQueries(bool enabled_time = false) {
  }

  if (FLAGS_verbose) {
-    std::cout << "Put uesr key comparison: \n" << hist_put.ToString()
-              << "Get uesr key comparison: \n" << hist_get.ToString()
-              << "MultiGet uesr key comparison: \n" << hist_get.ToString();
+    std::cout << "Put user key comparison: \n"
+              << hist_put.ToString() << "Get user key comparison: \n"
+              << hist_get.ToString() << "MultiGet user key comparison: \n"
+              << hist_get.ToString();
    std::cout << "Put(): Pre and Post Process Time: \n"
              << hist_write_pre_post.ToString() << " Writing WAL time: \n"
              << hist_write_wal_time.ToString() << "\n"
@@ -428,7 +431,10 @@ void ProfileQueries(bool enabled_time = false) {
    hist_get.Add(get_perf_context()->user_key_comparison_count);

    get_perf_context()->Reset();
-    db->MultiGet(read_options, multiget_keys, &values);
+    auto statuses = db->MultiGet(read_options, multiget_keys, &values);
+    for (const auto& s : statuses) {
+      ASSERT_OK(s);
+    }
    hist_mget_snapshot.Add(get_perf_context()->get_snapshot_time);
    hist_mget_memtable.Add(get_perf_context()->get_from_memtable_time);
    hist_mget_files.Add(get_perf_context()->get_from_output_files_time);
@@ -438,8 +444,9 @@ void ProfileQueries(bool enabled_time = false) {
  }

  if (FLAGS_verbose) {
-    std::cout << "ReadOnly Get uesr key comparison: \n" << hist_get.ToString()
-              << "ReadOnly MultiGet uesr key comparison: \n"
+    std::cout << "ReadOnly Get user key comparison: \n"
+              << hist_get.ToString()
+              << "ReadOnly MultiGet user key comparison: \n"
              << hist_mget.ToString();

    std::cout << "ReadOnly Get(): Time to get snapshot: \n"
@@ -539,7 +546,7 @@ TEST_F(PerfContextTest, SeekKeyComparison) {

    get_perf_context()->Reset();
    timer.Start();
-    db->Put(write_options, key, value);
+    ASSERT_OK(db->Put(write_options, key, value));
    auto put_time = timer.ElapsedNanos();
    hist_put_time.Add(put_time);
    hist_wal_time.Add(get_perf_context()->write_wal_time);
@@ -573,7 +580,7 @@ TEST_F(PerfContextTest, SeekKeyComparison) {
    iter->Next();
    hist_next.Add(get_perf_context()->user_key_comparison_count);
  }
-
+  ASSERT_OK(iter->status());
  if (FLAGS_verbose) {
    std::cout << "Seek:\n" << hist_seek.ToString() << "Next:\n"
              << hist_next.ToString();
@@ -835,7 +842,7 @@ TEST_F(PerfContextTest, CPUTimer) {
    std::string value = "v" + i_str;
    max_str = max_str > i_str ? max_str : i_str;

-    db->Put(write_options, key, value);
+    ASSERT_OK(db->Put(write_options, key, value));
  }
  std::string last_key = "k" + max_str;
  std::string last_value = "v" + max_str;
@@ -185,7 +185,7 @@ TEST_F(PeriodicWorkSchedulerTest, MultiInstances) {
  ASSERT_EQ(expected_run, pst_st_counter);

  for (int i = half; i < kInstanceNum; i++) {
-    dbs[i]->Close();
+    ASSERT_OK(dbs[i]->Close());
    delete dbs[i];
  }
 }
@@ -217,7 +217,7 @@ TEST_F(PeriodicWorkSchedulerTest, MultiEnv) {
  ASSERT_EQ(dbi->TEST_GetPeriodicWorkScheduler(),
            dbfull()->TEST_GetPeriodicWorkScheduler());

-  db->Close();
+  ASSERT_OK(db->Close());
  delete db;
  Close();
 }
@@ -311,7 +311,7 @@ TEST(SamePrefixTest, InDomainTest) {
    ASSERT_OK(db->Put(write_options, "HHKB pro2", "Mar 24, 2006"));
    ASSERT_OK(db->Put(write_options, "HHKB pro2 Type-S", "June 29, 2011"));
    ASSERT_OK(db->Put(write_options, "Realforce 87u", "idk"));
-    db->Flush(FlushOptions());
+    ASSERT_OK(db->Flush(FlushOptions()));
    std::string result;
    auto db_iter = db->NewIterator(ReadOptions());

@@ -331,7 +331,7 @@ TEST(SamePrefixTest, InDomainTest) {
    ASSERT_OK(db->Put(write_options, "pikachu", "1"));
    ASSERT_OK(db->Put(write_options, "Meowth", "1"));
    ASSERT_OK(db->Put(write_options, "Mewtwo", "idk"));
-    db->Flush(FlushOptions());
+    ASSERT_OK(db->Flush(FlushOptions()));
    std::string result;
    auto db_iter = db->NewIterator(ReadOptions());

@@ -351,7 +351,7 @@ TEST_F(PrefixTest, TestResult) {
      std::cout << "*** Mem table: " << options.memtable_factory->Name()
                << " number of buckets: " << num_buckets
                << std::endl;
-      DestroyDB(kDbName, Options());
+      ASSERT_OK(DestroyDB(kDbName, Options()));
      auto db = OpenDb();
      WriteOptions write_options;
      ReadOptions read_options;
@@ -528,7 +528,7 @@ TEST_F(PrefixTest, PrefixValid) {
    while (NextOptions(num_buckets)) {
      std::cout << "*** Mem table: " << options.memtable_factory->Name()
                << " number of buckets: " << num_buckets << std::endl;
-      DestroyDB(kDbName, Options());
+      ASSERT_OK(DestroyDB(kDbName, Options()));
      auto db = OpenDb();
      WriteOptions write_options;
      ReadOptions read_options;
@@ -543,7 +543,7 @@ TEST_F(PrefixTest, PrefixValid) {
      PutKey(db.get(), write_options, 12345, 8, v18);
      PutKey(db.get(), write_options, 12345, 9, v19);
      PutKey(db.get(), write_options, 12346, 8, v16);
-      db->Flush(FlushOptions());
+      ASSERT_OK(db->Flush(FlushOptions()));
      TestKey test_key(12346, 8);
      std::string s;
      ASSERT_OK(db->Delete(write_options, TestKeyToSlice(s, test_key)));
@@ -581,7 +581,7 @@ TEST_F(PrefixTest, DynamicPrefixIterator) {
  while (NextOptions(FLAGS_bucket_count)) {
    std::cout << "*** Mem table: " << options.memtable_factory->Name()
        << std::endl;
-    DestroyDB(kDbName, Options());
+    ASSERT_OK(DestroyDB(kDbName, Options()));
    auto db = OpenDb();
    WriteOptions write_options;
    ReadOptions read_options;
@@ -689,7 +689,7 @@ TEST_F(PrefixTest, PrefixSeekModePrev) {
  for (size_t m = 1; m < 100; m++) {
    std::cout << "[" + std::to_string(m) + "]" + "*** Mem table: "
              << options.memtable_factory->Name() << std::endl;
-    DestroyDB(kDbName, Options());
+    ASSERT_OK(DestroyDB(kDbName, Options()));
    auto db = OpenDb();
    WriteOptions write_options;
    ReadOptions read_options;
@@ -714,7 +714,7 @@ TEST_F(PrefixTest, PrefixSeekModePrev) {
        }
      }
      if (i < 2) {
-        db->Flush(FlushOptions());
+        ASSERT_OK(db->Flush(FlushOptions()));
      }
    }