Fix the logic of setting read_amp_bytes_per_bit from OPTIONS file (#7680 )

Summary: Instead of using `EncodeFixed32` which always serialize a integer to little endian, we should use the local machine's endianness when populating a native data structure during options parsing. Without this fix, `read_amp_bytes_per_bit` may be populated incorrectly on big-endian machines. Pull Request resolved: https://github.com/facebook/rocksdb/pull/7680 Test Plan: make check Reviewed By: pdillinger Differential Revision: D24999166 Pulled By: riversand963 fbshipit-source-id: dc603cff6e17f8fa32479ce6df93b93082e6b0c4
Bump version and update HISTORY.md
2026-11-06 11:02:36 -05:00 · 2020-11-17 10:27:16 -08:00 · 2020-11-15 15:13:05 -08:00 · 2020-11-15 15:10:54 -08:00 · 2020-11-15 15:08:05 -08:00 · 2020-10-14 10:55:43 -07:00
440 changed files with 23877 additions and 7725 deletions
@@ -0,0 +1,54 @@
+#! /bin/bash
+
+# Work around issue with parallel make output causing random error, as in
+# make[1]: write error: stdout
+# Probably due to a kernel bug:
+#   https://bugs.launchpad.net/ubuntu/+source/linux-signed/+bug/1814393
+# Seems to affect image ubuntu-1604:201903-01 and ubuntu-1604:202004-01
+
+cd "$(dirname $0)"
+
+if [ ! -x cat_ignore_eagain.out ]; then
+    cc -x c -o cat_ignore_eagain.out - << EOF
+#include <unistd.h>
+#include <errno.h>
+#include <stdio.h>
+int main() {
+    int n, m, p;
+    char buf[1024];
+    for (;;) {
+        n = read(STDIN_FILENO, buf, 1024);
+        if (n > 0 && n <= 1024) {
+            for (m = 0; m < n;) {
+                p = write(STDOUT_FILENO, buf + m, n - m);
+                if (p < 0) {
+                    if (errno == EAGAIN) {
+                        // ignore but pause a bit
+                        usleep(100);
+                    } else {
+                        perror("write failed");
+                        return 42;
+                    }
+                } else {
+                    m += p;
+                }
+            }
+        } else if (n < 0) {
+            if (errno == EAGAIN) {
+                // ignore but pause a bit
+                usleep(100);
+            } else {
+                // Some non-ignorable error
+                perror("read failed");
+                return 43;
+            }
+        } else {
+            // EOF
+            return 0;
+        }
+    }
+}
+EOF
+fi
+
+exec ./cat_ignore_eagain.out
@@ -6,25 +6,191 @@ orbs:
 executors:
  windows-2xlarge:
    machine:
-      image: 'windows-server-2019-vs2019:stable'
+      image: 'windows-server-2019-vs2019:201908-06'
      resource_class: windows.2xlarge
      shell: bash.exe

 jobs:
-  build:
+  build-linux:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: sudo apt-get update -y && sudo apt-get install -y libgflags-dev
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 PRINT_PARALLEL_OUTPUTS=1 GTEST_THROW_ON_FAILURE=0 GTEST_OUTPUT="xml:/tmp/test-results/" make V=1 J=32 -j32 check | .circleci/cat_ignore_eagain
+      - store_test_results:
+          path: /tmp/test-results
+
+  build-linux-shared_lib-alt_namespace-status_checked:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: sudo apt-get update -y && sudo apt-get install -y libgflags-dev
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 PRINT_PARALLEL_OUTPUTS=1 ASSERT_STATUS_CHECKED=1 LIB_MODE=shared OPT="-DROCKSDB_NAMESPACE=alternative_rocksdb_ns" GTEST_THROW_ON_FAILURE=0 GTEST_OUTPUT="xml:/tmp/test-results/" make V=1 -j32 all check_some | .circleci/cat_ignore_eagain
+      - store_test_results:
+          path: /tmp/test-results
+
+  build-linux-release:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: make V=1 -j32 release | .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 V=1 -j32 release | .circleci/cat_ignore_eagain
+      - run: ./db_stress --version # ensure with gflags
+
+  build-linux-lite:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: sudo apt-get update -y && sudo apt-get install -y libgflags-dev
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 PRINT_PARALLEL_OUTPUTS=1 LITE=1 GTEST_THROW_ON_FAILURE=0 GTEST_OUTPUT="xml:/tmp/test-results/" make V=1 J=32 -j32 check | .circleci/cat_ignore_eagain
+      - store_test_results:
+          path: /tmp/test-results
+
+  build-linux-lite-release:
+    machine:
+      image: ubuntu-1604:201903-01
+    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: 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: LITE=1 make V=1 -j32 release | .circleci/cat_ignore_eagain
+      - run: ./db_stress --version # ensure with gflags
+
+  build-linux-clang-no-test:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    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
+
+  build-linux-clang10-no-test:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: echo "deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: echo "deb-src http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: sudo apt-get update -y && sudo apt-get install -y clang-10 libgflags-dev
+      - run: CC=clang-10 CXX=clang++-10 ROCKSDB_DISABLE_ALIGNED_NEW=1 USE_CLANG=1 make V=1 -j32 all | .circleci/cat_ignore_eagain # aligned new doesn't work for reason we haven't figured out
+
+  build-linux-clang10-asan:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: echo "deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: echo "deb-src http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: sudo apt-get update -y && sudo apt-get install -y clang-10 libgflags-dev
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 COMPILE_WITH_ASAN=1 CC=clang-10 CXX=clang++-10 ROCKSDB_DISABLE_ALIGNED_NEW=1 USE_CLANG=1 PRINT_PARALLEL_OUTPUTS=1 GTEST_THROW_ON_FAILURE=0 GTEST_OUTPUT="xml:/tmp/test-results/" make V=1 -j32 check | .circleci/cat_ignore_eagain # aligned new doesn't work for reason we haven't figured out
+      - store_test_results:
+          path: /tmp/test-results
+
+  build-linux-clang10-mini-tsan:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: echo "deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: echo "deb-src http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: sudo apt-get update -y && sudo apt-get install -y clang-10 libgflags-dev
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 COMPILE_WITH_TSAN=1 CC=clang-10 CXX=clang++-10 ROCKSDB_DISABLE_ALIGNED_NEW=1 USE_CLANG=1 PRINT_PARALLEL_OUTPUTS=1 EXCLUDE_TESTS_REGEX="TransactionStressTest|SnapshotConcurrentAccess|SeqAdvanceConcurrent|DeadlockStress|MultiThreadedDBTest.MultiThreaded|WriteUnpreparedStressTest.ReadYourOwnWriteStress|DBAsBaseDB/TransactionStressTest|FlushCloseWALFiles|BackgroundPurgeCFDropTest" GTEST_THROW_ON_FAILURE=0 GTEST_OUTPUT="xml:/tmp/test-results/" make V=1 -j32 check | .circleci/cat_ignore_eagain # aligned new doesn't work for reason we haven't figured out. Exclude FlushCloseWALFiles and BackgroundPurgeCFDropTest for occasional failures.
+      - store_test_results:
+          path: /tmp/test-results
+
+  build-linux-clang10-ubsan:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: echo "deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: echo "deb-src http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: sudo apt-get update -y && sudo apt-get install -y clang-10 libgflags-dev
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 COMPILE_WITH_UBSAN=1 OPT="-fsanitize-blacklist=.circleci/ubsan_suppression_list.txt" CC=clang-10 CXX=clang++-10 ROCKSDB_DISABLE_ALIGNED_NEW=1 USE_CLANG=1 PRINT_PARALLEL_OUTPUTS=1 make V=1 -j32 ubsan_check | .circleci/cat_ignore_eagain # aligned new doesn't work for reason we haven't figured out
+
+  build-linux-clang10-clang-analyze:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: echo "deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: echo "deb-src http://apt.llvm.org/xenial/ llvm-toolchain-xenial-10 main" | sudo tee -a /etc/apt/sources.list
+      - run: sudo apt-get update -y && sudo apt-get install -y clang-10 libgflags-dev clang-tools-10
+      - run: SKIP_FORMAT_BUCK_CHECKS=1 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 PRINT_PARALLEL_OUTPUTS=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.
+
+  build-linux-cmake:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    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
+
+  build-linux-unity:
+    docker: # executor type
+      - image: gcc:latest
+    resource_class: xlarge
+    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
+
+  build-windows:
    executor: windows-2xlarge
-    
+    parameters:
+      extra_cmake_opt:
+        default: ""
+        type: string
+      vs_year:
+        default: "2019"
+        type: string
+      cmake_generator:
+        default: "Visual Studio 16 2019"
+        type: string
    environment:
      THIRDPARTY_HOME: C:/Users/circleci/thirdparty
      CMAKE_HOME: C:/Users/circleci/thirdparty/cmake-3.16.4-win64-x64
      CMAKE_BIN: C:/Users/circleci/thirdparty/cmake-3.16.4-win64-x64/bin/cmake.exe
-      CMAKE_GENERATOR: Visual Studio 16 2019
      SNAPPY_HOME: C:/Users/circleci/thirdparty/snappy-1.1.7
      SNAPPY_INCLUDE: C:/Users/circleci/thirdparty/snappy-1.1.7;C:/Users/circleci/thirdparty/snappy-1.1.7/build
      SNAPPY_LIB_DEBUG: C:/Users/circleci/thirdparty/snappy-1.1.7/build/Debug/snappy.lib
-
+      VS_YEAR: <<parameters.vs_year>>
+      CMAKE_GENERATOR: <<parameters.cmake_generator>>
    steps:
      - checkout
+      - run:
+          name: "Setup VS"
+          command: |
+            if [[ "${VS_YEAR}" == "2017" ]]; then
+              powershell .circleci/vs2017_install.ps1
+            elif [[ "${VS_YEAR}" == "2015" ]]; then
+              powershell .circleci/vs2015_install.ps1
+            fi
      - run:
          name: "Install thirdparty dependencies"
          command: |
@@ -46,7 +212,7 @@ jobs:
          command: |
            mkdir build
            cd build
-            ${CMAKE_BIN} -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DJNI=1 ..
+            ${CMAKE_BIN} -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DJNI=1 << parameters.extra_cmake_opt >> ..
            cd ..
            msbuild.exe build/rocksdb.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
      - run:
@@ -54,3 +220,98 @@ jobs:
          shell: powershell.exe
          command: |
            build_tools\run_ci_db_test.ps1 -SuiteRun db_basic_test,db_test,db_test2,env_basic_test,env_test,db_merge_operand_test -Concurrency 16
+
+  build-linux-java:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    steps:
+      - checkout
+      - run: pyenv global 3.5.2
+      - run: sudo apt-get update -y && sudo apt-get install -y libgflags-dev
+      - run:
+          name: "Build RocksDBJava"
+          command: |
+            export JAVA_HOME=/usr/lib/jvm/jdk1.8.0
+            export PATH=$JAVA_HOME/bin:$PATH
+            echo "JAVA_HOME=${JAVA_HOME}"
+            which java && java -version
+            which javac && javac -version
+            SKIP_FORMAT_BUCK_CHECKS=1 PRINT_PARALLEL_OUTPUTS=1 make V=1 J=32 -j32 rocksdbjava jtest | .circleci/cat_ignore_eagain
+
+  build-examples:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: medium
+    steps:
+      - checkout # check out the code in the project directory
+      - run: pyenv global 3.5.2
+      - run: sudo apt-get update -y && sudo apt-get install -y libgflags-dev
+      - run:
+          name: "Build examples"
+          command: |
+            OPT=-DTRAVIS V=1 make -j4 static_lib && cd examples && make -j4 | ../.circleci/cat_ignore_eagain
+
+workflows:
+  build-linux:
+    jobs:
+      - build-linux
+  build-linux-shared_lib-alt_namespace-status_checked:
+    jobs:
+      - build-linux-shared_lib-alt_namespace-status_checked
+  build-linux-lite:
+    jobs:
+      - build-linux-lite
+  build-linux-release:
+    jobs:
+      - build-linux-release
+  build-linux-lite-release:
+    jobs:
+      - build-linux-lite-release
+  build-linux-clang-no-test:
+    jobs:
+      - build-linux-clang-no-test
+  build-linux-clang10-no-test:
+    jobs:
+      - build-linux-clang10-no-test
+  build-linux-clang10-asan:
+    jobs:
+      - build-linux-clang10-asan
+  build-linux-clang10-mini-tsan:
+    jobs:
+      - build-linux-clang10-mini-tsan
+  build-linux-clang10-ubsan:
+    jobs:
+      - build-linux-clang10-ubsan
+  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
+  build-windows:
+    jobs:
+      - build-windows
+  build-windows-vs2017:
+    jobs:
+      - build-windows:
+          vs_year: "2017"
+          cmake_generator: "Visual Studio 15 Win64"
+  build-windows-vs2015:
+    jobs:
+      - build-windows:
+          vs_year: "2015"
+          cmake_generator: "Visual Studio 14 Win64"
+  build-windows-cxx20:
+    jobs:
+      - build-windows:
+          extra_cmake_opt: -DCMAKE_CXX_STANDARD=20
+  build-java:
+    jobs:
+      - build-linux-java
+  build-examples:
+    jobs:
+      - build-examples
@@ -0,0 +1,6 @@
+# Supress UBSAN warnings related to stl_tree.h, e.g.
+# UndefinedBehaviorSanitizer: undefined-behavior /usr/bin/../lib/gcc/x86_64-linux-gnu/5.4.0/../../../../include/c++/5.4.0/bits/stl_tree.h:1505:43 in 
+# /usr/bin/../lib/gcc/x86_64-linux-gnu/5.4.0/../../../../include/c++/5.4.0/bits/stl_tree.h:1505:43:
+# runtime error: upcast of address 0x000001fa8820 with insufficient space for an object of type
+# 'std::_Rb_tree_node<std::pair<const std::__cxx11::basic_string<char>, rocksdb::(anonymous namespace)::LockHoldingInfo> >'
+src:*bits/stl_tree.h
@@ -0,0 +1,23 @@
+$VS_DOWNLOAD_LINK = "https://go.microsoft.com/fwlink/?LinkId=691126"
+$COLLECT_DOWNLOAD_LINK = "https://aka.ms/vscollect.exe"
+curl.exe --retry 3 -kL $VS_DOWNLOAD_LINK --output vs_installer.exe
+if ($LASTEXITCODE -ne 0) {
+    echo "Download of the VS 2015 installer failed"
+    exit 1
+}
+$VS_INSTALL_ARGS = @("/Quiet", "/NoRestart")
+$process = Start-Process "${PWD}\vs_installer.exe" -ArgumentList $VS_INSTALL_ARGS -NoNewWindow -Wait -PassThru
+Remove-Item -Path vs_installer.exe -Force
+$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]."
+    curl.exe --retry 3 -kL $COLLECT_DOWNLOAD_LINK --output Collect.exe
+    if ($LASTEXITCODE -ne 0) {
+        echo "Download of the VS Collect tool failed."
+        exit 1
+    }
+    Start-Process "${PWD}\Collect.exe" -NoNewWindow -Wait -PassThru
+    New-Item -Path "C:\w\build-results" -ItemType "directory" -Force
+    Copy-Item -Path "C:\Users\circleci\AppData\Local\Temp\vslogs.zip" -Destination "C:\w\build-results\"
+    exit 1
+}
@@ -0,0 +1,34 @@
+$VS_DOWNLOAD_LINK = "https://aka.ms/vs/15/release/vs_buildtools.exe"
+$COLLECT_DOWNLOAD_LINK = "https://aka.ms/vscollect.exe"
+$VS_INSTALL_ARGS = @("--nocache","--quiet","--wait", "--add Microsoft.VisualStudio.Workload.VCTools",
+                                                     "--add Microsoft.VisualStudio.Component.VC.Tools.14.13",
+                                                     "--add Microsoft.Component.MSBuild",
+                                                     "--add Microsoft.VisualStudio.Component.Roslyn.Compiler",
+                                                     "--add Microsoft.VisualStudio.Component.TextTemplating",
+                                                     "--add Microsoft.VisualStudio.Component.VC.CoreIde",
+                                                     "--add Microsoft.VisualStudio.Component.VC.Redist.14.Latest",
+                                                     "--add Microsoft.VisualStudio.ComponentGroup.NativeDesktop.Core",
+                                                     "--add Microsoft.VisualStudio.Component.VC.Tools.x86.x64",
+                                                     "--add Microsoft.VisualStudio.ComponentGroup.NativeDesktop.Win81")
+
+curl.exe --retry 3 -kL $VS_DOWNLOAD_LINK --output vs_installer.exe
+if ($LASTEXITCODE -ne 0) {
+    echo "Download of the VS 2017 installer failed"
+    exit 1
+}
+
+$process = Start-Process "${PWD}\vs_installer.exe" -ArgumentList $VS_INSTALL_ARGS -NoNewWindow -Wait -PassThru
+Remove-Item -Path vs_installer.exe -Force
+$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]."
+    curl.exe --retry 3 -kL $COLLECT_DOWNLOAD_LINK --output Collect.exe
+    if ($LASTEXITCODE -ne 0) {
+        echo "Download of the VS Collect tool failed."
+        exit 1
+    }
+    Start-Process "${PWD}\Collect.exe" -NoNewWindow -Wait -PassThru
+    New-Item -Path "C:\w\build-results" -ItemType "directory" -Force
+    Copy-Item -Path "C:\Users\circleci\AppData\Local\Temp\vslogs.zip" -Destination "C:\w\build-results\"
+    exit 1
+}
@@ -35,7 +35,7 @@ jobs:
        args: https://raw.githubusercontent.com/llvm-mirror/clang/master/tools/clang-format/clang-format-diff.py

    - name: Check format
-      run: make check-format
+      run: VERBOSE_CHECK=1 make check-format

    - name: Compare buckify output
      run: make check-buck-targets
@@ -85,3 +85,5 @@ buckifier/*.pyc
 buckifier/__pycache__

 compile_commands.json
+clang-format-diff.py
+.py3/
@@ -53,6 +53,8 @@ env:
  - JOB_NAME=cmake-gcc9 # 3-5 minutes
  - JOB_NAME=cmake-gcc9-c++20 # 3-5 minutes
  - JOB_NAME=cmake-mingw # 3 minutes
+  - JOB_NAME=make-gcc4.8
+  - JOB_NAME=status_checked

 matrix:
  exclude:
@@ -64,6 +66,8 @@ matrix:
    env: JOB_NAME=cmake-gcc9-c++20
  - os: osx
    env: JOB_NAME=cmake-mingw
+  - os: osx
+    env: JOB_NAME=make-gcc4.8
  - os: osx
    arch: ppc64le
  - os: osx
@@ -71,9 +75,15 @@ matrix:
  - os : linux
    arch: arm64
    env: JOB_NAME=cmake-mingw
+  - os : linux
+    arch: arm64
+    env: JOB_NAME=make-gcc4.8
  - os: linux
    arch: ppc64le
    env: JOB_NAME=cmake-mingw
+  - os: linux
+    arch: ppc64le
+    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
@@ -197,6 +207,17 @@ matrix:
    os: linux
    arch: ppc64le
    env: JOB_NAME=cmake-gcc9-c++20
+  - if: type = pull_request
+    os : osx
+    env: JOB_NAME=status_checked
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=status_checked
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=status_checked

 install:
  - if [ "${TRAVIS_OS_NAME}" == osx ]; then
@@ -213,14 +234,13 @@ install:
  - if [ "${JOB_NAME}" == cmake-mingw ]; then
      sudo apt-get install -y mingw-w64 ;
    fi
+  - if [ "${JOB_NAME}" == make-gcc4.8 ]; then
+      sudo apt-get install -y g++-4.8 ;
+      CC=gcc-4.8 && CXX=g++-4.8;
+    fi
  - if [[ "${JOB_NAME}" == cmake* ]] && [ "${TRAVIS_OS_NAME}" == linux ]; then
-      CMAKE_DIST_URL="https://rocksdb-deps.s3-us-west-2.amazonaws.com/cmake/cmake-3.14.5-Linux-$(uname -m).tar.bz2";
-      TAR_OPT="--strip-components=1 -xj";
-      if [ "aarch64" == "$(uname -m)" ]; then
-        sudo apt-get install -y libuv1 librhash0;
-        sudo apt-get upgrade -y libstdc++6;
-      fi;
-      mkdir cmake-dist && curl --silent --fail --show-error --location "${CMAKE_DIST_URL}" | tar -C cmake-dist ${TAR_OPT} && export PATH=$PWD/cmake-dist/bin:$PATH;
+      sudo snap install cmake --beta --classic;
+      export PATH=/snap/bin:$PATH;
    fi
  - |
    if [[ "${JOB_NAME}" == java_test || "${JOB_NAME}" == cmake* ]]; then
@@ -258,7 +278,7 @@ script:
      OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" V=1 make -j4 tools && OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" V=1 ROCKSDBTESTS_START=db_iter_test ROCKSDBTESTS_END=options_file_test make -j4 check_some
      ;;
    3)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=options_file_test ROCKSDBTESTS_END=write_prepared_transaction_test make -j4 check_some
+      OPT=-DTRAVIS LIB_MODE=shared V=1 ROCKSDBTESTS_START=options_file_test ROCKSDBTESTS_END=write_prepared_transaction_test make -j4 check_some
      ;;
    4)
      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=write_prepared_transaction_test make -j4 check_some
@@ -269,7 +289,7 @@ script:
      OPT=-DTRAVIS V=1 make rocksdbjava jtest
      ;;
    lite_build)
-      OPT='-DTRAVIS -DROCKSDB_LITE' V=1 make -j4 static_lib tools
+      OPT='-DTRAVIS -DROCKSDB_LITE' V=1 make -j4 all
      ;;
    examples)
      OPT=-DTRAVIS V=1 make -j4 static_lib && cd examples && make -j4
@@ -285,7 +305,13 @@ script:
          ;;
      esac

-      mkdir build && cd build && cmake -DJNI=1 .. -DCMAKE_BUILD_TYPE=Release $OPT && make -j4 rocksdb rocksdbjni
+      mkdir build && cd build && cmake -DCMAKE_BUILD_TYPE=Release -DWITH_TESTS=0 -DWITH_GFLAGS=0 -DWITH_BENCHMARK_TOOLS=0 -DWITH_TOOLS=0 -DWITH_CORE_TOOLS=1 .. && make -j4 && cd  .. && rm -rf build && mkdir build && cd build && cmake -DJNI=1 .. -DCMAKE_BUILD_TYPE=Release $OPT && make -j4 rocksdb rocksdbjni
+      ;;
+    make-gcc4.8)
+      OPT=-DTRAVIS V=1 SKIP_LINK=1 make -j4 all && [ "Linking broken because libgflags compiled with newer ABI" ]
+      ;;
+    status_checked)
+      OPT=-DTRAVIS V=1 ASSERT_STATUS_CHECKED=1 make -j4 check_some
      ;;
    esac
 notifications:
@@ -95,7 +95,7 @@ if(MSVC)
  option(WITH_XPRESS "build with windows built in compression" OFF)
  include(${CMAKE_CURRENT_SOURCE_DIR}/thirdparty.inc)
 else()
-  if(CMAKE_SYSTEM_NAME MATCHES "FreeBSD")
+  if(CMAKE_SYSTEM_NAME MATCHES "FreeBSD" AND NOT CMAKE_SYSTEM_NAME MATCHES "kFreeBSD")
    # FreeBSD has jemalloc as default malloc
    # but it does not have all the jemalloc files in include/...
    set(WITH_JEMALLOC ON)
@@ -107,18 +107,35 @@ else()
    endif()
  endif()

-  # No config file for this
+  set(GFLAGS_LIB)
  if(WITH_GFLAGS)
-    find_package(gflags REQUIRED)
+    # Config with namespace available since gflags 2.2.2
+    option(GFLAGS_USE_TARGET_NAMESPACE "Use gflags import target with namespace." ON)
+    find_package(gflags CONFIG)
+    if(gflags_FOUND)
+      if(TARGET ${GFLAGS_TARGET})
+        # Config with GFLAGS_TARGET available since gflags 2.2.0
+        set(GFLAGS_LIB ${GFLAGS_TARGET})
+      else()
+        # Config with GFLAGS_LIBRARIES available since gflags 2.1.0
+        set(GFLAGS_LIB ${GFLAGS_LIBRARIES})
+      endif()
+    else()
+      find_package(gflags REQUIRED)
+        set(GFLAGS_LIB gflags::gflags)
+    endif()
+    include_directories(${GFLAGS_INCLUDE_DIR})
+    list(APPEND THIRDPARTY_LIBS ${GFLAGS_LIB})
    add_definitions(-DGFLAGS=1)
-    include_directories(${gflags_INCLUDE_DIR})
-    list(APPEND THIRDPARTY_LIBS gflags::gflags)
  endif()

  if(WITH_SNAPPY)
-    find_package(snappy REQUIRED)
+    find_package(Snappy CONFIG)
+    if(NOT Snappy_FOUND)
+      find_package(Snappy REQUIRED)
+    endif()
    add_definitions(-DSNAPPY)
-    list(APPEND THIRDPARTY_LIBS snappy::snappy)
+    list(APPEND THIRDPARTY_LIBS Snappy::snappy)
  endif()

  if(WITH_ZLIB)
@@ -434,6 +451,8 @@ elseif(CMAKE_SYSTEM_NAME MATCHES "Linux")
  add_definitions(-DOS_LINUX)
 elseif(CMAKE_SYSTEM_NAME MATCHES "SunOS")
  add_definitions(-DOS_SOLARIS)
+elseif(CMAKE_SYSTEM_NAME MATCHES "kFreeBSD")
+  add_definitions(-DOS_GNU_KFREEBSD)
 elseif(CMAKE_SYSTEM_NAME MATCHES "FreeBSD")
  add_definitions(-DOS_FREEBSD)
 elseif(CMAKE_SYSTEM_NAME MATCHES "NetBSD")
@@ -492,7 +511,11 @@ if(HAVE_PTHREAD_MUTEX_ADAPTIVE_NP)
 endif()

 include(CheckCXXSymbolExists)
-check_cxx_symbol_exists(malloc_usable_size malloc.h HAVE_MALLOC_USABLE_SIZE)
+if(CMAKE_SYSTEM_NAME MATCHES "^FreeBSD")
+  check_cxx_symbol_exists(malloc_usable_size malloc_np.h HAVE_MALLOC_USABLE_SIZE)
+else()
+  check_cxx_symbol_exists(malloc_usable_size malloc.h HAVE_MALLOC_USABLE_SIZE)
+endif()
 if(HAVE_MALLOC_USABLE_SIZE)
  add_definitions(-DROCKSDB_MALLOC_USABLE_SIZE)
 endif()
@@ -509,7 +532,6 @@ endif()

 include_directories(${PROJECT_SOURCE_DIR})
 include_directories(${PROJECT_SOURCE_DIR}/include)
-include_directories(SYSTEM ${PROJECT_SOURCE_DIR}/third-party/gtest-1.8.1/fused-src)
 if(WITH_FOLLY_DISTRIBUTED_MUTEX)
  include_directories(${PROJECT_SOURCE_DIR}/third-party/folly)
 endif()
@@ -526,6 +548,9 @@ set(SOURCES
        db/blob/blob_file_addition.cc
        db/blob/blob_file_garbage.cc
        db/blob/blob_file_meta.cc
+        db/blob/blob_log_format.cc
+        db/blob/blob_log_reader.cc
+        db/blob/blob_log_writer.cc
        db/builder.cc
        db/c.cc
        db/column_family.cc
@@ -537,6 +562,7 @@ set(SOURCES
        db/compaction/compaction_picker_fifo.cc
        db/compaction/compaction_picker_level.cc
        db/compaction/compaction_picker_universal.cc
+        db/compaction/sst_partitioner.cc
        db/convenience.cc
        db/db_filesnapshot.cc
        db/db_impl/db_impl.cc
@@ -676,6 +702,7 @@ set(SOURCES
        table/plain/plain_table_index.cc
        table/plain/plain_table_key_coding.cc
        table/plain/plain_table_reader.cc
+        table/sst_file_dumper.cc
        table/sst_file_reader.cc
        table/sst_file_writer.cc
        table/table_properties.cc
@@ -717,9 +744,6 @@ set(SOURCES
        utilities/blob_db/blob_db_impl_filesnapshot.cc
        utilities/blob_db/blob_dump_tool.cc
        utilities/blob_db/blob_file.cc
-        utilities/blob_db/blob_log_reader.cc
-        utilities/blob_db/blob_log_writer.cc
-        utilities/blob_db/blob_log_format.cc
        utilities/cassandra/cassandra_compaction_filter.cc
        utilities/cassandra/format.cc
        utilities/cassandra/merge_operator.cc
@@ -728,6 +752,8 @@ set(SOURCES
        utilities/debug.cc
        utilities/env_mirror.cc
        utilities/env_timed.cc
+        utilities/fault_injection_env.cc
+        utilities/fault_injection_fs.cc
        utilities/leveldb_options/leveldb_options.cc
        utilities/memory/memory_util.cc
        utilities/merge_operators/bytesxor.cc
@@ -790,9 +816,13 @@ if(WIN32)
    port/win/env_win.cc
    port/win/env_default.cc
    port/win/port_win.cc
-    port/win/win_logger.cc
-    port/win/win_thread.cc)
-
+    port/win/win_logger.cc)
+  if(NOT MINGW)
+    # Mingw only supports std::thread when using
+    # posix threads.
+    list(APPEND SOURCES
+      port/win/win_thread.cc)
+  endif()
 if(WITH_XPRESS)
  list(APPEND SOURCES
    port/win/xpress_win.cc)
@@ -839,12 +869,12 @@ else()
 endif()

 add_library(${ROCKSDB_STATIC_LIB} STATIC ${SOURCES})
-target_link_libraries(${ROCKSDB_STATIC_LIB}
+target_link_libraries(${ROCKSDB_STATIC_LIB} PRIVATE
  ${THIRDPARTY_LIBS} ${SYSTEM_LIBS})

 if(ROCKSDB_BUILD_SHARED)
  add_library(${ROCKSDB_SHARED_LIB} SHARED ${SOURCES})
-  target_link_libraries(${ROCKSDB_SHARED_LIB}
+  target_link_libraries(${ROCKSDB_SHARED_LIB} PRIVATE
    ${THIRDPARTY_LIBS} ${SYSTEM_LIBS})

  if(WIN32)
@@ -872,6 +902,16 @@ else()
 endif()

 option(WITH_JNI "build with JNI" OFF)
+# Tests are excluded from Release builds
+CMAKE_DEPENDENT_OPTION(WITH_TESTS "build with tests" ON
+  "CMAKE_BUILD_TYPE STREQUAL Debug" OFF)
+option(WITH_BENCHMARK_TOOLS "build with benchmarks" ON)
+option(WITH_CORE_TOOLS "build with ldb and sst_dump" ON)
+option(WITH_TOOLS "build with tools" ON)
+
+if(WITH_TESTS OR WITH_BENCHMARK_TOOLS OR WITH_TOOLS OR WITH_JNI OR JNI)
+  include_directories(SYSTEM ${PROJECT_SOURCE_DIR}/third-party/gtest-1.8.1/fused-src)
+endif()
 if(WITH_JNI OR JNI)
  message(STATUS "JNI library is enabled")
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/java)
@@ -909,6 +949,8 @@ if(NOT WIN32 OR ROCKSDB_INSTALL_ON_WINDOWS)

  install(DIRECTORY include/rocksdb COMPONENT devel DESTINATION "${CMAKE_INSTALL_INCLUDEDIR}")

+  install(DIRECTORY "${PROJECT_SOURCE_DIR}/cmake/modules" COMPONENT devel DESTINATION ${package_config_destination})
+
  install(
    TARGETS ${ROCKSDB_STATIC_LIB}
    EXPORT RocksDBTargets
@@ -945,16 +987,22 @@ if(NOT WIN32 OR ROCKSDB_INSTALL_ON_WINDOWS)
  )
 endif()

-# Tests are excluded from Release builds
-CMAKE_DEPENDENT_OPTION(WITH_TESTS "build with tests" ON
-  "CMAKE_BUILD_TYPE STREQUAL Debug" OFF)
-if(WITH_TESTS)
+option(WITH_ALL_TESTS "Build all test, rather than a small subset" ON)
+
+if(WITH_TESTS OR WITH_BENCHMARK_TOOLS)
  add_subdirectory(third-party/gtest-1.8.1/fused-src/gtest)
  add_library(testharness STATIC
  test_util/testharness.cc)
  target_link_libraries(testharness gtest)
-
+endif()
+  
+if(WITH_TESTS)
  set(TESTS
+        db/db_basic_test.cc
+        env/env_basic_test.cc
+  )
+  if(WITH_ALL_TESTS)
+    list(APPEND TESTS
        cache/cache_test.cc
        cache/lru_cache_test.cc
        db/blob/blob_file_addition_test.cc
@@ -969,7 +1017,6 @@ if(WITH_TESTS)
        db/comparator_db_test.cc
        db/corruption_test.cc
        db/cuckoo_table_db_test.cc
-        db/db_basic_test.cc
        db/db_with_timestamp_basic_test.cc
        db/db_block_cache_test.cc
        db/db_bloom_filter_test.cc
@@ -1032,7 +1079,6 @@ if(WITH_TESTS)
        db/write_batch_test.cc
        db/write_callback_test.cc
        db/write_controller_test.cc
-        env/env_basic_test.cc
        env/env_test.cc
        env/io_posix_test.cc
        env/mock_env_test.cc
@@ -1053,6 +1099,7 @@ if(WITH_TESTS)
        options/options_settable_test.cc
        options/options_test.cc
        table/block_based/block_based_filter_block_test.cc
+        table/block_based/block_based_table_reader_test.cc
        table/block_based/block_test.cc
        table/block_based/data_block_hash_index_test.cc
        table/block_based/full_filter_block_test.cc
@@ -1064,6 +1111,7 @@ if(WITH_TESTS)
        table/sst_file_reader_test.cc
        table/table_test.cc
        table/block_fetcher_test.cc
+        test_util/testutil_test.cc
        tools/block_cache_analyzer/block_cache_trace_analyzer_test.cc
        tools/ldb_cmd_test.cc
        tools/reduce_levels_test.cc
@@ -1113,7 +1161,8 @@ if(WITH_TESTS)
        utilities/transactions/write_unprepared_transaction_test.cc
        utilities/ttl/ttl_test.cc
        utilities/write_batch_with_index/write_batch_with_index_test.cc
-  )
+    )
+  endif()
  if(WITH_LIBRADOS)
    list(APPEND TESTS utilities/env_librados_test.cc)
  endif()
@@ -1126,8 +1175,6 @@ if(WITH_TESTS)
      db/db_test_util.cc
      monitoring/thread_status_updater_debug.cc
      table/mock_table.cc
-      test_util/fault_injection_test_env.cc
-      test_util/fault_injection_test_fs.cc
      utilities/cassandra/test_utils.cc
  )
  enable_testing()
@@ -1153,11 +1200,15 @@ if(WITH_TESTS)
        EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
        OUTPUT_NAME ${exename}${ARTIFACT_SUFFIX}
      )
-      target_link_libraries(${CMAKE_PROJECT_NAME}_${exename}${ARTIFACT_SUFFIX} testutillib${ARTIFACT_SUFFIX} testharness gtest ${ROCKSDB_LIB})
+      target_link_libraries(${CMAKE_PROJECT_NAME}_${exename}${ARTIFACT_SUFFIX} testutillib${ARTIFACT_SUFFIX} testharness gtest ${GFLAGS_LIB} ${ROCKSDB_LIB})
      if(NOT "${exename}" MATCHES "db_sanity_test")
        add_test(NAME ${exename} COMMAND ${exename}${ARTIFACT_SUFFIX})
        add_dependencies(check ${CMAKE_PROJECT_NAME}_${exename}${ARTIFACT_SUFFIX})
      endif()
+      if("${exename}" MATCHES "env_librados_test")
+        # env_librados_test.cc uses librados directly
+        target_link_libraries(${CMAKE_PROJECT_NAME}_${exename}${ARTIFACT_SUFFIX} rados)
+      endif()
  endforeach(sourcefile ${TESTS})

  if(WIN32)
@@ -1181,47 +1232,44 @@ if(WITH_TESTS)
  endif()
 endif()

-option(WITH_BENCHMARK_TOOLS "build with benchmarks" ON)
 if(WITH_BENCHMARK_TOOLS)
  add_executable(db_bench
    tools/db_bench.cc
    tools/db_bench_tool.cc)
  target_link_libraries(db_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${GFLAGS_LIB})

  add_executable(cache_bench
    cache/cache_bench.cc)
  target_link_libraries(cache_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${GFLAGS_LIB})

  add_executable(memtablerep_bench
    memtable/memtablerep_bench.cc)
  target_link_libraries(memtablerep_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${GFLAGS_LIB})

  add_executable(range_del_aggregator_bench
    db/range_del_aggregator_bench.cc)
  target_link_libraries(range_del_aggregator_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${GFLAGS_LIB})

  add_executable(table_reader_bench
    table/table_reader_bench.cc)
  target_link_libraries(table_reader_bench
-    ${ROCKSDB_LIB} testharness)
+    ${ROCKSDB_LIB} testharness ${GFLAGS_LIB})

  add_executable(filter_bench
    util/filter_bench.cc)
  target_link_libraries(filter_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${GFLAGS_LIB})

  add_executable(hash_table_bench
    utilities/persistent_cache/hash_table_bench.cc)
  target_link_libraries(hash_table_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${GFLAGS_LIB})
 endif()

-option(WITH_CORE_TOOLS "build with ldb and sst_dump" ON)
-option(WITH_TOOLS "build with tools" ON)
 if(WITH_CORE_TOOLS OR WITH_TOOLS)
  add_subdirectory(tools)
  add_custom_target(core_tools
@@ -1233,3 +1281,8 @@ if(WITH_TOOLS)
  add_custom_target(tools
    DEPENDS ${tool_deps})
 endif()
+
+option(WITH_EXAMPLES "build with examples" OFF)
+if(WITH_EXAMPLES)
+  add_subdirectory(examples)
+endif()
@@ -1,11 +1,112 @@
 # Rocksdb Change Log
-## Unreleased
+## 6.12.8 (2020-11-15)
+### Bug Fixes
+* Fix a bug of encoding and parsing BlockBasedTableOptions::read_amp_bytes_per_bit as a 64-bit integer.
+* 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.
+
+## 6.12.7 (2020-10-14)
+### Other
+Fix build issue to enable RocksJava release for ppc64le
+
+## 6.12.6 (2020-10-13)
+### Bug Fixes
+* Fix false positive flush/compaction `Status::Corruption` failure when `paranoid_file_checks == true` and range tombstones were written to the compaction output files.
+
+## 6.12.5 (2020-10-12)
+### Bug Fixes
+* Since 6.12, memtable lookup should report unrecognized value_type as corruption (#7121).
+* 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.
+* Fixed a bug when indexes are partitioned (`index_type == kTwoLevelIndexSearch`), some index partitions are pinned in memory (`BlockBasedTableOptions::pin_l0_filter_and_index_blocks_in_cache`), and partitions reads could be mixed between block cache and directly from the file (e.g., with `enable_index_compression == 1` and `mmap_read == 1`, partitions that were stored uncompressed due to poor compression ratio would be read directly from the file via mmap, while partitions that were stored compressed would be read from block cache). The bug could cause index partitions to be mistakenly considered empty during reads leading to wrong read results.
+
+## 6.12.4 (2020-09-18)
+### Public API Change
+* Reworked `BackupableDBOptions::share_files_with_checksum_naming` (new in 6.12) with some minor improvements and to better support those who were extracting files sizes from backup file names.
+
+## 6.12.3 (2020-09-16)
+### Bug fixes
+* Fixed a bug in size-amp-triggered and periodic-triggered universal compaction, where the compression settings for the first input level were used rather than the compression settings for the output (bottom) level.
+
+## 6.12.2 (2020-09-14)
+### Public API Change
+* BlobDB now exposes the start of the expiration range of TTL blob files via the `GetLiveFilesMetaData` API.
+
+## 6.12.1 (2020-08-20)
+### Bug fixes
+* BackupEngine::CreateNewBackup could fail intermittently with non-OK status when backing up a read-write DB configured with a DBOptions::file_checksum_gen_factory. This issue has been worked-around such that CreateNewBackup should succeed, but (until fully fixed) BackupEngine might not see all checksums available in the DB.
+
+## 6.12 (2020-07-28)
+### Public API Change
+* Encryption file classes now exposed for inheritance in env_encryption.h
+* File I/O listener is extended to cover more I/O operations. Now class `EventListener` in listener.h contains new callback functions: `OnFileFlushFinish()`, `OnFileSyncFinish()`, `OnFileRangeSyncFinish()`, `OnFileTruncateFinish()`, and ``OnFileCloseFinish()``.
+* `FileOperationInfo` now reports `duration` measured by `std::chrono::steady_clock` and `start_ts` measured by `std::chrono::system_clock` instead of start and finish timestamps measured by `system_clock`. Note that `system_clock` is called before `steady_clock` in program order at operation starts.
+* `DB::GetDbSessionId(std::string& session_id)` is added. `session_id` stores a unique identifier that gets reset every time the DB is opened. This DB session ID should be unique among all open DB instances on all hosts, and should be unique among re-openings of the same or other DBs. This identifier is recorded in the LOG file on the line starting with "DB Session ID:".
+* `DB::OpenForReadOnly()` now returns `Status::NotFound` when the specified DB directory does not exist. Previously the error returned depended on the underlying `Env`. This change is available in all 6.11 releases as well.
+* A parameter `verify_with_checksum` is added to `BackupEngine::VerifyBackup`, which is false by default. If it is ture, `BackupEngine::VerifyBackup` verifies checksums and file sizes of backup files. Pass `false` for `verify_with_checksum` to maintain the previous behavior and performance of `BackupEngine::VerifyBackup`, by only verifying sizes of backup files.
+
+
+### Behavior Changes
+* Best-efforts recovery ignores CURRENT file completely. If CURRENT file is missing during recovery, best-efforts recovery still proceeds with MANIFEST file(s).
+* In best-efforts recovery, an error that is not Corruption or IOError::kNotFound or IOError::kPathNotFound will be overwritten silently. Fix this by checking all non-ok cases and return early.
+* When `file_checksum_gen_factory` is set to `GetFileChecksumGenCrc32cFactory()`, BackupEngine will compare the crc32c checksums of table files computed when creating a backup to the expected checksums stored in the DB manifest, and will fail `CreateNewBackup()` on mismatch (corruption). If the `file_checksum_gen_factory` is not set or set to any other customized factory, there is no checksum verification to detect if SST files in a DB are corrupt when read, copied, and independently checksummed by BackupEngine.
+* When a DB sets `stats_dump_period_sec > 0`, either as the initial value for DB open or as a dynamic option change, the first stats dump is staggered in the following X seconds, where X is an integer in `[0, stats_dump_period_sec)`. Subsequent stats dumps are still spaced `stats_dump_period_sec` seconds apart.
+* When the paranoid_file_checks option is true, a hash is generated of all keys and values are generated when the SST file is written, and then the values are read back in to validate the file.  A corruption is signaled if the two hashes do not match.
+
+### Bug fixes
+* Compressed block cache was automatically disabled with read-only DBs by mistake. Now it is fixed: compressed block cache will be in effective with read-only DB too.
+* Fix a bug of wrong iterator result if another thread finishes an update and a DB flush between two statement.
+* Disable file deletion after MANIFEST write/sync failure until db re-open or Resume() so that subsequent re-open will not see MANIFEST referencing deleted SSTs.
+* Fix a bug when index_type == kTwoLevelIndexSearch in PartitionedIndexBuilder to update FlushPolicy to point to internal key partitioner when it changes from user-key mode to internal-key mode in index partition.
+* Make compaction report InternalKey corruption while iterating over the input.
+* Fix a bug which may cause MultiGet to be slow because it may read more data than requested, but this won't affect correctness. The bug was introduced in 6.10 release.
+* Fail recovery and report once hitting a physical log record checksum mismatch, while reading MANIFEST. RocksDB should not continue processing the MANIFEST any further.
+
+### New Features
+* DB identity (`db_id`) and DB session identity (`db_session_id`) are added to table properties and stored in SST files. SST files generated from SstFileWriter and Repairer have DB identity “SST Writer” and “DB Repairer”, respectively. Their DB session IDs are generated in the same way as `DB::GetDbSessionId`. The session ID for SstFileWriter (resp., Repairer) resets every time `SstFileWriter::Open` (resp., `Repairer::Run`) is called.
+* Added experimental option BlockBasedTableOptions::optimize_filters_for_memory for reducing allocated memory size of Bloom filters (~10% savings with Jemalloc) while preserving the same general accuracy. To have an effect, the option requires format_version=5 and malloc_usable_size. Enabling this option is forward and backward compatible with existing format_version=5.
+* `BackupableDBOptions::share_files_with_checksum_naming` is added with new default behavior for naming backup files with `share_files_with_checksum`, to address performance and backup integrity issues. See API comments for details.
+* Added auto resume function to automatically recover the DB from background Retryable IO Error. When retryable IOError happens during flush and WAL write, the error is mapped to Hard Error and DB will be in read mode. When retryable IO Error happens during compaction, the error will be mapped to Soft Error. DB is still in write/read mode. Autoresume function will create a thread for a DB to call DB->ResumeImpl() to try the recover for Retryable IO Error during flush and WAL write. Compaction will be rescheduled by itself if retryable IO Error happens. Auto resume may also cause other Retryable IO Error during the recovery, so the recovery will fail. Retry the auto resume may solve the issue, so we use max_bgerror_resume_count to decide how many resume cycles will be tried in total. If it is <=0, auto resume retryable IO Error is disabled. Default is INT_MAX, which will lead to a infinit auto resume. bgerror_resume_retry_interval decides the time interval between two auto resumes.
+* Option `max_subcompactions` can be set dynamically using DB::SetDBOptions().
+* Added experimental ColumnFamilyOptions::sst_partitioner_factory to define determine the partitioning of sst files. This helps compaction to split the files on interesting boundaries (key prefixes) to make propagation of sst files less write amplifying (covering the whole key space).
+
+### Performance Improvements
+* Eliminate key copies for internal comparisons while accessing ingested block-based tables.
+* Reduce key comparisons during random access in all block-based tables.
+* BackupEngine avoids unnecessary repeated checksum computation for backing up a table file to the `shared_checksum` directory when using `share_files_with_checksum_naming = kUseDbSessionId` (new default), except on SST files generated before this version of RocksDB, which fall back on using `kLegacyCrc32cAndFileSize`.
+
+## 6.11 (6/12/2020)
 ### Bug Fixes
 * Fix consistency checking error swallowing in some cases when options.force_consistency_checks = true.
+* Fix possible false NotFound status from batched MultiGet using index type kHashSearch.
+* Fix corruption caused by enabling delete triggered compaction (NewCompactOnDeletionCollectorFactory) in universal compaction mode, along with parallel compactions. The bug can result in two parallel compactions picking the same input files, resulting in the DB resurrecting older and deleted versions of some keys.
+* Fix a use-after-free bug in best-efforts recovery. column_family_memtables_ needs to point to valid ColumnFamilySet.
+* Let best-efforts recovery ignore corrupted files during table loading.
+* Fix corrupt key read from ingested file when iterator direction switches from reverse to forward at a key that is a prefix of another key in the same file. It is only possible in files with a non-zero global seqno.
+* Fix abnormally large estimate from GetApproximateSizes when a range starts near the end of one SST file and near the beginning of another. Now GetApproximateSizes consistently and fairly includes the size of SST metadata in addition to data blocks, attributing metadata proportionally among the data blocks based on their size.
+* Fix potential file descriptor leakage in PosixEnv's IsDirectory() and NewRandomAccessFile().
+* Fix false negative from the VerifyChecksum() API when there is a checksum mismatch in an index partition block in a BlockBasedTable format table file (index_type is kTwoLevelIndexSearch).
+* Fix sst_dump to return non-zero exit code if the specified file is not a recognized SST file or fails requested checks.
+* Fix incorrect results from batched MultiGet for duplicate keys, when the duplicate key matches the largest key of an SST file and the value type for the key in the file is a merge value.

 ### Public API Change
 * Flush(..., column_family) may return Status::ColumnFamilyDropped() instead of Status::InvalidArgument() if column_family is dropped while processing the flush request.
 * BlobDB now explicitly disallows using the default column family's storage directories as blob directory.
+* DeleteRange now returns `Status::InvalidArgument` if the range's end key comes before its start key according to the user comparator. Previously the behavior was undefined.
+* ldb now uses options.force_consistency_checks = true by default and "--disable_consistency_checks" is added to disable it.
+* DB::OpenForReadOnly no longer creates files or directories if the named DB does not exist, unless create_if_missing is set to true.
+* The consistency checks that validate LSM state changes (table file additions/deletions during flushes and compactions) are now stricter, more efficient, and no longer optional, i.e. they are performed even if `force_consistency_checks` is `false`.
+* Disable delete triggered compaction (NewCompactOnDeletionCollectorFactory) in universal compaction mode and num_levels = 1 in order to avoid a corruption bug.
+* `pin_l0_filter_and_index_blocks_in_cache` no longer applies to L0 files larger than `1.5 * write_buffer_size` to give more predictable memory usage. Such L0 files may exist due to intra-L0 compaction, external file ingestion, or user dynamically changing `write_buffer_size` (note, however, that files that are already pinned will continue being pinned, even after such a dynamic change).
+* In point-in-time wal recovery mode, fail database recovery in case of IOError while reading the WAL to avoid data loss.
+* A new method `Env::LowerThreadPoolCPUPriority(Priority, CpuPriority)` is added to `Env` to be able to lower to a specific priority such as `CpuPriority::kIdle`.
+
+### New Features
+* sst_dump to add a new --readahead_size argument. Users can specify read size when scanning the data. Sst_dump also tries to prefetch tail part of the SST files so usually some number of I/Os are saved there too.
+* Generate file checksum in SstFileWriter if Options.file_checksum_gen_factory is set. The checksum and checksum function name are stored in ExternalSstFileInfo after the sst file write is finished.
+* Add a value_size_soft_limit in read options which limits the cumulative value size of keys read in batches in MultiGet. Once the cumulative value size of found keys exceeds read_options.value_size_soft_limit, all the remaining keys are returned with status Abort without further finding their values. By default the value_size_soft_limit is std::numeric_limits<uint64_t>::max().
+* Enable SST file ingestion with file checksum information when calling IngestExternalFiles(const std::vector<IngestExternalFileArg>& args). Added files_checksums and files_checksum_func_names to IngestExternalFileArg such that user can ingest the sst files with their file checksum information. Added verify_file_checksum to IngestExternalFileOptions (default is True). To be backward compatible, if DB does not enable file checksum or user does not provide checksum information (vectors of files_checksums and files_checksum_func_names are both empty), verification of file checksum is always sucessful. If DB enables file checksum, DB will always generate the checksum for each ingested SST file during Prepare stage of ingestion and store the checksum in Manifest, unless verify_file_checksum is False and checksum information is provided by the application. In this case, we only verify the checksum function name and directly store the ingested checksum in Manifest. If verify_file_checksum is set to True, DB will verify the ingested checksum and function name with the genrated ones. Any mismatch will fail the ingestion. Note that, if IngestExternalFileOptions::write_global_seqno is True, the seqno will be changed in the ingested file. Therefore, the checksum of the file will be changed. In this case, a new checksum will be generated after the seqno is updated and be stored in the Manifest.
+
+### Performance Improvements
+* Eliminate redundant key comparisons during random access in block-based tables.

 ## 6.10 (5/2/2020)
 ### Bug Fixes
@@ -22,7 +123,6 @@
 * Add a ConfigOptions argument to the APIs dealing with converting options to and from strings and files.  The ConfigOptions is meant to replace some of the options (such as input_strings_escaped and ignore_unknown_options) and allow for more parameters to be passed in the future without changing the function signature.
 * Add NewFileChecksumGenCrc32cFactory to the file checksum public API, such that the builtin Crc32c based file checksum generator factory can be used by applications.
 * Add IsDirectory to Env and FS to indicate if a path is a directory.
-* DeleteRange now returns `Status::InvalidArgument` if the range's end key comes before its start key according to the user comparator. Previously the behavior was undefined.

 ### New Features
 * Added support for pipelined & parallel compression optimization for `BlockBasedTableBuilder`. This optimization makes block building, block compression and block appending a pipeline, and uses multiple threads to accelerate block compression. Users can set `CompressionOptions::parallel_threads` greater than 1 to enable compression parallelism. This feature is experimental for now.
@@ -10,7 +10,9 @@ This is the list of all known third-party language bindings for RocksDB. If some
 * Ruby - http://rubygems.org/gems/rocksdb-ruby
 * Haskell - https://hackage.haskell.org/package/rocksdb-haskell
 * PHP - https://github.com/Photonios/rocksdb-php
-* C# - https://github.com/warrenfalk/rocksdb-sharp
+* C#
+    * https://github.com/warrenfalk/rocksdb-sharp
+    * https://github.com/curiosity-ai/rocksdb-sharp
 * Rust
    * https://github.com/pingcap/rust-rocksdb (used in production fork of https://github.com/spacejam/rust-rocksdb)
    * https://github.com/spacejam/rust-rocksdb
@@ -1,7 +1,8 @@
 ## RocksDB: A Persistent Key-Value Store for Flash and RAM Storage

-[![Linux/Mac Build Status](https://travis-ci.org/facebook/rocksdb.svg?branch=master)](https://travis-ci.org/facebook/rocksdb)
-[![Windows Build status](https://ci.appveyor.com/api/projects/status/fbgfu0so3afcno78/branch/master?svg=true)](https://ci.appveyor.com/project/Facebook/rocksdb/branch/master)
+[![CircleCI Status](https://circleci.com/gh/facebook/rocksdb.svg?style=svg)](https://circleci.com/gh/facebook/rocksdb)
+[![TravisCI Status](https://travis-ci.org/facebook/rocksdb.svg?branch=master)](https://travis-ci.org/facebook/rocksdb)
+[![Appveyor Build status](https://ci.appveyor.com/api/projects/status/fbgfu0so3afcno78/branch/master?svg=true)](https://ci.appveyor.com/project/Facebook/rocksdb/branch/master)
 [![PPC64le Build Status](http://140.211.168.68:8080/buildStatus/icon?job=Rocksdb)](http://140.211.168.68:8080/job/Rocksdb)

 RocksDB is developed and maintained by Facebook Database Engineering Team.
@@ -1,4 +1,4 @@
-# This file @generated by `python buckifier/buckify_rocksdb.py`
+# This file @generated by `python3 buckifier/buckify_rocksdb.py`
 # --> DO NOT EDIT MANUALLY <--
 # This file is a Facebook-specific integration for buck builds, so can
 # only be validated by Facebook employees.
@@ -26,7 +26,6 @@ ROCKSDB_EXTERNAL_DEPS = [
    ("lz4", None, "lz4"),
    ("zstd", None),
    ("tbb", None),
-    ("googletest", None, "gtest"),
 ]

 ROCKSDB_OS_DEPS = [
@@ -109,6 +108,11 @@ ROCKSDB_OS_DEPS += ([(
    ["third-party//jemalloc:headers"],
 )] if sanitizer == "" else [])

+ROCKSDB_LIB_DEPS = [
+    ":rocksdb_lib",
+    ":rocksdb_test_lib",
+] if not is_opt_mode else [":rocksdb_lib"]
+
 cpp_library(
    name = "rocksdb_lib",
    srcs = [
@@ -120,6 +124,9 @@ cpp_library(
        "db/blob/blob_file_addition.cc",
        "db/blob/blob_file_garbage.cc",
        "db/blob/blob_file_meta.cc",
+        "db/blob/blob_log_format.cc",
+        "db/blob/blob_log_reader.cc",
+        "db/blob/blob_log_writer.cc",
        "db/builder.cc",
        "db/c.cc",
        "db/column_family.cc",
@@ -131,6 +138,7 @@ cpp_library(
        "db/compaction/compaction_picker_fifo.cc",
        "db/compaction/compaction_picker_level.cc",
        "db/compaction/compaction_picker_universal.cc",
+        "db/compaction/sst_partitioner.cc",
        "db/convenience.cc",
        "db/db_filesnapshot.cc",
        "db/db_impl/db_impl.cc",
@@ -186,6 +194,7 @@ cpp_library(
        "env/env_hdfs.cc",
        "env/env_posix.cc",
        "env/file_system.cc",
+        "env/file_system_tracer.cc",
        "env/fs_posix.cc",
        "env/io_posix.cc",
        "env/mock_env.cc",
@@ -275,6 +284,7 @@ cpp_library(
        "table/plain/plain_table_index.cc",
        "table/plain/plain_table_key_coding.cc",
        "table/plain/plain_table_reader.cc",
+        "table/sst_file_dumper.cc",
        "table/sst_file_reader.cc",
        "table/sst_file_writer.cc",
        "table/table_properties.cc",
@@ -287,6 +297,7 @@ cpp_library(
        "tools/ldb_tool.cc",
        "tools/sst_dump_tool.cc",
        "trace_replay/block_cache_tracer.cc",
+        "trace_replay/io_tracer.cc",
        "trace_replay/trace_replay.cc",
        "util/build_version.cc",
        "util/coding.cc",
@@ -314,9 +325,6 @@ cpp_library(
        "utilities/blob_db/blob_db_impl_filesnapshot.cc",
        "utilities/blob_db/blob_dump_tool.cc",
        "utilities/blob_db/blob_file.cc",
-        "utilities/blob_db/blob_log_format.cc",
-        "utilities/blob_db/blob_log_reader.cc",
-        "utilities/blob_db/blob_log_writer.cc",
        "utilities/cassandra/cassandra_compaction_filter.cc",
        "utilities/cassandra/format.cc",
        "utilities/cassandra/merge_operator.cc",
@@ -326,6 +334,8 @@ cpp_library(
        "utilities/debug.cc",
        "utilities/env_mirror.cc",
        "utilities/env_timed.cc",
+        "utilities/fault_injection_env.cc",
+        "utilities/fault_injection_fs.cc",
        "utilities/leveldb_options/leveldb_options.cc",
        "utilities/memory/memory_util.cc",
        "utilities/merge_operators/bytesxor.cc",
@@ -379,8 +389,6 @@ cpp_library(
    srcs = [
        "db/db_test_util.cc",
        "table/mock_table.cc",
-        "test_util/fault_injection_test_env.cc",
-        "test_util/fault_injection_test_fs.cc",
        "test_util/testharness.cc",
        "test_util/testutil.cc",
        "tools/block_cache_analyzer/block_cache_trace_analyzer.cc",
@@ -394,7 +402,9 @@ cpp_library(
    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
    deps = [":rocksdb_lib"],
-    external_deps = ROCKSDB_EXTERNAL_DEPS,
+    external_deps = ROCKSDB_EXTERNAL_DEPS + [
+        ("googletest", None, "gtest"),
+    ],
 )

 cpp_library(
@@ -437,13 +447,31 @@ cpp_library(
    os_deps = ROCKSDB_OS_DEPS,
    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-    deps = [
-        ":rocksdb_lib",
-        ":rocksdb_test_lib",
-    ],
+    deps = ROCKSDB_LIB_DEPS,
    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"],
+    )
+
+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",
+    )
+
 cpp_library(
    name = "env_basic_test_lib",
    srcs = ["env/env_basic_test.cc"],
@@ -515,6 +543,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "block_based_table_reader_test",
+        "table/block_based/block_based_table_reader_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "block_cache_trace_analyzer_test",
        "tools/block_cache_analyzer/block_cache_trace_analyzer_test.cc",
@@ -550,13 +585,6 @@ ROCKS_TESTS = [
        [],
        [],
    ],
-    [
-        "c_test",
-        "db/c_test.c",
-        "serial",
-        [],
-        [],
-    ],
    [
        "cache_simulator_test",
        "utilities/simulator_cache/cache_simulator_test.cc",
@@ -623,7 +651,7 @@ ROCKS_TESTS = [
    [
        "column_family_test",
        "db/column_family_test.cc",
-        "serial",
+        "parallel",
        [],
        [],
    ],
@@ -917,7 +945,7 @@ ROCKS_TESTS = [
    [
        "db_test2",
        "db/db_test2.cc",
-        "serial",
+        "parallel",
        [],
        [],
    ],
@@ -1145,6 +1173,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "io_tracer_test",
+        "trace_replay/io_tracer_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "iostats_context_test",
        "monitoring/iostats_context_test.cc",
@@ -1460,6 +1495,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "testutil_test",
+        "test_util/testutil_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "thread_list_test",
        "util/thread_list_test.cc",
@@ -1613,18 +1655,17 @@ ROCKS_TESTS = [
 # Do not build the tests in opt mode, since SyncPoint and other test code
 # will not be included.
 [
-    test_binary(
-        extra_compiler_flags = extra_compiler_flags,
-        extra_deps = extra_deps,
-        parallelism = parallelism,
-        rocksdb_arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
-        rocksdb_compiler_flags = ROCKSDB_COMPILER_FLAGS,
-        rocksdb_external_deps = ROCKSDB_EXTERNAL_DEPS,
-        rocksdb_os_deps = ROCKSDB_OS_DEPS,
-        rocksdb_os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
-        rocksdb_preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-        test_cc = test_cc,
-        test_name = test_name,
+    cpp_unittest(
+        name = test_name,
+        srcs = [test_cc],
+        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"] + extra_deps,
+        external_deps = ROCKSDB_EXTERNAL_DEPS + [
+            ("googletest", None, "gtest"),
+        ],
    )
    for test_name, test_cc, parallelism, extra_deps, extra_compiler_flags in ROCKS_TESTS
    if not is_opt_mode
@@ -24,15 +24,6 @@ environment:
    - APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2017
      CMAKE_GENERATOR: Visual Studio 15 Win64
      DEV_ENV: C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\Common7\IDE\devenv.com
-    - APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2019
-      CMAKE_GENERATOR: Visual Studio 16
-      CMAKE_PLATEFORM_NAME: x64
-      DEV_ENV: C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\Common7\IDE\devenv.com
-    - APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2019
-      CMAKE_GENERATOR: Visual Studio 16
-      CMAKE_PLATEFORM_NAME: x64
-      CMAKE_OPT: -DCMAKE_CXX_STANDARD=20
-      DEV_ENV: C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\Common7\IDE\devenv.com

 install:
  - md %THIRDPARTY_HOME%
@@ -68,7 +59,7 @@ before_build:
  - md %APPVEYOR_BUILD_FOLDER%\build
  - cd %APPVEYOR_BUILD_FOLDER%\build
  - if DEFINED CMAKE_PLATEFORM_NAME (set "PLATEFORM_OPT=-A %CMAKE_PLATEFORM_NAME%")
-  - cmake .. -G "%CMAKE_GENERATOR%" %PLATEFORM_OPT% %CMAKE_OPT% -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DLZ4=1 -DZSTD=1 -DXPRESS=1 -DJNI=1
+  - cmake .. -G "%CMAKE_GENERATOR%" %PLATEFORM_OPT% %CMAKE_OPT% -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DLZ4=1 -DZSTD=1 -DXPRESS=1 -DJNI=1 -DWITH_ALL_TESTS=0
  - cd ..

 build:
@@ -79,7 +70,7 @@ build:
 test:

 test_script:
-  - ps: build_tools\run_ci_db_test.ps1 -SuiteRun db_basic_test,db_with_timestamp_basic_test,db_test2,db_test,env_basic_test,env_test,db_merge_operand_test -Concurrency 8
+  - ps: build_tools\run_ci_db_test.ps1 -SuiteRun db_basic_test,env_basic_test -Concurrency 8

 on_failure:
  - cmd: 7z a build-failed.zip %APPVEYOR_BUILD_FOLDER%\build\ && appveyor PushArtifact build-failed.zip
@@ -20,10 +20,10 @@ from util import ColorString
 # User can pass extra dependencies as a JSON object via command line, and this
 # script can include these dependencies in the generate TARGETS file.
 # Usage:
-# $python buckifier/buckify_rocksdb.py
+# $python3 buckifier/buckify_rocksdb.py
 # (This generates a TARGET file without user-specified dependency for unit
 # tests.)
-# $python buckifier/buckify_rocksdb.py \
+# $python3 buckifier/buckify_rocksdb.py \
 #        '{"fake": { \
 #                      "extra_deps": [":test_dep", "//fakes/module:mock1"],  \
 #                      "extra_compiler_flags": ["-DROCKSDB_LITE", "-Os"], \
@@ -48,8 +48,8 @@ def parse_src_mk(repo_path):
        if '=' in line:
            current_src = line.split('=')[0].strip()
            src_files[current_src] = []
-        elif '.cc' in line:
-            src_path = line.split('.cc')[0].strip() + '.cc'
+        elif '.c' in line:
+            src_path = line.split('\\')[0].strip()
            src_files[current_src].append(src_path)
    return src_files

@@ -69,27 +69,11 @@ def get_cc_files(repo_path):
    return cc_files


-# Get tests from Makefile
-def get_tests(repo_path):
+# Get parallel tests from Makefile
+def get_parallel_tests(repo_path):
    Makefile = repo_path + "/Makefile"

-    # Dictionary TEST_NAME => IS_PARALLEL
-    tests = {}
-
-    found_tests = False
-    for line in open(Makefile):
-        line = line.strip()
-        if line.startswith("TESTS ="):
-            found_tests = True
-        elif found_tests:
-            if line.endswith("\\"):
-                # remove the trailing \
-                line = line[:-1]
-                line = line.strip()
-                tests[line] = False
-            else:
-                # we consumed all the tests
-                break
+    s = set({})

    found_parallel_tests = False
    for line in open(Makefile):
@@ -101,13 +85,12 @@ def get_tests(repo_path):
                # remove the trailing \
                line = line[:-1]
                line = line.strip()
-                tests[line] = True
+                s.add(line)
            else:
                # we consumed all the parallel tests
                break

-    return tests
-
+    return s

 # Parse extra dependencies passed by user from command line
 def get_dependencies():
@@ -140,13 +123,14 @@ def generate_targets(repo_path, deps_map):
    src_mk = parse_src_mk(repo_path)
    # get all .cc files
    cc_files = get_cc_files(repo_path)
-    # get tests from Makefile
-    tests = get_tests(repo_path)
+    # get parallel tests from Makefile
+    parallel_tests = get_parallel_tests(repo_path)

-    if src_mk is None or cc_files is None or tests is None:
+    if src_mk is None or cc_files is None or parallel_tests is None:
        return False

    TARGETS = TARGETSBuilder("%s/TARGETS" % repo_path)
+
    # rocksdb_lib
    TARGETS.add_library(
        "rocksdb_lib",
@@ -159,7 +143,10 @@ def generate_targets(repo_path, deps_map):
        src_mk.get("TEST_LIB_SOURCES", []) +
        src_mk.get("EXP_LIB_SOURCES", []) +
        src_mk.get("ANALYZER_LIB_SOURCES", []),
-        [":rocksdb_lib"])
+        [":rocksdb_lib"],
+        extra_external_deps=""" + [
+        ("googletest", None, "gtest"),
+    ]""")
    # rocksdb_tools_lib
    TARGETS.add_library(
        "rocksdb_tools_lib",
@@ -168,40 +155,50 @@ def generate_targets(repo_path, deps_map):
        ["test_util/testutil.cc"],
        [":rocksdb_lib"])
    # rocksdb_stress_lib
-    TARGETS.add_library(
+    TARGETS.add_rocksdb_library(
        "rocksdb_stress_lib",
        src_mk.get("ANALYZER_LIB_SOURCES", [])
        + src_mk.get('STRESS_LIB_SOURCES', [])
-        + ["test_util/testutil.cc"],
-        [":rocksdb_lib", ":rocksdb_test_lib"])
+        + ["test_util/testutil.cc"])
+
+    print("Extra dependencies:\n{0}".format(json.dumps(deps_map)))
+
+    # Dictionary test executable name -> relative source file path
+    test_source_map = {}
+    print(src_mk)
+
+    # c_test.c is added through TARGETS.add_c_test(). If there
+    # are more than one .c test file, we need to extend
+    # TARGETS.add_c_test() to include other C tests too.
+    for test_src in src_mk.get("TEST_MAIN_SOURCES_C", []):
+        if test_src != 'db/c_test.c':
+            print("Don't know how to deal with " + test_src)
+            return False
+    TARGETS.add_c_test()
+
+    for test_src in src_mk.get("TEST_MAIN_SOURCES", []):
+        test = test_src.split('.c')[0].strip().split('/')[-1].strip()
+        test_source_map[test] = test_src
+        print("" + test + " " + test_src)

-    print("Extra dependencies:\n{0}".format(str(deps_map)))
-    # test for every test we found in the Makefile
    for target_alias, deps in deps_map.items():
-        for test in sorted(tests):
-            match_src = [src for src in cc_files if ("/%s.c" % test) in src]
-            if len(match_src) == 0:
-                print(ColorString.warning("Cannot find .cc file for %s" % test))
-                continue
-            elif len(match_src) > 1:
-                print(ColorString.warning("Found more than one .cc for %s" % test))
-                print(match_src)
+        for test, test_src in sorted(test_source_map.items()):
+            if len(test) == 0:
+                print(ColorString.warning("Failed to get test name for %s" % test_src))
                continue

-            assert(len(match_src) == 1)
-            is_parallel = tests[test]
            test_target_name = \
                test if not target_alias else test + "_" + target_alias
            TARGETS.register_test(
                test_target_name,
-                match_src[0],
-                is_parallel,
-                deps['extra_deps'],
-                deps['extra_compiler_flags'])
+                test_src,
+                test in parallel_tests,
+                json.dumps(deps['extra_deps']),
+                json.dumps(deps['extra_compiler_flags']))

            if test in _EXPORTED_TEST_LIBS:
                test_library = "%s_lib" % test_target_name
-                TARGETS.add_library(test_library, match_src, [":rocksdb_test_lib"])
+                TARGETS.add_library(test_library, [test_src], [":rocksdb_test_lib"])
    TARGETS.flush_tests()

    print(ColorString.info("Generated TARGETS Summary:"))
@@ -220,6 +217,7 @@ def get_rocksdb_path():

    return rocksdb_path

+
 def exit_with_error(msg):
    print(ColorString.error(msg))
    sys.exit(1)
@@ -15,7 +15,7 @@ echo Backup original TARGETS file.

 cp TARGETS TARGETS.bkp

-python buckifier/buckify_rocksdb.py
+${PYTHON:-python3} buckifier/buckify_rocksdb.py

 TGT_DIFF=`git diff TARGETS | head -n 1`

@@ -24,8 +24,9 @@ then
  mv TARGETS.bkp TARGETS
  exit 0
 else
-  echo "Please run 'python buckifier/buckify_rocksdb.py' to update TARGETS file."
+  echo "Please run '${PYTHON:-python3} buckifier/buckify_rocksdb.py' to update TARGETS file."
  echo "Do not manually update TARGETS file."
+  ${PYTHON:-python3} --version
  mv TARGETS.bkp TARGETS
  exit 1
 fi
@@ -28,7 +28,8 @@ class TARGETSBuilder(object):
    def __init__(self, path):
        self.path = path
        self.targets_file = open(path, 'w')
-        self.targets_file.write(targets_cfg.rocksdb_target_header)
+        header = targets_cfg.rocksdb_target_header_template
+        self.targets_file.write(header)
        self.total_lib = 0
        self.total_bin = 0
        self.total_test = 0
@@ -37,17 +38,35 @@ class TARGETSBuilder(object):
    def __del__(self):
        self.targets_file.close()

-    def add_library(self, name, srcs, deps=None, headers=None):
+    def add_library(self, name, srcs, deps=None, headers=None,
+                    extra_external_deps=""):
        headers_attr_prefix = ""
        if headers is None:
            headers_attr_prefix = "auto_"
            headers = "AutoHeaders.RECURSIVE_GLOB"
+        else:
+            headers = "[" + pretty_list(headers) + "]"
        self.targets_file.write(targets_cfg.library_template.format(
            name=name,
            srcs=pretty_list(srcs),
            headers_attr_prefix=headers_attr_prefix,
            headers=headers,
-            deps=pretty_list(deps)))
+            deps=pretty_list(deps),
+            extra_external_deps=extra_external_deps))
+        self.total_lib = self.total_lib + 1
+
+    def add_rocksdb_library(self, name, srcs, headers=None):
+        headers_attr_prefix = ""
+        if headers is None:
+            headers_attr_prefix = "auto_"
+            headers = "AutoHeaders.RECURSIVE_GLOB"
+        else:
+            headers = "[" + pretty_list(headers) + "]"
+        self.targets_file.write(targets_cfg.rocksdb_library_template.format(
+            name=name,
+            srcs=pretty_list(srcs),
+            headers_attr_prefix=headers_attr_prefix,
+            headers=headers))
        self.total_lib = self.total_lib + 1

    def add_binary(self, name, srcs, deps=None):
@@ -57,6 +76,30 @@ class TARGETSBuilder(object):
            pretty_list(deps)))
        self.total_bin = self.total_bin + 1

+    def add_c_test(self):
+        self.targets_file.write("""
+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"],
+    )
+
+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",
+    )
+""")
+
    def register_test(self,
                      test_name,
                      src,
@@ -4,7 +4,8 @@ from __future__ import division
 from __future__ import print_function
 from __future__ import unicode_literals

-rocksdb_target_header = """# This file \100generated by `python buckifier/buckify_rocksdb.py`
+rocksdb_target_header_template = \
+    """# This file \100generated by `python3 buckifier/buckify_rocksdb.py`
 # --> DO NOT EDIT MANUALLY <--
 # This file is a Facebook-specific integration for buck builds, so can
 # only be validated by Facebook employees.
@@ -32,7 +33,6 @@ ROCKSDB_EXTERNAL_DEPS = [
    ("lz4", None, "lz4"),
    ("zstd", None),
    ("tbb", None),
-    ("googletest", None, "gtest"),
 ]

 ROCKSDB_OS_DEPS = [
@@ -114,6 +114,11 @@ ROCKSDB_OS_DEPS += ([(
    "linux",
    ["third-party//jemalloc:headers"],
 )] if sanitizer == "" else [])
+
+ROCKSDB_LIB_DEPS = [
+    ":rocksdb_lib",
+    ":rocksdb_test_lib",
+] if not is_opt_mode else [":rocksdb_lib"]
 """


@@ -128,6 +133,21 @@ cpp_library(
    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
    deps = [{deps}],
+    external_deps = ROCKSDB_EXTERNAL_DEPS{extra_external_deps},
+)
+"""
+
+rocksdb_library_template = """
+cpp_library(
+    name = "{name}",
+    srcs = [{srcs}],
+    {headers_attr_prefix}headers = {headers},
+    arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
+    compiler_flags = ROCKSDB_COMPILER_FLAGS,
+    os_deps = ROCKSDB_OS_DEPS,
+    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
+    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    deps = ROCKSDB_LIB_DEPS,
    external_deps = ROCKSDB_EXTERNAL_DEPS,
 )
 """
@@ -162,18 +182,17 @@ ROCKS_TESTS = [
 # Do not build the tests in opt mode, since SyncPoint and other test code
 # will not be included.
 [
-    test_binary(
-        extra_compiler_flags = extra_compiler_flags,
-        extra_deps = extra_deps,
-        parallelism = parallelism,
-        rocksdb_arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
-        rocksdb_compiler_flags = ROCKSDB_COMPILER_FLAGS,
-        rocksdb_external_deps = ROCKSDB_EXTERNAL_DEPS,
-        rocksdb_os_deps = ROCKSDB_OS_DEPS,
-        rocksdb_os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
-        rocksdb_preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-        test_cc = test_cc,
-        test_name = test_name,
+    cpp_unittest(
+        name = test_name,
+        srcs = [test_cc],
+        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"] + extra_deps,
+        external_deps = ROCKSDB_EXTERNAL_DEPS + [
+            ("googletest", None, "gtest"),
+        ],
    )
    for test_name, test_cc, parallelism, extra_deps, extra_compiler_flags in ROCKS_TESTS
    if not is_opt_mode
@@ -89,6 +89,16 @@ if test -z "$CXX"; then
    fi
 fi

+if test -z "$AR"; then
+    if [ -x "$(command -v gcc-ar)" ]; then
+        AR=gcc-ar
+    elif [ -x "$(command -v llvm-ar)" ]; then
+        AR=llvm-ar
+    else
+        AR=ar
+    fi
+fi
+
 # Detect OS
 if test -z "$TARGET_OS"; then
    TARGET_OS=`uname -s`
@@ -151,7 +161,7 @@ case "$TARGET_OS" in
        else
            PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -latomic"
        fi
-        PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lpthread -lrt"
+        PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lpthread -lrt -ldl"
        if test $ROCKSDB_USE_IO_URING; then
            # check for liburing
            $CXX $CFLAGS -x c++ - -luring -o /dev/null 2>/dev/null  <<EOF
@@ -192,6 +202,17 @@ EOF
        PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lpthread"
        # PORT_FILES=port/freebsd/freebsd_specific.cc
        ;;
+    GNU/kFreeBSD)
+        PLATFORM=OS_GNU_KFREEBSD
+        COMMON_FLAGS="$COMMON_FLAGS -DOS_GNU_KFREEBSD"
+        if [ -z "$USE_CLANG" ]; then
+            COMMON_FLAGS="$COMMON_FLAGS -fno-builtin-memcmp"
+        else
+            PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -latomic"
+        fi
+        PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lpthread -lrt"
+        # PORT_FILES=port/gnu_kfreebsd/gnu_kfreebsd_specific.cc
+        ;;
    NetBSD)
        PLATFORM=OS_NETBSD
        COMMON_FLAGS="$COMMON_FLAGS -fno-builtin-memcmp -D_REENTRANT -DOS_NETBSD"
@@ -285,24 +306,32 @@ EOF
        # Test whether gflags library is installed
        # http://gflags.github.io/gflags/
        # check if the namespace is gflags
-        $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null << EOF
+        if $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null << EOF
          #include <gflags/gflags.h>
+          using namespace GFLAGS_NAMESPACE;
          int main() {}
 EOF
-        if [ "$?" = 0 ]; then
-            COMMON_FLAGS="$COMMON_FLAGS -DGFLAGS=1"
-            PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lgflags"
-        else
-          # check if namespace is google
-          $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null << EOF
+        then
+          COMMON_FLAGS="$COMMON_FLAGS -DGFLAGS=1"
+          PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lgflags"
+        # check if namespace is gflags
+        elif $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null << EOF
+            #include <gflags/gflags.h>
+            using namespace gflags;
+            int main() {}
+EOF
+        then
+          COMMON_FLAGS="$COMMON_FLAGS -DGFLAGS=1 -DGFLAGS_NAMESPACE=gflags"
+          PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lgflags"
+        # check if namespace is google
+        elif $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null << EOF
            #include <gflags/gflags.h>
            using namespace google;
            int main() {}
 EOF
-          if [ "$?" = 0 ]; then
-              COMMON_FLAGS="$COMMON_FLAGS -DGFLAGS=google"
-              PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lgflags"
-          fi
+        then
+          COMMON_FLAGS="$COMMON_FLAGS -DGFLAGS=1 -DGFLAGS_NAMESPACE=google"
+          PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lgflags"
        fi
    fi

@@ -430,7 +459,7 @@ EOF
            COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_MALLOC_USABLE_SIZE"
        fi
    fi
-    
+
    if ! test $ROCKSDB_DISABLE_MEMKIND; then
        # Test whether memkind library is installed
        $CXX $CFLAGS $COMMON_FLAGS -lmemkind -x c++ - -o /dev/null 2>/dev/null  <<EOF
@@ -446,7 +475,7 @@ EOF
            JAVA_LDFLAGS="$JAVA_LDFLAGS -lmemkind"
        fi
    fi
-    
+
    if ! test $ROCKSDB_DISABLE_PTHREAD_MUTEX_ADAPTIVE_NP; then
        # Test whether PTHREAD_MUTEX_ADAPTIVE_NP mutex type is available
        $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null  <<EOF
@@ -612,6 +641,13 @@ else
  if test "$USE_SSE"; then
    TRY_SSE_ETC="1"
  fi
+
+  if [[ "${PLATFORM}" == "OS_MACOSX" ]]; then
+    # For portability compile for macOS 10.12 (2016) or newer
+    COMMON_FLAGS="$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"
+  fi
 fi

 if test "$TRY_SSE_ETC"; then
@@ -730,6 +766,7 @@ ROCKSDB_PATCH=`build_tools/version.sh patch`

 echo "CC=$CC" >> "$OUTPUT"
 echo "CXX=$CXX" >> "$OUTPUT"
+echo "AR=$AR" >> "$OUTPUT"
 echo "PLATFORM=$PLATFORM" >> "$OUTPUT"
 echo "PLATFORM_LDFLAGS=$PLATFORM_LDFLAGS" >> "$OUTPUT"
 echo "JAVA_LDFLAGS=$JAVA_LDFLAGS" >> "$OUTPUT"
@@ -109,6 +109,7 @@ if [ -z "$USE_CLANG" ]; then
  # gcc
  CC="$GCC_BASE/bin/gcc"
  CXX="$GCC_BASE/bin/g++"
+  AR="$GCC_BASE/bin/gcc-ar"

  CFLAGS+=" -B$BINUTILS/gold"
  CFLAGS+=" -isystem $GLIBC_INCLUDE"
@@ -119,6 +120,7 @@ else
  CLANG_INCLUDE="$CLANG_LIB/clang/stable/include"
  CC="$CLANG_BIN/clang"
  CXX="$CLANG_BIN/clang++"
+  AR="$CLANG_BIN/llvm-ar"

  KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include"

@@ -69,6 +69,7 @@ if [ -z "$USE_CLANG" ]; then
  # gcc
  CC="$GCC_BASE/bin/gcc"
  CXX="$GCC_BASE/bin/g++"
+  CXX="$GCC_BASE/bin/gcc-ar"

  CFLAGS="-B$BINUTILS/gold -m64 -mtune=generic"
  CFLAGS+=" -isystem $GLIBC_INCLUDE"
@@ -81,6 +82,7 @@ else
  CLANG_INCLUDE="$CLANG_LIB/clang/*/include"
  CC="$CLANG_BIN/clang"
  CXX="$CLANG_BIN/clang++"
+  AR="$CLANG_BIN/llvm-ar"

  KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include/"

@@ -118,6 +118,7 @@ if [ -z "$USE_CLANG" ]; then
  # gcc
  CC="$GCC_BASE/bin/gcc"
  CXX="$GCC_BASE/bin/g++"
+  AR="$GCC_BASE/bin/gcc-ar"

  CFLAGS+=" -B$BINUTILS/gold"
  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
@@ -128,6 +129,7 @@ else
  CLANG_INCLUDE="$CLANG_LIB/clang/stable/include"
  CC="$CLANG_BIN/clang"
  CXX="$CLANG_BIN/clang++"
+  AR="$CLANG_BIN/llvm-ar"

  KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include"

@@ -26,53 +26,77 @@ while getopts ':ch' OPTION; do
  esac
 done

-if [ -z $CLANG_FORMAT_DIFF ]
-then
-CLANG_FORMAT_DIFF="clang-format-diff.py"
-fi
+REPO_ROOT="$(git rev-parse --show-toplevel)"

-# Check clang-format-diff.py
-if ! which $CLANG_FORMAT_DIFF &> /dev/null
-then
-  if [ ! -f ./clang-format-diff.py ]
-  then
-    echo "You didn't have clang-format-diff.py and/or clang-format available in your computer!"
-    echo "You can download clang-format-diff.py by running: "
-    echo "    curl --location http://goo.gl/iUW1u2 -o ${CLANG_FORMAT_DIFF}"
-    echo "You can download clang-format by running:"
-    echo "    brew install clang-format"
-    echo "  Or"
-    echo "    apt install clang-format"
-    echo "  This might work too:"
-    echo "    yum install git-clang-format"
-    echo "Then, move both files (i.e. ${CLANG_FORMAT_DIFF} and clang-format) to some directory within PATH=${PATH}"
-    echo "and make sure ${CLANG_FORMAT_DIFF} is executable."
-    exit 128
+if [ "$CLANG_FORMAT_DIFF" ]; then
+  echo "Note: CLANG_FORMAT_DIFF='$CLANG_FORMAT_DIFF'"
+  # Dry run to confirm dependencies like argparse
+  if $CLANG_FORMAT_DIFF --help >/dev/null < /dev/null; then
+    true #Good
  else
-    if [ -x ./clang-format-diff.py ]
-    then
-      PATH=$PATH:.
+    exit 128
+  fi
+else
+  # First try directly executing the possibilities
+  if clang-format-diff.py --help &> /dev/null < /dev/null; then
+    CLANG_FORMAT_DIFF=clang-format-diff.py
+  elif $REPO_ROOT/clang-format-diff.py --help &> /dev/null < /dev/null; then
+    CLANG_FORMAT_DIFF=$REPO_ROOT/clang-format-diff.py
+  else
+    # This probably means we need to directly invoke the interpreter.
+    # But first find clang-format-diff.py
+    if [ -f "$REPO_ROOT/clang-format-diff.py" ]; then
+      CFD_PATH="$REPO_ROOT/clang-format-diff.py"
+    elif which clang-format-diff.py &> /dev/null; then
+      CFD_PATH="$(which clang-format-diff.py)"
    else
-      CLANG_FORMAT_DIFF="python ./clang-format-diff.py"
+      echo "You didn't have clang-format-diff.py and/or clang-format available in your computer!"
+      echo "You can download clang-format-diff.py by running: "
+      echo "    curl --location http://goo.gl/iUW1u2 -o ${CLANG_FORMAT_DIFF}"
+      echo "You can download clang-format by running:"
+      echo "    brew install clang-format"
+      echo "  Or"
+      echo "    apt install clang-format"
+      echo "  This might work too:"
+      echo "    yum install git-clang-format"
+      echo "Then, move both files (i.e. ${CLANG_FORMAT_DIFF} and clang-format) to some directory within PATH=${PATH}"
+      echo "and make sure ${CLANG_FORMAT_DIFF} is executable."
+      exit 128
+    fi
+    # Check argparse pre-req on interpreter, or it will fail
+    if echo import argparse | ${PYTHON:-python3}; then
+      true # Good
+    else
+      echo "To run clang-format-diff.py, we'll need the library "argparse" to be"
+      echo "installed. You can try either of the follow ways to install it:"
+      echo "  1. Manually download argparse: https://pypi.python.org/pypi/argparse"
+      echo "  2. easy_install argparse (if you have easy_install)"
+      echo "  3. pip install argparse (if you have pip)"
+      exit 129
+    fi
+    # Unfortunately, some machines have a Python2 clang-format-diff.py
+    # installed but only a Python3 interpreter installed. Rather than trying
+    # different Python versions that might be installed, we can try migrating
+    # the code to Python3 if it looks like Python2
+    if grep -q "print '" "$CFD_PATH" && \
+       ${PYTHON:-python3} --version | grep -q 'ython 3'; then
+      if [ ! -f "$REPO_ROOT/.py3/clang-format-diff.py" ]; then
+        echo "Migrating $CFD_PATH to Python3 in a hidden file"
+        mkdir -p "$REPO_ROOT/.py3"
+        ${PYTHON:-python3} -m lib2to3 -w -n -o "$REPO_ROOT/.py3" "$CFD_PATH" > /dev/null || exit 128
+      fi
+      CFD_PATH="$REPO_ROOT/.py3/clang-format-diff.py"
+    fi
+    CLANG_FORMAT_DIFF="${PYTHON:-python3} $CFD_PATH"
+    # This had better work after all those checks
+    if $CLANG_FORMAT_DIFF --help >/dev/null < /dev/null; then
+      true #Good
+    else
+      exit 128
    fi
  fi
 fi

-# Check argparse, a library that clang-format-diff.py requires.
-python 2>/dev/null << EOF
-import argparse
-EOF
-
-if [ "$?" != 0 ]
-then
-  echo "To run clang-format-diff.py, we'll need the library "argparse" to be"
-  echo "installed. You can try either of the follow ways to install it:"
-  echo "  1. Manually download argparse: https://pypi.python.org/pypi/argparse"
-  echo "  2. easy_install argparse (if you have easy_install)"
-  echo "  3. pip install argparse (if you have pip)"
-  exit 129
-fi
-
 # TODO(kailiu) following work is not complete since we still need to figure
 # out how to add the modified files done pre-commit hook to git's commit index.
 #
@@ -124,6 +148,10 @@ then
 elif [ $CHECK_ONLY ]
 then
  echo "Your change has unformatted code. Please run make format!"
+  if [ $VERBOSE_CHECK ]; then
+    clang-format --version
+    echo "$diffs"
+  fi
  exit 1
 fi

@@ -421,6 +421,66 @@ STRESS_CRASH_TEST_COMMANDS="[
    }
 ]"

+#
+# RocksDB blackbox stress/crash test
+#
+BLACKBOX_STRESS_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb Blackbox Stress and Crash Test',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Build and run RocksDB debug stress tests',
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            {
+                'name':'Build and run RocksDB debug blackbox crash tests',
+                'timeout': 86400,
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 blackbox_crash_test || $CONTRUN_NAME=blackbox_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
+#
+# RocksDB whitebox stress/crash test
+#
+WHITEBOX_STRESS_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb Whitebox Stress and Crash Test',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Build and run RocksDB debug stress tests',
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            {
+                'name':'Build and run RocksDB debug whitebox crash tests',
+                'timeout': 86400,
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 whitebox_crash_test || $CONTRUN_NAME=whitebox_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
 #
 # RocksDB stress/crash test with atomic flush
 #
@@ -549,6 +609,54 @@ ASAN_CRASH_TEST_COMMANDS="[
    }
 ]"

+#
+# RocksDB blackbox crash testing under address sanitizer
+#
+ASAN_BLACKBOX_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb blackbox crash test under ASAN',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Build and run RocksDB debug blackbox asan_crash_test',
+                'timeout': 86400,
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 blackbox_asan_crash_test || $CONTRUN_NAME=blackbox_asan_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
+#
+# RocksDB whitebox crash testing under address sanitizer
+#
+ASAN_WHITEBOX_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb whitebox crash test under ASAN',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Build and run RocksDB debug whitebox asan_crash_test',
+                'timeout': 86400,
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 whitebox_asan_crash_test || $CONTRUN_NAME=whitebox_asan_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
 #
 # RocksDB crash testing with atomic flush under address sanitizer
 #
@@ -619,7 +727,7 @@ UBSAN_TEST_COMMANDS="[
 ]"

 #
-# RocksDB crash testing under udnefined behavior sanitizer
+# RocksDB crash testing under undefined behavior sanitizer
 #
 UBSAN_CRASH_TEST_COMMANDS="[
    {
@@ -642,6 +750,54 @@ UBSAN_CRASH_TEST_COMMANDS="[
    }
 ]"

+#
+# RocksDB crash testing under undefined behavior sanitizer
+#
+UBSAN_BLACKBOX_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb blackbox crash test under UBSAN',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Build and run RocksDB debug blackbox ubsan_crash_test',
+                'timeout': 86400,
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 blackbox_ubsan_crash_test || $CONTRUN_NAME=blackbox_ubsan_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
+#
+# RocksDB crash testing under undefined behavior sanitizer
+#
+UBSAN_WHITEBOX_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb whitebox crash test under UBSAN',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Build and run RocksDB debug whitebox ubsan_crash_test',
+                'timeout': 86400,
+                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 whitebox_ubsan_crash_test || $CONTRUN_NAME=whitebox_ubsan_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
 #
 # RocksDB crash testing with atomic flush under undefined behavior sanitizer
 #
@@ -760,6 +916,54 @@ TSAN_CRASH_TEST_COMMANDS="[
    }
 ]"

+#
+# RocksDB blackbox crash test under TSAN
+#
+TSAN_BLACKBOX_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb Blackbox Crash Test under TSAN',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Compile and run',
+                'timeout': 86400,
+                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 blackbox_crash_test || $CONTRUN_NAME=tsan_blackbox_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
+#
+# RocksDB whitebox crash test under TSAN
+#
+TSAN_WHITEBOX_CRASH_TEST_COMMANDS="[
+    {
+        'name':'Rocksdb Whitebox Crash Test under TSAN',
+        'oncall':'$ONCALL',
+        'executeLocal': 'true',
+        'timeout': 86400,
+        'steps': [
+            $CLEANUP_ENV,
+            {
+                'name':'Compile and run',
+                'timeout': 86400,
+                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 whitebox_crash_test || $CONTRUN_NAME=tsan_whitebox_crash_test $TASK_CREATION_TOOL',
+                'user':'root',
+                $PARSER
+            },
+            $UPLOAD_DB_DIR,
+        ],
+        $REPORT
+    }
+]"
+
 #
 # RocksDB crash test with atomic flush under TSAN
 #
@@ -994,6 +1198,12 @@ case $1 in
  stress_crash)
    echo $STRESS_CRASH_TEST_COMMANDS
    ;;
+  blackbox_stress_crash)
+    echo $BLACKBOX_STRESS_CRASH_TEST_COMMANDS
+    ;;
+  whitebox_stress_crash)
+    echo $WHITEBOX_STRESS_CRASH_TEST_COMMANDS
+    ;;
  stress_crash_with_atomic_flush)
    echo $STRESS_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS
    ;;
@@ -1009,6 +1219,12 @@ case $1 in
  asan_crash)
    echo $ASAN_CRASH_TEST_COMMANDS
    ;;
+  blackbox_asan_crash)
+    echo $ASAN_BLACKBOX_CRASH_TEST_COMMANDS
+    ;;
+  whitebox_asan_crash)
+    echo $ASAN_WHITEBOX_CRASH_TEST_COMMANDS
+    ;;
  asan_crash_with_atomic_flush)
    echo $ASAN_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS
    ;;
@@ -1021,6 +1237,12 @@ case $1 in
  ubsan_crash)
    echo $UBSAN_CRASH_TEST_COMMANDS
    ;;
+  blackbox_ubsan_crash)
+    echo $UBSAN_BLACKBOX_CRASH_TEST_COMMANDS
+    ;;
+  whitebox_ubsan_crash)
+    echo $UBSAN_WHITEBOX_CRASH_TEST_COMMANDS
+    ;;
  ubsan_crash_with_atomic_flush)
    echo $UBSAN_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS
    ;;
@@ -1036,6 +1258,12 @@ case $1 in
  tsan_crash)
    echo $TSAN_CRASH_TEST_COMMANDS
    ;;
+  blackbox_tsan_crash)
+    echo $TSAN_BLACKBOX_CRASH_TEST_COMMANDS
+    ;;
+  whitebox_tsan_crash)
+    echo $TSAN_WHITEBOX_CRASH_TEST_COMMANDS
+    ;;
  tsan_crash_with_atomic_flush)
    echo $TSAN_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS
    ;;
@@ -14,7 +14,30 @@
 #include "util/string_util.h"

 namespace ROCKSDB_NAMESPACE {
-Status Cache::CreateFromString(const ConfigOptions& /*opts*/,
+#ifndef ROCKSDB_LITE
+static std::unordered_map<std::string, OptionTypeInfo>
+    lru_cache_options_type_info = {
+        {"capacity",
+         {offsetof(struct LRUCacheOptions, capacity), OptionType::kSizeT,
+          OptionVerificationType::kNormal, OptionTypeFlags::kMutable,
+          offsetof(struct LRUCacheOptions, capacity)}},
+        {"num_shard_bits",
+         {offsetof(struct LRUCacheOptions, num_shard_bits), OptionType::kInt,
+          OptionVerificationType::kNormal, OptionTypeFlags::kMutable,
+          offsetof(struct LRUCacheOptions, num_shard_bits)}},
+        {"strict_capacity_limit",
+         {offsetof(struct LRUCacheOptions, strict_capacity_limit),
+          OptionType::kBoolean, OptionVerificationType::kNormal,
+          OptionTypeFlags::kMutable,
+          offsetof(struct LRUCacheOptions, strict_capacity_limit)}},
+        {"high_pri_pool_ratio",
+         {offsetof(struct LRUCacheOptions, high_pri_pool_ratio),
+          OptionType::kDouble, OptionVerificationType::kNormal,
+          OptionTypeFlags::kMutable,
+          offsetof(struct LRUCacheOptions, high_pri_pool_ratio)}}};
+#endif  // ROCKSDB_LITE
+
+Status Cache::CreateFromString(const ConfigOptions& config_options,
                               const std::string& value,
                               std::shared_ptr<Cache>* result) {
  Status status;
@@ -24,12 +47,14 @@ Status Cache::CreateFromString(const ConfigOptions& /*opts*/,
  } else {
 #ifndef ROCKSDB_LITE
    LRUCacheOptions cache_opts;
-    if (!ParseOptionHelper(reinterpret_cast<char*>(&cache_opts),
-                           OptionType::kLRUCacheOptions, value)) {
-      status = Status::InvalidArgument("Invalid cache options");
+    status = OptionTypeInfo::ParseStruct(
+        config_options, "", &lru_cache_options_type_info, "", value,
+        reinterpret_cast<char*>(&cache_opts));
+    if (status.ok()) {
+      cache = NewLRUCache(cache_opts);
    }
-    cache = NewLRUCache(cache_opts);
 #else
+    (void)config_options;
    status = Status::NotSupported("Cannot load cache in LITE mode ", value);
 #endif  //! ROCKSDB_LITE
  }
@@ -182,7 +182,7 @@ struct CacheHandle {
  void (*deleter)(const Slice&, void* value);

  // Flags and counters associated with the cache handle:
-  //   lowest bit: n-cache bit
+  //   lowest bit: in-cache bit
  //   second lowest bit: usage bit
  //   the rest bits: reference count
  // The handle is unused when flags equals to 0. The thread decreases the count
@@ -1,3 +1,54 @@
@PACKAGE_INIT@
+
+list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_LIST_DIR}/modules")
+
+include(CMakeFindDependencyMacro)
+
+set(GFLAGS_USE_TARGET_NAMESPACE @GFLAGS_USE_TARGET_NAMESPACE@)
+
+if(@WITH_JEMALLOC@)
+  find_dependency(JeMalloc)
+endif()
+
+if(@WITH_GFLAGS@)
+  find_dependency(gflags CONFIG)
+  if(NOT gflags_FOUND)
+    find_dependency(gflags)
+  endif()
+endif()
+
+if(@WITH_SNAPPY@)
+  find_dependency(Snappy CONFIG)
+  if(NOT Snappy_FOUND)
+    find_dependency(Snappy)
+  endif()
+endif()
+
+if(@WITH_ZLIB@)
+  find_dependency(ZLIB)
+endif()
+
+if(@WITH_BZ2@)
+  find_dependency(BZip2)
+endif()
+
+if(@WITH_LZ4@)
+  find_dependency(lz4)
+endif()
+
+if(@WITH_ZSTD@)
+  find_dependency(zstd)
+endif()
+
+if(@WITH_NUMA@)
+  find_dependency(NUMA)
+endif()
+
+if(@WITH_TBB@)
+  find_dependency(TBB)
+endif()
+
+find_dependency(Threads)
+
 include("${CMAKE_CURRENT_LIST_DIR}/RocksDBTargets.cmake")
 check_required_components(RocksDB)
@@ -0,0 +1,29 @@
+# - Find Snappy
+# Find the snappy compression library and includes
+#
+# Snappy_INCLUDE_DIRS - where to find snappy.h, etc.
+# Snappy_LIBRARIES - List of libraries when using snappy.
+# Snappy_FOUND - True if snappy found.
+
+find_path(Snappy_INCLUDE_DIRS
+  NAMES snappy.h
+  HINTS ${snappy_ROOT_DIR}/include)
+
+find_library(Snappy_LIBRARIES
+  NAMES snappy
+  HINTS ${snappy_ROOT_DIR}/lib)
+
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(Snappy DEFAULT_MSG Snappy_LIBRARIES Snappy_INCLUDE_DIRS)
+
+mark_as_advanced(
+  Snappy_LIBRARIES
+  Snappy_INCLUDE_DIRS)
+
+if(Snappy_FOUND AND NOT (TARGET Snappy::snappy))
+  add_library (Snappy::snappy UNKNOWN IMPORTED)
+  set_target_properties(Snappy::snappy
+    PROPERTIES
+      IMPORTED_LOCATION ${Snappy_LIBRARIES}
+      INTERFACE_INCLUDE_DIRECTORIES ${Snappy_INCLUDE_DIRS})
+endif()
@@ -1,8 +1,8 @@
 # - Find gflags library
 # Find the gflags includes and library
 #
-# gflags_INCLUDE_DIR - where to find gflags.h.
-# gflags_LIBRARIES - List of libraries when using gflags.
+# GFLAGS_INCLUDE_DIR - where to find gflags.h.
+# GFLAGS_LIBRARIES - List of libraries when using gflags.
 # gflags_FOUND - True if gflags found.

 find_path(GFLAGS_INCLUDE_DIR
@@ -1,29 +0,0 @@
-# - Find Snappy
-# Find the snappy compression library and includes
-#
-# snappy_INCLUDE_DIRS - where to find snappy.h, etc.
-# snappy_LIBRARIES - List of libraries when using snappy.
-# snappy_FOUND - True if snappy found.
-
-find_path(snappy_INCLUDE_DIRS
-  NAMES snappy.h
-  HINTS ${snappy_ROOT_DIR}/include)
-
-find_library(snappy_LIBRARIES
-  NAMES snappy
-  HINTS ${snappy_ROOT_DIR}/lib)
-
-include(FindPackageHandleStandardArgs)
-find_package_handle_standard_args(snappy DEFAULT_MSG snappy_LIBRARIES snappy_INCLUDE_DIRS)
-
-mark_as_advanced(
-  snappy_LIBRARIES
-  snappy_INCLUDE_DIRS)
-
-if(snappy_FOUND AND NOT (TARGET snappy::snappy))
-  add_library (snappy::snappy UNKNOWN IMPORTED)
-  set_target_properties(snappy::snappy
-    PROPERTIES
-      IMPORTED_LOCATION ${snappy_LIBRARIES}
-      INTERFACE_INCLUDE_DIRECTORIES ${snappy_INCLUDE_DIRS})
-endif()
@@ -12,7 +12,7 @@ fi
 ROOT=".."
 # Fetch right version of gcov
 if [ -d /mnt/gvfs/third-party -a -z "$CXX" ]; then
-  source $ROOT/build_tools/fbcode_config.sh
+  source $ROOT/build_tools/fbcode_config_platform007.sh
  GCOV=$GCC_BASE/bin/gcov
 else
  GCOV=$(which gcov)
@@ -56,8 +56,9 @@ Status ArenaWrappedDBIter::Refresh() {
  // TODO(yiwu): For last_seq_same_as_publish_seq_==false, this is not the
  // correct behavior. Will be corrected automatically when we take a snapshot
  // here for the case of WritePreparedTxnDB.
-  SequenceNumber latest_seq = db_impl_->GetLatestSequenceNumber();
  uint64_t cur_sv_number = cfd_->GetSuperVersionNumber();
+  TEST_SYNC_POINT("ArenaWrappedDBIter::Refresh:1");
+  TEST_SYNC_POINT("ArenaWrappedDBIter::Refresh:2");
  if (sv_number_ != cur_sv_number) {
    Env* env = db_iter_->env();
    db_iter_->~DBIter();
@@ -65,6 +66,7 @@ Status ArenaWrappedDBIter::Refresh() {
    new (&arena_) Arena();

    SuperVersion* sv = cfd_->GetReferencedSuperVersion(db_impl_);
+    SequenceNumber latest_seq = db_impl_->GetLatestSequenceNumber();
    if (read_callback_) {
      read_callback_->Refresh(latest_seq);
    }
@@ -78,7 +80,7 @@ Status ArenaWrappedDBIter::Refresh() {
        latest_seq, /* allow_unprepared_value */ true);
    SetIterUnderDBIter(internal_iter);
  } else {
-    db_iter_->set_sequence(latest_seq);
+    db_iter_->set_sequence(db_impl_->GetLatestSequenceNumber());
    db_iter_->set_valid(false);
  }
  return Status::OK();
@@ -38,8 +38,15 @@ std::string BlobFileMetaData::DebugString() const {
 std::ostream& operator<<(std::ostream& os, const BlobFileMetaData& meta) {
  const auto& shared_meta = meta.GetSharedMeta();
  assert(shared_meta);
+  os << (*shared_meta);

-  os << (*shared_meta) << " garbage_blob_count: " << meta.GetGarbageBlobCount()
+  os << " linked_ssts: {";
+  for (uint64_t file_number : meta.GetLinkedSsts()) {
+    os << ' ' << file_number;
+  }
+  os << " }";
+
+  os << " garbage_blob_count: " << meta.GetGarbageBlobCount()
     << " garbage_blob_bytes: " << meta.GetGarbageBlobBytes();

  return os;
@@ -5,12 +5,13 @@

 #pragma once

-#include "rocksdb/rocksdb_namespace.h"
-
 #include <cassert>
 #include <iosfwd>
 #include <memory>
 #include <string>
+#include <unordered_set>
+
+#include "rocksdb/rocksdb_namespace.h"

 namespace ROCKSDB_NAMESPACE {

@@ -90,11 +91,15 @@ std::ostream& operator<<(std::ostream& os,

 class BlobFileMetaData {
 public:
+  using LinkedSsts = std::unordered_set<uint64_t>;
+
  static std::shared_ptr<BlobFileMetaData> Create(
      std::shared_ptr<SharedBlobFileMetaData> shared_meta,
-      uint64_t garbage_blob_count, uint64_t garbage_blob_bytes) {
-    return std::shared_ptr<BlobFileMetaData>(new BlobFileMetaData(
-        std::move(shared_meta), garbage_blob_count, garbage_blob_bytes));
+      LinkedSsts linked_ssts, uint64_t garbage_blob_count,
+      uint64_t garbage_blob_bytes) {
+    return std::shared_ptr<BlobFileMetaData>(
+        new BlobFileMetaData(std::move(shared_meta), std::move(linked_ssts),
+                             garbage_blob_count, garbage_blob_bytes));
  }

  BlobFileMetaData(const BlobFileMetaData&) = delete;
@@ -128,6 +133,8 @@ class BlobFileMetaData {
    return shared_meta_->GetChecksumValue();
  }

+  const LinkedSsts& GetLinkedSsts() const { return linked_ssts_; }
+
  uint64_t GetGarbageBlobCount() const { return garbage_blob_count_; }
  uint64_t GetGarbageBlobBytes() const { return garbage_blob_bytes_; }

@@ -135,8 +142,10 @@ class BlobFileMetaData {

 private:
  BlobFileMetaData(std::shared_ptr<SharedBlobFileMetaData> shared_meta,
-                   uint64_t garbage_blob_count, uint64_t garbage_blob_bytes)
+                   LinkedSsts linked_ssts, uint64_t garbage_blob_count,
+                   uint64_t garbage_blob_bytes)
      : shared_meta_(std::move(shared_meta)),
+        linked_ssts_(std::move(linked_ssts)),
        garbage_blob_count_(garbage_blob_count),
        garbage_blob_bytes_(garbage_blob_bytes) {
    assert(shared_meta_);
@@ -145,6 +154,7 @@ class BlobFileMetaData {
  }

  std::shared_ptr<SharedBlobFileMetaData> shared_meta_;
+  LinkedSsts linked_ssts_;
  uint64_t garbage_blob_count_;
  uint64_t garbage_blob_bytes_;
 };
@@ -5,13 +5,12 @@
 //
 #ifndef ROCKSDB_LITE

-#include "utilities/blob_db/blob_log_format.h"
+#include "db/blob/blob_log_format.h"

 #include "util/coding.h"
 #include "util/crc32c.h"

 namespace ROCKSDB_NAMESPACE {
-namespace blob_db {

 void BlobLogHeader::EncodeTo(std::string* dst) {
  assert(dst != nullptr);
@@ -144,6 +143,5 @@ Status BlobLogRecord::CheckBlobCRC() const {
  return Status::OK();
 }

-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -9,7 +9,6 @@

 #ifndef ROCKSDB_LITE

-#include <limits>
 #include <memory>
 #include <utility>

@@ -19,11 +18,9 @@
 #include "rocksdb/types.h"

 namespace ROCKSDB_NAMESPACE {
-namespace blob_db {

 constexpr uint32_t kMagicNumber = 2395959;  // 0x00248f37
 constexpr uint32_t kVersion1 = 1;
-constexpr uint64_t kNoExpiration = std::numeric_limits<uint64_t>::max();

 using ExpirationRange = std::pair<uint64_t, uint64_t>;

@@ -128,6 +125,5 @@ struct BlobLogRecord {
  Status CheckBlobCRC() const;
 };

-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -5,7 +5,7 @@
 //
 #ifndef ROCKSDB_LITE

-#include "utilities/blob_db/blob_log_reader.h"
+#include "db/blob/blob_log_reader.h"

 #include <algorithm>

@@ -14,17 +14,17 @@
 #include "util/stop_watch.h"

 namespace ROCKSDB_NAMESPACE {
-namespace blob_db {

-Reader::Reader(std::unique_ptr<RandomAccessFileReader>&& file_reader, Env* env,
-               Statistics* statistics)
+BlobLogReader::BlobLogReader(
+    std::unique_ptr<RandomAccessFileReader>&& file_reader, Env* env,
+    Statistics* statistics)
    : file_(std::move(file_reader)),
      env_(env),
      statistics_(statistics),
      buffer_(),
      next_byte_(0) {}

-Status Reader::ReadSlice(uint64_t size, Slice* slice, char* buf) {
+Status BlobLogReader::ReadSlice(uint64_t size, Slice* slice, char* buf) {
  StopWatch read_sw(env_, statistics_, BLOB_DB_BLOB_FILE_READ_MICROS);
  Status s = file_->Read(IOOptions(), next_byte_, static_cast<size_t>(size),
                         slice, buf, nullptr);
@@ -39,7 +39,7 @@ Status Reader::ReadSlice(uint64_t size, Slice* slice, char* buf) {
  return s;
 }

-Status Reader::ReadHeader(BlobLogHeader* header) {
+Status BlobLogReader::ReadHeader(BlobLogHeader* header) {
  assert(file_.get() != nullptr);
  assert(next_byte_ == 0);
  Status s = ReadSlice(BlobLogHeader::kSize, &buffer_, header_buf_);
@@ -54,8 +54,8 @@ Status Reader::ReadHeader(BlobLogHeader* header) {
  return header->DecodeFrom(buffer_);
 }

-Status Reader::ReadRecord(BlobLogRecord* record, ReadLevel level,
-                          uint64_t* blob_offset) {
+Status BlobLogReader::ReadRecord(BlobLogRecord* record, ReadLevel level,
+                                 uint64_t* blob_offset) {
  Status s = ReadSlice(BlobLogRecord::kHeaderSize, &buffer_, header_buf_);
  if (!s.ok()) {
    return s;
@@ -101,6 +101,5 @@ Status Reader::ReadRecord(BlobLogRecord* record, ReadLevel level,
  return s;
 }

-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -10,27 +10,26 @@
 #include <memory>
 #include <string>

+#include "db/blob/blob_log_format.h"
 #include "file/random_access_file_reader.h"
 #include "rocksdb/env.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/statistics.h"
 #include "rocksdb/status.h"
-#include "utilities/blob_db/blob_log_format.h"

 namespace ROCKSDB_NAMESPACE {

 class SequentialFileReader;
 class Logger;

-namespace blob_db {
-
 /**
- * Reader is a general purpose log stream reader implementation. The actual job
- * of reading from the device is implemented by the SequentialFile interface.
+ * BlobLogReader is a general purpose log stream reader implementation. The
+ * actual job of reading from the device is implemented by the SequentialFile
+ * interface.
 *
 * Please see Writer for details on the file and record layout.
 */
-class Reader {
+class BlobLogReader {
 public:
  enum ReadLevel {
    kReadHeader,
@@ -39,14 +38,14 @@ class Reader {
  };

  // Create a reader that will return log records from "*file".
-  // "*file" must remain live while this Reader is in use.
-  Reader(std::unique_ptr<RandomAccessFileReader>&& file_reader, Env* env,
-         Statistics* statistics);
+  // "*file" must remain live while this BlobLogReader is in use.
+  BlobLogReader(std::unique_ptr<RandomAccessFileReader>&& file_reader, Env* env,
+                Statistics* statistics);
  // No copying allowed
-  Reader(const Reader&) = delete;
-  Reader& operator=(const Reader&) = delete;
+  BlobLogReader(const BlobLogReader&) = delete;
+  BlobLogReader& operator=(const BlobLogReader&) = delete;

-  ~Reader() = default;
+  ~BlobLogReader() = default;

  Status ReadHeader(BlobLogHeader* header);

@@ -77,6 +76,5 @@ class Reader {
  uint64_t next_byte_;
 };

-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -4,42 +4,42 @@
 //  (found in the LICENSE.Apache file in the root directory).
 #ifndef ROCKSDB_LITE

-#include "utilities/blob_db/blob_log_writer.h"
+#include "db/blob/blob_log_writer.h"

 #include <cstdint>
 #include <string>

+#include "db/blob/blob_log_format.h"
 #include "file/writable_file_writer.h"
 #include "monitoring/statistics.h"
 #include "rocksdb/env.h"
+#include "test_util/sync_point.h"
 #include "util/coding.h"
 #include "util/stop_watch.h"
-#include "utilities/blob_db/blob_log_format.h"

 namespace ROCKSDB_NAMESPACE {
-namespace blob_db {

-Writer::Writer(std::unique_ptr<WritableFileWriter>&& dest, Env* env,
-               Statistics* statistics, uint64_t log_number, uint64_t bpsync,
-               bool use_fs, uint64_t boffset)
+BlobLogWriter::BlobLogWriter(std::unique_ptr<WritableFileWriter>&& dest,
+                             Env* env, Statistics* statistics,
+                             uint64_t log_number, bool use_fs, uint64_t boffset)
    : dest_(std::move(dest)),
      env_(env),
      statistics_(statistics),
      log_number_(log_number),
      block_offset_(boffset),
-      bytes_per_sync_(bpsync),
-      next_sync_offset_(0),
      use_fsync_(use_fs),
      last_elem_type_(kEtNone) {}

-Status Writer::Sync() {
+Status BlobLogWriter::Sync() {
+  TEST_SYNC_POINT("BlobLogWriter::Sync");
+
  StopWatch sync_sw(env_, statistics_, BLOB_DB_BLOB_FILE_SYNC_MICROS);
  Status s = dest_->Sync(use_fsync_);
  RecordTick(statistics_, BLOB_DB_BLOB_FILE_SYNCED);
  return s;
 }

-Status Writer::WriteHeader(BlobLogHeader& header) {
+Status BlobLogWriter::WriteHeader(BlobLogHeader& header) {
  assert(block_offset_ == 0);
  assert(last_elem_type_ == kEtNone);
  std::string str;
@@ -56,7 +56,7 @@ Status Writer::WriteHeader(BlobLogHeader& header) {
  return s;
 }

-Status Writer::AppendFooter(BlobLogFooter& footer) {
+Status BlobLogWriter::AppendFooter(BlobLogFooter& footer) {
  assert(block_offset_ != 0);
  assert(last_elem_type_ == kEtFileHdr || last_elem_type_ == kEtRecord);

@@ -66,7 +66,10 @@ Status Writer::AppendFooter(BlobLogFooter& footer) {
  Status s = dest_->Append(Slice(str));
  if (s.ok()) {
    block_offset_ += str.size();
-    s = dest_->Close();
+    s = Sync();
+    if (s.ok()) {
+      s = dest_->Close();
+    }
    dest_.reset();
  }

@@ -76,9 +79,9 @@ Status Writer::AppendFooter(BlobLogFooter& footer) {
  return s;
 }

-Status Writer::AddRecord(const Slice& key, const Slice& val,
-                         uint64_t expiration, uint64_t* key_offset,
-                         uint64_t* blob_offset) {
+Status BlobLogWriter::AddRecord(const Slice& key, const Slice& val,
+                                uint64_t expiration, uint64_t* key_offset,
+                                uint64_t* blob_offset) {
  assert(block_offset_ != 0);
  assert(last_elem_type_ == kEtFileHdr || last_elem_type_ == kEtRecord);

@@ -89,8 +92,8 @@ Status Writer::AddRecord(const Slice& key, const Slice& val,
  return s;
 }

-Status Writer::AddRecord(const Slice& key, const Slice& val,
-                         uint64_t* key_offset, uint64_t* blob_offset) {
+Status BlobLogWriter::AddRecord(const Slice& key, const Slice& val,
+                                uint64_t* key_offset, uint64_t* blob_offset) {
  assert(block_offset_ != 0);
  assert(last_elem_type_ == kEtFileHdr || last_elem_type_ == kEtRecord);

@@ -101,8 +104,8 @@ Status Writer::AddRecord(const Slice& key, const Slice& val,
  return s;
 }

-void Writer::ConstructBlobHeader(std::string* buf, const Slice& key,
-                                 const Slice& val, uint64_t expiration) {
+void BlobLogWriter::ConstructBlobHeader(std::string* buf, const Slice& key,
+                                        const Slice& val, uint64_t expiration) {
  BlobLogRecord record;
  record.key = key;
  record.value = val;
@@ -110,9 +113,10 @@ void Writer::ConstructBlobHeader(std::string* buf, const Slice& key,
  record.EncodeHeaderTo(buf);
 }

-Status Writer::EmitPhysicalRecord(const std::string& headerbuf,
-                                  const Slice& key, const Slice& val,
-                                  uint64_t* key_offset, uint64_t* blob_offset) {
+Status BlobLogWriter::EmitPhysicalRecord(const std::string& headerbuf,
+                                         const Slice& key, const Slice& val,
+                                         uint64_t* key_offset,
+                                         uint64_t* blob_offset) {
  StopWatch write_sw(env_, statistics_, BLOB_DB_BLOB_FILE_WRITE_MICROS);
  Status s = dest_->Append(Slice(headerbuf));
  if (s.ok()) {
@@ -134,6 +138,5 @@ Status Writer::EmitPhysicalRecord(const std::string& headerbuf,
  return s;
 }

-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -10,40 +10,38 @@
 #include <memory>
 #include <string>

+#include "db/blob/blob_log_format.h"
 #include "rocksdb/env.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/statistics.h"
 #include "rocksdb/status.h"
 #include "rocksdb/types.h"
-#include "utilities/blob_db/blob_log_format.h"

 namespace ROCKSDB_NAMESPACE {

 class WritableFileWriter;

-namespace blob_db {
-
 /**
- * Writer is the blob log stream writer. It provides an append-only
+ * BlobLogWriter is the blob log stream writer. It provides an append-only
 * abstraction for writing blob data.
 *
 *
 * Look at blob_db_format.h to see the details of the record formats.
 */

-class Writer {
+class BlobLogWriter {
 public:
  // Create a writer that will append data to "*dest".
  // "*dest" must be initially empty.
-  // "*dest" must remain live while this Writer is in use.
-  Writer(std::unique_ptr<WritableFileWriter>&& dest, Env* env,
-         Statistics* statistics, uint64_t log_number, uint64_t bpsync,
-         bool use_fsync, uint64_t boffset = 0);
+  // "*dest" must remain live while this BlobLogWriter is in use.
+  BlobLogWriter(std::unique_ptr<WritableFileWriter>&& dest, Env* env,
+                Statistics* statistics, uint64_t log_number, bool use_fsync,
+                uint64_t boffset = 0);
  // No copying allowed
-  Writer(const Writer&) = delete;
-  Writer& operator=(const Writer&) = delete;
+  BlobLogWriter(const BlobLogWriter&) = delete;
+  BlobLogWriter& operator=(const BlobLogWriter&) = delete;

-  ~Writer() = default;
+  ~BlobLogWriter() = default;

  static void ConstructBlobHeader(std::string* buf, const Slice& key,
                                  const Slice& val, uint64_t expiration);
@@ -68,20 +66,14 @@ class Writer {

  uint64_t get_log_number() const { return log_number_; }

-  bool ShouldSync() const { return block_offset_ > next_sync_offset_; }
-
  Status Sync();

-  void ResetSyncPointer() { next_sync_offset_ += bytes_per_sync_; }
-
 private:
  std::unique_ptr<WritableFileWriter> dest_;
  Env* env_;
  Statistics* statistics_;
  uint64_t log_number_;
  uint64_t block_offset_;  // Current offset in block
-  uint64_t bytes_per_sync_;
-  uint64_t next_sync_offset_;
  bool use_fsync_;

 public:
@@ -89,6 +81,5 @@ class Writer {
  ElemType last_elem_type_;
 };

-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -46,7 +46,7 @@ class DBBlobIndexTest : public DBTestBase {
  ColumnFamilyHandle* cfh() { return dbfull()->DefaultColumnFamily(); }

  ColumnFamilyData* cfd() {
-    return reinterpret_cast<ColumnFamilyHandleImpl*>(cfh())->cfd();
+    return static_cast_with_check<ColumnFamilyHandleImpl>(cfh())->cfd();
  }

  Status PutBlobIndex(WriteBatch* batch, const Slice& key,
@@ -51,7 +51,8 @@ TableBuilder* NewTableBuilder(
    uint64_t sample_for_compression, const CompressionOptions& compression_opts,
    int level, const bool skip_filters, const uint64_t creation_time,
    const uint64_t oldest_key_time, const uint64_t target_file_size,
-    const uint64_t file_creation_time) {
+    const uint64_t file_creation_time, const std::string& db_id,
+    const std::string& db_session_id) {
  assert((column_family_id ==
          TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
         column_family_name.empty());
@@ -61,7 +62,7 @@ TableBuilder* NewTableBuilder(
                          sample_for_compression, compression_opts,
                          skip_filters, column_family_name, level,
                          creation_time, oldest_key_time, target_file_size,
-                          file_creation_time),
+                          file_creation_time, db_id, db_session_id),
      column_family_id, file);
 }

@@ -85,13 +86,16 @@ Status BuildTable(
    EventLogger* event_logger, int job_id, const Env::IOPriority io_priority,
    TableProperties* table_properties, int level, const uint64_t creation_time,
    const uint64_t oldest_key_time, Env::WriteLifeTimeHint write_hint,
-    const uint64_t file_creation_time) {
+    const uint64_t file_creation_time, const std::string& db_id,
+    const std::string& db_session_id) {
  assert((column_family_id ==
          TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
         column_family_name.empty());
  // Reports the IOStats for flush for every following bytes.
  const size_t kReportFlushIOStatsEvery = 1048576;
+  uint64_t paranoid_hash = 0;
  Status s;
+  IOStatus io_s;
  meta->fd.file_size = 0;
  iter->SeekToFirst();
  std::unique_ptr<CompactionRangeDelAggregator> range_del_agg(
@@ -107,7 +111,6 @@ Status BuildTable(
      ioptions.listeners, dbname, column_family_name, fname, job_id, reason);
 #endif  // !ROCKSDB_LITE
  TableProperties tp;
-
  if (iter->Valid() || !range_del_agg->IsEmpty()) {
    TableBuilder* builder;
    std::unique_ptr<WritableFileWriter> file_writer;
@@ -122,7 +125,11 @@ Status BuildTable(
      bool use_direct_writes = file_options.use_direct_writes;
      TEST_SYNC_POINT_CALLBACK("BuildTable:create_file", &use_direct_writes);
 #endif  // !NDEBUG
-      s = NewWritableFile(fs, fname, &file, file_options);
+      io_s = NewWritableFile(fs, fname, &file, file_options);
+      s = io_s;
+      if (io_status->ok()) {
+        *io_status = io_s;
+      }
      if (!s.ok()) {
        EventHelpers::LogAndNotifyTableFileCreationFinished(
            event_logger, ioptions.listeners, dbname, column_family_name, fname,
@@ -142,7 +149,7 @@ Status BuildTable(
          column_family_name, file_writer.get(), compression,
          sample_for_compression, compression_opts_for_flush, level,
          false /* skip_filters */, creation_time, oldest_key_time,
-          0 /*target_file_size*/, file_creation_time);
+          0 /*target_file_size*/, file_creation_time, db_id, db_session_id);
    }

    MergeHelper merge(env, internal_comparator.user_comparator(),
@@ -161,6 +168,11 @@ Status BuildTable(
      const Slice& key = c_iter.key();
      const Slice& value = c_iter.value();
      const ParsedInternalKey& ikey = c_iter.ikey();
+      if (paranoid_file_checks) {
+        // Generate a rolling 64-bit hash of the key and values
+        paranoid_hash = Hash64(key.data(), key.size(), paranoid_hash);
+        paranoid_hash = Hash64(value.data(), value.size(), paranoid_hash);
+      }
      builder->Add(key, value);
      meta->UpdateBoundaries(key, value, ikey.sequence, ikey.type);

@@ -191,7 +203,10 @@ Status BuildTable(
    } else {
      s = builder->Finish();
    }
-    *io_status = builder->io_status();
+    io_s = builder->io_status();
+    if (io_status->ok()) {
+      *io_status = io_s;
+    }

    if (s.ok() && !empty) {
      uint64_t file_size = builder->FileSize();
@@ -210,16 +225,16 @@ Status BuildTable(
      StopWatch sw(env, ioptions.statistics, TABLE_SYNC_MICROS);
      *io_status = file_writer->Sync(ioptions.use_fsync);
    }
-    if (io_status->ok() && !empty) {
+    if (s.ok() && io_status->ok() && !empty) {
      *io_status = file_writer->Close();
    }
-    if (io_status->ok() && !empty) {
+    if (s.ok() && io_status->ok() && !empty) {
      // Add the checksum information to file metadata.
      meta->file_checksum = file_writer->GetFileChecksum();
      meta->file_checksum_func_name = file_writer->GetFileChecksumFuncName();
    }

-    if (!io_status->ok()) {
+    if (s.ok()) {
      s = *io_status;
    }

@@ -239,14 +254,24 @@ Status BuildTable(
          (internal_stats == nullptr) ? nullptr
                                      : internal_stats->GetFileReadHist(0),
          TableReaderCaller::kFlush, /*arena=*/nullptr,
-          /*skip_filter=*/false, level, /*smallest_compaction_key=*/nullptr,
+          /*skip_filter=*/false, level,
+          MaxFileSizeForL0MetaPin(mutable_cf_options),
+          /*smallest_compaction_key=*/nullptr,
          /*largest_compaction_key*/ nullptr,
          /*allow_unprepared_value*/ false));
      s = it->status();
      if (s.ok() && paranoid_file_checks) {
+        uint64_t check_hash = 0;
        for (it->SeekToFirst(); it->Valid(); it->Next()) {
+          // Generate a rolling 64-bit hash of the key and values
+          check_hash = Hash64(it->key().data(), it->key().size(), check_hash);
+          check_hash =
+              Hash64(it->value().data(), it->value().size(), check_hash);
        }
        s = it->status();
+        if (s.ok() && check_hash != paranoid_hash) {
+          s = Status::Corruption("Paraniod checksums do not match");
+        }
      }
    }
  }
@@ -51,7 +51,8 @@ TableBuilder* NewTableBuilder(
    const CompressionOptions& compression_opts, int level,
    const bool skip_filters = false, const uint64_t creation_time = 0,
    const uint64_t oldest_key_time = 0, const uint64_t target_file_size = 0,
-    const uint64_t file_creation_time = 0);
+    const uint64_t file_creation_time = 0, const std::string& db_id = "",
+    const std::string& db_session_id = "");

 // Build a Table file from the contents of *iter.  The generated file
 // will be named according to number specified in meta. On success, the rest of
@@ -83,6 +84,7 @@ extern Status BuildTable(
    TableProperties* table_properties = nullptr, int level = -1,
    const uint64_t creation_time = 0, const uint64_t oldest_key_time = 0,
    Env::WriteLifeTimeHint write_hint = Env::WLTH_NOT_SET,
-    const uint64_t file_creation_time = 0);
+    const uint64_t file_creation_time = 0, const std::string& db_id = "",
+    const std::string& db_session_id = "");

 }  // namespace ROCKSDB_NAMESPACE
@@ -595,6 +595,15 @@ void rocksdb_backup_engine_restore_db_from_latest_backup(
                                                       restore_options->rep));
 }

+void rocksdb_backup_engine_restore_db_from_backup(
+    rocksdb_backup_engine_t* be, const char* db_dir, const char* wal_dir,
+    const rocksdb_restore_options_t* restore_options, const uint32_t backup_id,
+    char** errptr) {
+  SaveError(errptr, be->rep->RestoreDBFromBackup(backup_id, std::string(db_dir),
+                                                 std::string(wal_dir),
+                                                 restore_options->rep));
+}
+
 const rocksdb_backup_engine_info_t* rocksdb_backup_engine_get_backup_info(
    rocksdb_backup_engine_t* be) {
  rocksdb_backup_engine_info_t* result = new rocksdb_backup_engine_info_t;
@@ -996,6 +1005,55 @@ void rocksdb_multi_get_cf(
  }
 }

+unsigned char rocksdb_key_may_exist(rocksdb_t* db,
+                                    const rocksdb_readoptions_t* options,
+                                    const char* key, size_t key_len,
+                                    char** value, size_t* val_len,
+                                    const char* timestamp, size_t timestamp_len,
+                                    unsigned char* value_found) {
+  std::string tmp;
+  std::string time;
+  if (timestamp) {
+    time.assign(timestamp, timestamp_len);
+  }
+  bool found = false;
+  const bool result = db->rep->KeyMayExist(options->rep, Slice(key, key_len),
+                                           &tmp, timestamp ? &time : nullptr,
+                                           value_found ? &found : nullptr);
+  if (value_found) {
+    *value_found = found;
+    if (found) {
+      *val_len = tmp.size();
+      *value = CopyString(tmp);
+    }
+  }
+  return result;
+}
+
+unsigned char rocksdb_key_may_exist_cf(
+    rocksdb_t* db, const rocksdb_readoptions_t* options,
+    rocksdb_column_family_handle_t* column_family, const char* key,
+    size_t key_len, char** value, size_t* val_len, const char* timestamp,
+    size_t timestamp_len, unsigned char* value_found) {
+  std::string tmp;
+  std::string time;
+  if (timestamp) {
+    time.assign(timestamp, timestamp_len);
+  }
+  bool found = false;
+  const bool result = db->rep->KeyMayExist(
+      options->rep, column_family->rep, Slice(key, key_len), &tmp,
+      timestamp ? &time : nullptr, value_found ? &found : nullptr);
+  if (value_found) {
+    *value_found = found;
+    if (found) {
+      *val_len = tmp.size();
+      *value = CopyString(tmp);
+    }
+  }
+  return result;
+}
+
 rocksdb_iterator_t* rocksdb_create_iterator(
    rocksdb_t* db,
    const rocksdb_readoptions_t* options) {
@@ -1466,6 +1524,11 @@ void rocksdb_writebatch_delete(
  b->rep.Delete(Slice(key, klen));
 }

+void rocksdb_writebatch_singledelete(rocksdb_writebatch_t* b, const char* key,
+                                     size_t klen) {
+  b->rep.SingleDelete(Slice(key, klen));
+}
+
 void rocksdb_writebatch_delete_cf(
    rocksdb_writebatch_t* b,
    rocksdb_column_family_handle_t* column_family,
@@ -1473,6 +1536,12 @@ void rocksdb_writebatch_delete_cf(
  b->rep.Delete(column_family->rep, Slice(key, klen));
 }

+void rocksdb_writebatch_singledelete_cf(
+    rocksdb_writebatch_t* b, rocksdb_column_family_handle_t* column_family,
+    const char* key, size_t klen) {
+  b->rep.SingleDelete(column_family->rep, Slice(key, klen));
+}
+
 void rocksdb_writebatch_deletev(
    rocksdb_writebatch_t* b,
    int num_keys, const char* const* keys_list,
@@ -1723,6 +1792,11 @@ void rocksdb_writebatch_wi_delete(
  b->rep->Delete(Slice(key, klen));
 }

+void rocksdb_writebatch_wi_singledelete(rocksdb_writebatch_wi_t* b,
+                                        const char* key, size_t klen) {
+  b->rep->SingleDelete(Slice(key, klen));
+}
+
 void rocksdb_writebatch_wi_delete_cf(
    rocksdb_writebatch_wi_t* b,
    rocksdb_column_family_handle_t* column_family,
@@ -1730,6 +1804,12 @@ void rocksdb_writebatch_wi_delete_cf(
  b->rep->Delete(column_family->rep, Slice(key, klen));
 }

+void rocksdb_writebatch_wi_singledelete_cf(
+    rocksdb_writebatch_wi_t* b, rocksdb_column_family_handle_t* column_family,
+    const char* key, size_t klen) {
+  b->rep->SingleDelete(column_family->rep, Slice(key, klen));
+}
+
 void rocksdb_writebatch_wi_deletev(
    rocksdb_writebatch_wi_t* b,
    int num_keys, const char* const* keys_list,
@@ -2154,6 +2234,10 @@ void rocksdb_options_destroy(rocksdb_options_t* options) {
  delete options;
 }

+rocksdb_options_t* rocksdb_options_create_copy(rocksdb_options_t* options) {
+  return new rocksdb_options_t(*options);
+}
+
 void rocksdb_options_increase_parallelism(
    rocksdb_options_t* opt, int total_threads) {
  opt->rep.IncreaseParallelism(total_threads);
@@ -2179,6 +2263,10 @@ void rocksdb_options_set_allow_ingest_behind(
  opt->rep.allow_ingest_behind = v;
 }

+unsigned char rocksdb_options_get_allow_ingest_behind(rocksdb_options_t* opt) {
+  return opt->rep.allow_ingest_behind;
+}
+
 void rocksdb_options_set_compaction_filter(
    rocksdb_options_t* opt,
    rocksdb_compactionfilter_t* filter) {
@@ -2196,6 +2284,10 @@ void rocksdb_options_compaction_readahead_size(
  opt->rep.compaction_readahead_size = s;
 }

+size_t rocksdb_options_get_compaction_readahead_size(rocksdb_options_t* opt) {
+  return opt->rep.compaction_readahead_size;
+}
+
 void rocksdb_options_set_comparator(
    rocksdb_options_t* opt,
    rocksdb_comparator_t* cmp) {
@@ -2208,27 +2300,43 @@ void rocksdb_options_set_merge_operator(
  opt->rep.merge_operator = std::shared_ptr<MergeOperator>(merge_operator);
 }

-
 void rocksdb_options_set_create_if_missing(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.create_if_missing = v;
 }

+unsigned char rocksdb_options_get_create_if_missing(rocksdb_options_t* opt) {
+  return opt->rep.create_if_missing;
+}
+
 void rocksdb_options_set_create_missing_column_families(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.create_missing_column_families = v;
 }

+unsigned char rocksdb_options_get_create_missing_column_families(
+    rocksdb_options_t* opt) {
+  return opt->rep.create_missing_column_families;
+}
+
 void rocksdb_options_set_error_if_exists(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.error_if_exists = v;
 }

+unsigned char rocksdb_options_get_error_if_exists(rocksdb_options_t* opt) {
+  return opt->rep.error_if_exists;
+}
+
 void rocksdb_options_set_paranoid_checks(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.paranoid_checks = v;
 }

+unsigned char rocksdb_options_get_paranoid_checks(rocksdb_options_t* opt) {
+  return opt->rep.paranoid_checks;
+}
+
 void rocksdb_options_set_db_paths(rocksdb_options_t* opt,
                                  const rocksdb_dbpath_t** dbpath_values,
                                  size_t num_paths) {
@@ -2254,57 +2362,107 @@ void rocksdb_options_set_info_log_level(
  opt->rep.info_log_level = static_cast<InfoLogLevel>(v);
 }

+int rocksdb_options_get_info_log_level(rocksdb_options_t* opt) {
+  return static_cast<int>(opt->rep.info_log_level);
+}
+
 void rocksdb_options_set_db_write_buffer_size(rocksdb_options_t* opt,
                                              size_t s) {
  opt->rep.db_write_buffer_size = s;
 }

+size_t rocksdb_options_get_db_write_buffer_size(rocksdb_options_t* opt) {
+  return opt->rep.db_write_buffer_size;
+}
+
 void rocksdb_options_set_write_buffer_size(rocksdb_options_t* opt, size_t s) {
  opt->rep.write_buffer_size = s;
 }

+size_t rocksdb_options_get_write_buffer_size(rocksdb_options_t* opt) {
+  return opt->rep.write_buffer_size;
+}
+
 void rocksdb_options_set_max_open_files(rocksdb_options_t* opt, int n) {
  opt->rep.max_open_files = n;
 }

+int rocksdb_options_get_max_open_files(rocksdb_options_t* opt) {
+  return opt->rep.max_open_files;
+}
+
 void rocksdb_options_set_max_file_opening_threads(rocksdb_options_t* opt, int n) {
  opt->rep.max_file_opening_threads = n;
 }

+int rocksdb_options_get_max_file_opening_threads(rocksdb_options_t* opt) {
+  return opt->rep.max_file_opening_threads;
+}
+
 void rocksdb_options_set_max_total_wal_size(rocksdb_options_t* opt, uint64_t n) {
  opt->rep.max_total_wal_size = n;
 }

+uint64_t rocksdb_options_get_max_total_wal_size(rocksdb_options_t* opt) {
+  return opt->rep.max_total_wal_size;
+}
+
 void rocksdb_options_set_target_file_size_base(
    rocksdb_options_t* opt, uint64_t n) {
  opt->rep.target_file_size_base = n;
 }

+uint64_t rocksdb_options_get_target_file_size_base(rocksdb_options_t* opt) {
+  return opt->rep.target_file_size_base;
+}
+
 void rocksdb_options_set_target_file_size_multiplier(
    rocksdb_options_t* opt, int n) {
  opt->rep.target_file_size_multiplier = n;
 }

+int rocksdb_options_get_target_file_size_multiplier(rocksdb_options_t* opt) {
+  return opt->rep.target_file_size_multiplier;
+}
+
 void rocksdb_options_set_max_bytes_for_level_base(
    rocksdb_options_t* opt, uint64_t n) {
  opt->rep.max_bytes_for_level_base = n;
 }

+uint64_t rocksdb_options_get_max_bytes_for_level_base(rocksdb_options_t* opt) {
+  return opt->rep.max_bytes_for_level_base;
+}
+
 void rocksdb_options_set_level_compaction_dynamic_level_bytes(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.level_compaction_dynamic_level_bytes = v;
 }

+unsigned char rocksdb_options_get_level_compaction_dynamic_level_bytes(
+    rocksdb_options_t* opt) {
+  return opt->rep.level_compaction_dynamic_level_bytes;
+}
+
 void rocksdb_options_set_max_bytes_for_level_multiplier(rocksdb_options_t* opt,
                                                        double n) {
  opt->rep.max_bytes_for_level_multiplier = n;
 }

+double rocksdb_options_get_max_bytes_for_level_multiplier(
+    rocksdb_options_t* opt) {
+  return opt->rep.max_bytes_for_level_multiplier;
+}
+
 void rocksdb_options_set_max_compaction_bytes(rocksdb_options_t* opt,
                                              uint64_t n) {
  opt->rep.max_compaction_bytes = n;
 }

+uint64_t rocksdb_options_get_max_compaction_bytes(rocksdb_options_t* opt) {
+  return opt->rep.max_compaction_bytes;
+}
+
 void rocksdb_options_set_max_bytes_for_level_multiplier_additional(
    rocksdb_options_t* opt, int* level_values, size_t num_levels) {
  opt->rep.max_bytes_for_level_multiplier_additional.resize(num_levels);
@@ -2322,30 +2480,57 @@ void rocksdb_options_set_skip_stats_update_on_db_open(rocksdb_options_t* opt,
  opt->rep.skip_stats_update_on_db_open = val;
 }

+unsigned char rocksdb_options_get_skip_stats_update_on_db_open(
+    rocksdb_options_t* opt) {
+  return opt->rep.skip_stats_update_on_db_open;
+}
+
 void rocksdb_options_set_skip_checking_sst_file_sizes_on_db_open(
    rocksdb_options_t* opt, unsigned char val) {
  opt->rep.skip_checking_sst_file_sizes_on_db_open = val;
 }

+unsigned char rocksdb_options_get_skip_checking_sst_file_sizes_on_db_open(
+    rocksdb_options_t* opt) {
+  return opt->rep.skip_checking_sst_file_sizes_on_db_open;
+}
+
 void rocksdb_options_set_num_levels(rocksdb_options_t* opt, int n) {
  opt->rep.num_levels = n;
 }

+int rocksdb_options_get_num_levels(rocksdb_options_t* opt) {
+  return opt->rep.num_levels;
+}
+
 void rocksdb_options_set_level0_file_num_compaction_trigger(
    rocksdb_options_t* opt, int n) {
  opt->rep.level0_file_num_compaction_trigger = n;
 }

+int rocksdb_options_get_level0_file_num_compaction_trigger(
+    rocksdb_options_t* opt) {
+  return opt->rep.level0_file_num_compaction_trigger;
+}
+
 void rocksdb_options_set_level0_slowdown_writes_trigger(
    rocksdb_options_t* opt, int n) {
  opt->rep.level0_slowdown_writes_trigger = n;
 }

+int rocksdb_options_get_level0_slowdown_writes_trigger(rocksdb_options_t* opt) {
+  return opt->rep.level0_slowdown_writes_trigger;
+}
+
 void rocksdb_options_set_level0_stop_writes_trigger(
    rocksdb_options_t* opt, int n) {
  opt->rep.level0_stop_writes_trigger = n;
 }

+int rocksdb_options_get_level0_stop_writes_trigger(rocksdb_options_t* opt) {
+  return opt->rep.level0_stop_writes_trigger;
+}
+
 void rocksdb_options_set_max_mem_compaction_level(rocksdb_options_t* /*opt*/,
                                                  int /*n*/) {}

@@ -2353,10 +2538,26 @@ void rocksdb_options_set_wal_recovery_mode(rocksdb_options_t* opt,int mode) {
  opt->rep.wal_recovery_mode = static_cast<WALRecoveryMode>(mode);
 }

+int rocksdb_options_get_wal_recovery_mode(rocksdb_options_t* opt) {
+  return static_cast<int>(opt->rep.wal_recovery_mode);
+}
+
 void rocksdb_options_set_compression(rocksdb_options_t* opt, int t) {
  opt->rep.compression = static_cast<CompressionType>(t);
 }

+int rocksdb_options_get_compression(rocksdb_options_t* opt) {
+  return opt->rep.compression;
+}
+
+void rocksdb_options_set_bottommost_compression(rocksdb_options_t* opt, int t) {
+  opt->rep.bottommost_compression = static_cast<CompressionType>(t);
+}
+
+int rocksdb_options_get_bottommost_compression(rocksdb_options_t* opt) {
+  return opt->rep.bottommost_compression;
+}
+
 void rocksdb_options_set_compression_per_level(rocksdb_options_t* opt,
                                               int* level_values,
                                               size_t num_levels) {
@@ -2371,7 +2572,7 @@ void rocksdb_options_set_bottommost_compression_options(rocksdb_options_t* opt,
                                                        int w_bits, int level,
                                                        int strategy,
                                                        int max_dict_bytes,
-                                                        bool enabled) {
+                                                        unsigned char enabled) {
  opt->rep.bottommost_compression_opts.window_bits = w_bits;
  opt->rep.bottommost_compression_opts.level = level;
  opt->rep.bottommost_compression_opts.strategy = strategy;
@@ -2379,6 +2580,13 @@ void rocksdb_options_set_bottommost_compression_options(rocksdb_options_t* opt,
  opt->rep.bottommost_compression_opts.enabled = enabled;
 }

+void rocksdb_options_set_bottommost_compression_options_zstd_max_train_bytes(
+    rocksdb_options_t* opt, int zstd_max_train_bytes, unsigned char enabled) {
+  opt->rep.bottommost_compression_opts.zstd_max_train_bytes =
+      zstd_max_train_bytes;
+  opt->rep.bottommost_compression_opts.enabled = enabled;
+}
+
 void rocksdb_options_set_compression_options(rocksdb_options_t* opt, int w_bits,
                                             int level, int strategy,
                                             int max_dict_bytes) {
@@ -2388,6 +2596,11 @@ void rocksdb_options_set_compression_options(rocksdb_options_t* opt, int w_bits,
  opt->rep.compression_opts.max_dict_bytes = max_dict_bytes;
 }

+void rocksdb_options_set_compression_options_zstd_max_train_bytes(
+    rocksdb_options_t* opt, int zstd_max_train_bytes) {
+  opt->rep.compression_opts.zstd_max_train_bytes = zstd_max_train_bytes;
+}
+
 void rocksdb_options_set_prefix_extractor(
    rocksdb_options_t* opt, rocksdb_slicetransform_t* prefix_extractor) {
  opt->rep.prefix_extractor.reset(prefix_extractor);
@@ -2398,6 +2611,10 @@ void rocksdb_options_set_use_fsync(
  opt->rep.use_fsync = use_fsync;
 }

+int rocksdb_options_get_use_fsync(rocksdb_options_t* opt) {
+  return opt->rep.use_fsync;
+}
+
 void rocksdb_options_set_db_log_dir(
    rocksdb_options_t* opt, const char* db_log_dir) {
  opt->rep.db_log_dir = db_log_dir;
@@ -2412,16 +2629,28 @@ void rocksdb_options_set_WAL_ttl_seconds(rocksdb_options_t* opt, uint64_t ttl) {
  opt->rep.WAL_ttl_seconds = ttl;
 }

+uint64_t rocksdb_options_get_WAL_ttl_seconds(rocksdb_options_t* opt) {
+  return opt->rep.WAL_ttl_seconds;
+}
+
 void rocksdb_options_set_WAL_size_limit_MB(
    rocksdb_options_t* opt, uint64_t limit) {
  opt->rep.WAL_size_limit_MB = limit;
 }

+uint64_t rocksdb_options_get_WAL_size_limit_MB(rocksdb_options_t* opt) {
+  return opt->rep.WAL_size_limit_MB;
+}
+
 void rocksdb_options_set_manifest_preallocation_size(
    rocksdb_options_t* opt, size_t v) {
  opt->rep.manifest_preallocation_size = v;
 }

+size_t rocksdb_options_get_manifest_preallocation_size(rocksdb_options_t* opt) {
+  return opt->rep.manifest_preallocation_size;
+}
+
 // noop
 void rocksdb_options_set_purge_redundant_kvs_while_flush(
    rocksdb_options_t* /*opt*/, unsigned char /*v*/) {}
@@ -2431,41 +2660,86 @@ void rocksdb_options_set_use_direct_reads(rocksdb_options_t* opt,
  opt->rep.use_direct_reads = v;
 }

+unsigned char rocksdb_options_get_use_direct_reads(rocksdb_options_t* opt) {
+  return opt->rep.use_direct_reads;
+}
+
 void rocksdb_options_set_use_direct_io_for_flush_and_compaction(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.use_direct_io_for_flush_and_compaction = v;
 }

+unsigned char rocksdb_options_get_use_direct_io_for_flush_and_compaction(
+    rocksdb_options_t* opt) {
+  return opt->rep.use_direct_io_for_flush_and_compaction;
+}
+
 void rocksdb_options_set_allow_mmap_reads(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.allow_mmap_reads = v;
 }

+unsigned char rocksdb_options_get_allow_mmap_reads(rocksdb_options_t* opt) {
+  return opt->rep.allow_mmap_reads;
+}
+
 void rocksdb_options_set_allow_mmap_writes(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.allow_mmap_writes = v;
 }

+unsigned char rocksdb_options_get_allow_mmap_writes(rocksdb_options_t* opt) {
+  return opt->rep.allow_mmap_writes;
+}
+
 void rocksdb_options_set_is_fd_close_on_exec(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.is_fd_close_on_exec = v;
 }

+unsigned char rocksdb_options_get_is_fd_close_on_exec(rocksdb_options_t* opt) {
+  return opt->rep.is_fd_close_on_exec;
+}
+
 void rocksdb_options_set_skip_log_error_on_recovery(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.skip_log_error_on_recovery = v;
 }

+unsigned char rocksdb_options_get_skip_log_error_on_recovery(
+    rocksdb_options_t* opt) {
+  return opt->rep.skip_log_error_on_recovery;
+}
+
 void rocksdb_options_set_stats_dump_period_sec(
    rocksdb_options_t* opt, unsigned int v) {
  opt->rep.stats_dump_period_sec = v;
 }

+unsigned int rocksdb_options_get_stats_dump_period_sec(rocksdb_options_t* opt) {
+  return opt->rep.stats_dump_period_sec;
+}
+
+void rocksdb_options_set_stats_persist_period_sec(rocksdb_options_t* opt,
+                                                  unsigned int v) {
+  opt->rep.stats_persist_period_sec = v;
+}
+
+unsigned int rocksdb_options_get_stats_persist_period_sec(
+    rocksdb_options_t* opt) {
+  return opt->rep.stats_persist_period_sec;
+}
+
 void rocksdb_options_set_advise_random_on_open(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.advise_random_on_open = v;
 }

+unsigned char rocksdb_options_get_advise_random_on_open(
+    rocksdb_options_t* opt) {
+  return opt->rep.advise_random_on_open;
+}
+
 void rocksdb_options_set_access_hint_on_compaction_start(
    rocksdb_options_t* opt, int v) {
  switch(v) {
@@ -2488,139 +2762,271 @@ void rocksdb_options_set_access_hint_on_compaction_start(
  }
 }

+int rocksdb_options_get_access_hint_on_compaction_start(
+    rocksdb_options_t* opt) {
+  return opt->rep.access_hint_on_compaction_start;
+}
+
 void rocksdb_options_set_use_adaptive_mutex(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.use_adaptive_mutex = v;
 }

+unsigned char rocksdb_options_get_use_adaptive_mutex(rocksdb_options_t* opt) {
+  return opt->rep.use_adaptive_mutex;
+}
+
 void rocksdb_options_set_wal_bytes_per_sync(
    rocksdb_options_t* opt, uint64_t v) {
  opt->rep.wal_bytes_per_sync = v;
 }

+uint64_t rocksdb_options_get_wal_bytes_per_sync(rocksdb_options_t* opt) {
+  return opt->rep.wal_bytes_per_sync;
+}
+
 void rocksdb_options_set_bytes_per_sync(
    rocksdb_options_t* opt, uint64_t v) {
  opt->rep.bytes_per_sync = v;
 }

+uint64_t rocksdb_options_get_bytes_per_sync(rocksdb_options_t* opt) {
+  return opt->rep.bytes_per_sync;
+}
+
 void rocksdb_options_set_writable_file_max_buffer_size(rocksdb_options_t* opt,
                                                       uint64_t v) {
  opt->rep.writable_file_max_buffer_size = static_cast<size_t>(v);
 }

+uint64_t rocksdb_options_get_writable_file_max_buffer_size(
+    rocksdb_options_t* opt) {
+  return opt->rep.writable_file_max_buffer_size;
+}
+
 void rocksdb_options_set_allow_concurrent_memtable_write(rocksdb_options_t* opt,
                                                         unsigned char v) {
  opt->rep.allow_concurrent_memtable_write = v;
 }

+unsigned char rocksdb_options_get_allow_concurrent_memtable_write(
+    rocksdb_options_t* opt) {
+  return opt->rep.allow_concurrent_memtable_write;
+}
+
 void rocksdb_options_set_enable_write_thread_adaptive_yield(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.enable_write_thread_adaptive_yield = v;
 }

+unsigned char rocksdb_options_get_enable_write_thread_adaptive_yield(
+    rocksdb_options_t* opt) {
+  return opt->rep.enable_write_thread_adaptive_yield;
+}
+
 void rocksdb_options_set_max_sequential_skip_in_iterations(
    rocksdb_options_t* opt, uint64_t v) {
  opt->rep.max_sequential_skip_in_iterations = v;
 }

+uint64_t rocksdb_options_get_max_sequential_skip_in_iterations(
+    rocksdb_options_t* opt) {
+  return opt->rep.max_sequential_skip_in_iterations;
+}
+
 void rocksdb_options_set_max_write_buffer_number(rocksdb_options_t* opt, int n) {
  opt->rep.max_write_buffer_number = n;
 }

+int rocksdb_options_get_max_write_buffer_number(rocksdb_options_t* opt) {
+  return opt->rep.max_write_buffer_number;
+}
+
 void rocksdb_options_set_min_write_buffer_number_to_merge(rocksdb_options_t* opt, int n) {
  opt->rep.min_write_buffer_number_to_merge = n;
 }

+int rocksdb_options_get_min_write_buffer_number_to_merge(
+    rocksdb_options_t* opt) {
+  return opt->rep.min_write_buffer_number_to_merge;
+}
+
 void rocksdb_options_set_max_write_buffer_number_to_maintain(
    rocksdb_options_t* opt, int n) {
  opt->rep.max_write_buffer_number_to_maintain = n;
 }

+int rocksdb_options_get_max_write_buffer_number_to_maintain(
+    rocksdb_options_t* opt) {
+  return opt->rep.max_write_buffer_number_to_maintain;
+}
+
 void rocksdb_options_set_max_write_buffer_size_to_maintain(
    rocksdb_options_t* opt, int64_t n) {
  opt->rep.max_write_buffer_size_to_maintain = n;
 }

+int64_t rocksdb_options_get_max_write_buffer_size_to_maintain(
+    rocksdb_options_t* opt) {
+  return opt->rep.max_write_buffer_size_to_maintain;
+}
+
 void rocksdb_options_set_enable_pipelined_write(rocksdb_options_t* opt,
                                                unsigned char v) {
  opt->rep.enable_pipelined_write = v;
 }

+unsigned char rocksdb_options_get_enable_pipelined_write(
+    rocksdb_options_t* opt) {
+  return opt->rep.enable_pipelined_write;
+}
+
 void rocksdb_options_set_unordered_write(rocksdb_options_t* opt,
                                         unsigned char v) {
  opt->rep.unordered_write = v;
 }

+unsigned char rocksdb_options_get_unordered_write(rocksdb_options_t* opt) {
+  return opt->rep.unordered_write;
+}
+
 void rocksdb_options_set_max_subcompactions(rocksdb_options_t* opt,
                                            uint32_t n) {
  opt->rep.max_subcompactions = n;
 }

+uint32_t rocksdb_options_get_max_subcompactions(rocksdb_options_t* opt) {
+  return opt->rep.max_subcompactions;
+}
+
 void rocksdb_options_set_max_background_jobs(rocksdb_options_t* opt, int n) {
  opt->rep.max_background_jobs = n;
 }

+int rocksdb_options_get_max_background_jobs(rocksdb_options_t* opt) {
+  return opt->rep.max_background_jobs;
+}
+
 void rocksdb_options_set_max_background_compactions(rocksdb_options_t* opt, int n) {
  opt->rep.max_background_compactions = n;
 }

+int rocksdb_options_get_max_background_compactions(rocksdb_options_t* opt) {
+  return opt->rep.max_background_compactions;
+}
+
 void rocksdb_options_set_base_background_compactions(rocksdb_options_t* opt,
                                                     int n) {
  opt->rep.base_background_compactions = n;
 }

+int rocksdb_options_get_base_background_compactions(rocksdb_options_t* opt) {
+  return opt->rep.base_background_compactions;
+}
+
 void rocksdb_options_set_max_background_flushes(rocksdb_options_t* opt, int n) {
  opt->rep.max_background_flushes = n;
 }

+int rocksdb_options_get_max_background_flushes(rocksdb_options_t* opt) {
+  return opt->rep.max_background_flushes;
+}
+
 void rocksdb_options_set_max_log_file_size(rocksdb_options_t* opt, size_t v) {
  opt->rep.max_log_file_size = v;
 }

+size_t rocksdb_options_get_max_log_file_size(rocksdb_options_t* opt) {
+  return opt->rep.max_log_file_size;
+}
+
 void rocksdb_options_set_log_file_time_to_roll(rocksdb_options_t* opt, size_t v) {
  opt->rep.log_file_time_to_roll = v;
 }

+size_t rocksdb_options_get_log_file_time_to_roll(rocksdb_options_t* opt) {
+  return opt->rep.log_file_time_to_roll;
+}
+
 void rocksdb_options_set_keep_log_file_num(rocksdb_options_t* opt, size_t v) {
  opt->rep.keep_log_file_num = v;
 }

+size_t rocksdb_options_get_keep_log_file_num(rocksdb_options_t* opt) {
+  return opt->rep.keep_log_file_num;
+}
+
 void rocksdb_options_set_recycle_log_file_num(rocksdb_options_t* opt,
                                              size_t v) {
  opt->rep.recycle_log_file_num = v;
 }

+size_t rocksdb_options_get_recycle_log_file_num(rocksdb_options_t* opt) {
+  return opt->rep.recycle_log_file_num;
+}
+
 void rocksdb_options_set_soft_rate_limit(rocksdb_options_t* opt, double v) {
  opt->rep.soft_rate_limit = v;
 }

+double rocksdb_options_get_soft_rate_limit(rocksdb_options_t* opt) {
+  return opt->rep.soft_rate_limit;
+}
+
 void rocksdb_options_set_hard_rate_limit(rocksdb_options_t* opt, double v) {
  opt->rep.hard_rate_limit = v;
 }

+double rocksdb_options_get_hard_rate_limit(rocksdb_options_t* opt) {
+  return opt->rep.hard_rate_limit;
+}
+
 void rocksdb_options_set_soft_pending_compaction_bytes_limit(rocksdb_options_t* opt, size_t v) {
  opt->rep.soft_pending_compaction_bytes_limit = v;
 }

+size_t rocksdb_options_get_soft_pending_compaction_bytes_limit(
+    rocksdb_options_t* opt) {
+  return opt->rep.soft_pending_compaction_bytes_limit;
+}
+
 void rocksdb_options_set_hard_pending_compaction_bytes_limit(rocksdb_options_t* opt, size_t v) {
  opt->rep.hard_pending_compaction_bytes_limit = v;
 }

+size_t rocksdb_options_get_hard_pending_compaction_bytes_limit(
+    rocksdb_options_t* opt) {
+  return opt->rep.hard_pending_compaction_bytes_limit;
+}
+
 void rocksdb_options_set_rate_limit_delay_max_milliseconds(
    rocksdb_options_t* opt, unsigned int v) {
  opt->rep.rate_limit_delay_max_milliseconds = v;
 }

+unsigned int rocksdb_options_get_rate_limit_delay_max_milliseconds(
+    rocksdb_options_t* opt) {
+  return opt->rep.rate_limit_delay_max_milliseconds;
+}
+
 void rocksdb_options_set_max_manifest_file_size(
    rocksdb_options_t* opt, size_t v) {
  opt->rep.max_manifest_file_size = v;
 }

+size_t rocksdb_options_get_max_manifest_file_size(rocksdb_options_t* opt) {
+  return opt->rep.max_manifest_file_size;
+}
+
 void rocksdb_options_set_table_cache_numshardbits(
    rocksdb_options_t* opt, int v) {
  opt->rep.table_cache_numshardbits = v;
 }

+int rocksdb_options_get_table_cache_numshardbits(rocksdb_options_t* opt) {
+  return opt->rep.table_cache_numshardbits;
+}
+
 void rocksdb_options_set_table_cache_remove_scan_count_limit(
    rocksdb_options_t* /*opt*/, int /*v*/) {
  // this option is deprecated
@@ -2631,19 +3037,38 @@ void rocksdb_options_set_arena_block_size(
  opt->rep.arena_block_size = v;
 }

+size_t rocksdb_options_get_arena_block_size(rocksdb_options_t* opt) {
+  return opt->rep.arena_block_size;
+}
+
 void rocksdb_options_set_disable_auto_compactions(rocksdb_options_t* opt, int disable) {
  opt->rep.disable_auto_compactions = disable;
 }

+unsigned char rocksdb_options_get_disable_auto_compactions(
+    rocksdb_options_t* opt) {
+  return opt->rep.disable_auto_compactions;
+}
+
 void rocksdb_options_set_optimize_filters_for_hits(rocksdb_options_t* opt, int v) {
  opt->rep.optimize_filters_for_hits = v;
 }

+unsigned char rocksdb_options_get_optimize_filters_for_hits(
+    rocksdb_options_t* opt) {
+  return opt->rep.optimize_filters_for_hits;
+}
+
 void rocksdb_options_set_delete_obsolete_files_period_micros(
    rocksdb_options_t* opt, uint64_t v) {
  opt->rep.delete_obsolete_files_period_micros = v;
 }

+uint64_t rocksdb_options_get_delete_obsolete_files_period_micros(
+    rocksdb_options_t* opt) {
+  return opt->rep.delete_obsolete_files_period_micros;
+}
+
 void rocksdb_options_prepare_for_bulk_load(rocksdb_options_t* opt) {
  opt->rep.PrepareForBulkLoad();
 }
@@ -2657,11 +3082,20 @@ void rocksdb_options_set_memtable_prefix_bloom_size_ratio(
  opt->rep.memtable_prefix_bloom_size_ratio = v;
 }

+double rocksdb_options_get_memtable_prefix_bloom_size_ratio(
+    rocksdb_options_t* opt) {
+  return opt->rep.memtable_prefix_bloom_size_ratio;
+}
+
 void rocksdb_options_set_memtable_huge_page_size(rocksdb_options_t* opt,
                                                 size_t v) {
  opt->rep.memtable_huge_page_size = v;
 }

+size_t rocksdb_options_get_memtable_huge_page_size(rocksdb_options_t* opt) {
+  return opt->rep.memtable_huge_page_size;
+}
+
 void rocksdb_options_set_hash_skip_list_rep(
    rocksdb_options_t *opt, size_t bucket_count,
    int32_t skiplist_height, int32_t skiplist_branching_factor) {
@@ -2696,31 +3130,56 @@ void rocksdb_options_set_max_successive_merges(
  opt->rep.max_successive_merges = v;
 }

+size_t rocksdb_options_get_max_successive_merges(rocksdb_options_t* opt) {
+  return opt->rep.max_successive_merges;
+}
+
 void rocksdb_options_set_bloom_locality(
    rocksdb_options_t* opt, uint32_t v) {
  opt->rep.bloom_locality = v;
 }

+uint32_t rocksdb_options_get_bloom_locality(rocksdb_options_t* opt) {
+  return opt->rep.bloom_locality;
+}
+
 void rocksdb_options_set_inplace_update_support(
    rocksdb_options_t* opt, unsigned char v) {
  opt->rep.inplace_update_support = v;
 }

+unsigned char rocksdb_options_get_inplace_update_support(
+    rocksdb_options_t* opt) {
+  return opt->rep.inplace_update_support;
+}
+
 void rocksdb_options_set_inplace_update_num_locks(
    rocksdb_options_t* opt, size_t v) {
  opt->rep.inplace_update_num_locks = v;
 }

+size_t rocksdb_options_get_inplace_update_num_locks(rocksdb_options_t* opt) {
+  return opt->rep.inplace_update_num_locks;
+}
+
 void rocksdb_options_set_report_bg_io_stats(
    rocksdb_options_t* opt, int v) {
  opt->rep.report_bg_io_stats = v;
 }

+unsigned char rocksdb_options_get_report_bg_io_stats(rocksdb_options_t* opt) {
+  return opt->rep.report_bg_io_stats;
+}
+
 void rocksdb_options_set_compaction_style(rocksdb_options_t *opt, int style) {
  opt->rep.compaction_style =
      static_cast<ROCKSDB_NAMESPACE::CompactionStyle>(style);
 }

+int rocksdb_options_get_compaction_style(rocksdb_options_t* opt) {
+  return opt->rep.compaction_style;
+}
+
 void rocksdb_options_set_universal_compaction_options(rocksdb_options_t *opt, rocksdb_universal_compaction_options_t *uco) {
  opt->rep.compaction_options_universal = *(uco->rep);
 }
@@ -2750,6 +3209,10 @@ void rocksdb_options_set_atomic_flush(rocksdb_options_t* opt,
  opt->rep.atomic_flush = atomic_flush;
 }

+unsigned char rocksdb_options_get_atomic_flush(rocksdb_options_t* opt) {
+  return opt->rep.atomic_flush;
+}
+
 rocksdb_ratelimiter_t* rocksdb_ratelimiter_create(
    int64_t rate_bytes_per_sec,
    int64_t refill_period_us,
@@ -4456,6 +4919,10 @@ void rocksdb_approximate_memory_usage_destroy(rocksdb_memory_usage_t* usage) {
  delete usage;
 }

+void rocksdb_cancel_all_background_work(rocksdb_t* db, unsigned char wait) {
+  CancelAllBackgroundWork(db->rep, wait);
+}
+
 }  // end extern "C"

 #endif  // !ROCKSDB_LITE
@@ -1296,6 +1296,29 @@ int main(int argc, char** argv) {
      Free(&vals[i]);
    }

+    {
+      unsigned char value_found = 0;
+
+      CheckCondition(!rocksdb_key_may_exist(db, roptions, "invalid_key", 11,
+                                            NULL, NULL, NULL, 0, NULL));
+      CheckCondition(!rocksdb_key_may_exist(db, roptions, "invalid_key", 11,
+                                            &vals[0], &vals_sizes[0], NULL, 0,
+                                            &value_found));
+      if (value_found) {
+        Free(&vals[0]);
+      }
+
+      CheckCondition(!rocksdb_key_may_exist_cf(db, roptions, handles[1],
+                                               "invalid_key", 11, NULL, NULL,
+                                               NULL, 0, NULL));
+      CheckCondition(!rocksdb_key_may_exist_cf(db, roptions, handles[1],
+                                               "invalid_key", 11, &vals[0],
+                                               &vals_sizes[0], NULL, 0, NULL));
+      if (value_found) {
+        Free(&vals[0]);
+      }
+    }
+
    rocksdb_iterator_t* iter = rocksdb_create_iterator_cf(db, roptions, handles[1]);
    CheckCondition(!rocksdb_iter_valid(iter));
    rocksdb_iter_seek_to_first(iter);
@@ -1461,6 +1484,757 @@ int main(int argc, char** argv) {
    rocksdb_cuckoo_options_destroy(cuckoo_options);
  }

+  StartPhase("options");
+  {
+    rocksdb_options_t* o;
+    o = rocksdb_options_create();
+
+    // Set and check options.
+    rocksdb_options_set_allow_ingest_behind(o, 1);
+    CheckCondition(1 == rocksdb_options_get_allow_ingest_behind(o));
+
+    rocksdb_options_compaction_readahead_size(o, 10);
+    CheckCondition(10 == rocksdb_options_get_compaction_readahead_size(o));
+
+    rocksdb_options_set_create_if_missing(o, 1);
+    CheckCondition(1 == rocksdb_options_get_create_if_missing(o));
+
+    rocksdb_options_set_create_missing_column_families(o, 1);
+    CheckCondition(1 == rocksdb_options_get_create_missing_column_families(o));
+
+    rocksdb_options_set_error_if_exists(o, 1);
+    CheckCondition(1 == rocksdb_options_get_error_if_exists(o));
+
+    rocksdb_options_set_paranoid_checks(o, 1);
+    CheckCondition(1 == rocksdb_options_get_paranoid_checks(o));
+
+    rocksdb_options_set_info_log_level(o, 3);
+    CheckCondition(3 == rocksdb_options_get_info_log_level(o));
+
+    rocksdb_options_set_write_buffer_size(o, 100);
+    CheckCondition(100 == rocksdb_options_get_write_buffer_size(o));
+
+    rocksdb_options_set_db_write_buffer_size(o, 1000);
+    CheckCondition(1000 == rocksdb_options_get_db_write_buffer_size(o));
+
+    rocksdb_options_set_max_open_files(o, 21);
+    CheckCondition(21 == rocksdb_options_get_max_open_files(o));
+
+    rocksdb_options_set_max_file_opening_threads(o, 5);
+    CheckCondition(5 == rocksdb_options_get_max_file_opening_threads(o));
+
+    rocksdb_options_set_max_total_wal_size(o, 400);
+    CheckCondition(400 == rocksdb_options_get_max_total_wal_size(o));
+
+    rocksdb_options_set_num_levels(o, 7);
+    CheckCondition(7 == rocksdb_options_get_num_levels(o));
+
+    rocksdb_options_set_level0_file_num_compaction_trigger(o, 4);
+    CheckCondition(4 ==
+                   rocksdb_options_get_level0_file_num_compaction_trigger(o));
+
+    rocksdb_options_set_level0_slowdown_writes_trigger(o, 6);
+    CheckCondition(6 == rocksdb_options_get_level0_slowdown_writes_trigger(o));
+
+    rocksdb_options_set_level0_stop_writes_trigger(o, 8);
+    CheckCondition(8 == rocksdb_options_get_level0_stop_writes_trigger(o));
+
+    rocksdb_options_set_target_file_size_base(o, 256);
+    CheckCondition(256 == rocksdb_options_get_target_file_size_base(o));
+
+    rocksdb_options_set_target_file_size_multiplier(o, 3);
+    CheckCondition(3 == rocksdb_options_get_target_file_size_multiplier(o));
+
+    rocksdb_options_set_max_bytes_for_level_base(o, 1024);
+    CheckCondition(1024 == rocksdb_options_get_max_bytes_for_level_base(o));
+
+    rocksdb_options_set_level_compaction_dynamic_level_bytes(o, 1);
+    CheckCondition(1 ==
+                   rocksdb_options_get_level_compaction_dynamic_level_bytes(o));
+
+    rocksdb_options_set_max_bytes_for_level_multiplier(o, 2.0);
+    CheckCondition(2.0 ==
+                   rocksdb_options_get_max_bytes_for_level_multiplier(o));
+
+    rocksdb_options_set_skip_stats_update_on_db_open(o, 1);
+    CheckCondition(1 == rocksdb_options_get_skip_stats_update_on_db_open(o));
+
+    rocksdb_options_set_skip_checking_sst_file_sizes_on_db_open(o, 1);
+    CheckCondition(
+        1 == rocksdb_options_get_skip_checking_sst_file_sizes_on_db_open(o));
+
+    rocksdb_options_set_max_write_buffer_number(o, 97);
+    CheckCondition(97 == rocksdb_options_get_max_write_buffer_number(o));
+
+    rocksdb_options_set_min_write_buffer_number_to_merge(o, 23);
+    CheckCondition(23 ==
+                   rocksdb_options_get_min_write_buffer_number_to_merge(o));
+
+    rocksdb_options_set_max_write_buffer_number_to_maintain(o, 64);
+    CheckCondition(64 ==
+                   rocksdb_options_get_max_write_buffer_number_to_maintain(o));
+
+    rocksdb_options_set_max_write_buffer_size_to_maintain(o, 50000);
+    CheckCondition(50000 ==
+                   rocksdb_options_get_max_write_buffer_size_to_maintain(o));
+
+    rocksdb_options_set_enable_pipelined_write(o, 1);
+    CheckCondition(1 == rocksdb_options_get_enable_pipelined_write(o));
+
+    rocksdb_options_set_unordered_write(o, 1);
+    CheckCondition(1 == rocksdb_options_get_unordered_write(o));
+
+    rocksdb_options_set_max_subcompactions(o, 123456);
+    CheckCondition(123456 == rocksdb_options_get_max_subcompactions(o));
+
+    rocksdb_options_set_max_background_jobs(o, 2);
+    CheckCondition(2 == rocksdb_options_get_max_background_jobs(o));
+
+    rocksdb_options_set_max_background_compactions(o, 3);
+    CheckCondition(3 == rocksdb_options_get_max_background_compactions(o));
+
+    rocksdb_options_set_base_background_compactions(o, 4);
+    CheckCondition(4 == rocksdb_options_get_base_background_compactions(o));
+
+    rocksdb_options_set_max_background_flushes(o, 5);
+    CheckCondition(5 == rocksdb_options_get_max_background_flushes(o));
+
+    rocksdb_options_set_max_log_file_size(o, 6);
+    CheckCondition(6 == rocksdb_options_get_max_log_file_size(o));
+
+    rocksdb_options_set_log_file_time_to_roll(o, 7);
+    CheckCondition(7 == rocksdb_options_get_log_file_time_to_roll(o));
+
+    rocksdb_options_set_keep_log_file_num(o, 8);
+    CheckCondition(8 == rocksdb_options_get_keep_log_file_num(o));
+
+    rocksdb_options_set_recycle_log_file_num(o, 9);
+    CheckCondition(9 == rocksdb_options_get_recycle_log_file_num(o));
+
+    rocksdb_options_set_soft_rate_limit(o, 2.0);
+    CheckCondition(2.0 == rocksdb_options_get_soft_rate_limit(o));
+
+    rocksdb_options_set_hard_rate_limit(o, 4.0);
+    CheckCondition(4.0 == rocksdb_options_get_hard_rate_limit(o));
+
+    rocksdb_options_set_soft_pending_compaction_bytes_limit(o, 10);
+    CheckCondition(10 ==
+                   rocksdb_options_get_soft_pending_compaction_bytes_limit(o));
+
+    rocksdb_options_set_hard_pending_compaction_bytes_limit(o, 11);
+    CheckCondition(11 ==
+                   rocksdb_options_get_hard_pending_compaction_bytes_limit(o));
+
+    rocksdb_options_set_rate_limit_delay_max_milliseconds(o, 1);
+    CheckCondition(1 ==
+                   rocksdb_options_get_rate_limit_delay_max_milliseconds(o));
+
+    rocksdb_options_set_max_manifest_file_size(o, 12);
+    CheckCondition(12 == rocksdb_options_get_max_manifest_file_size(o));
+
+    rocksdb_options_set_table_cache_numshardbits(o, 13);
+    CheckCondition(13 == rocksdb_options_get_table_cache_numshardbits(o));
+
+    rocksdb_options_set_arena_block_size(o, 14);
+    CheckCondition(14 == rocksdb_options_get_arena_block_size(o));
+
+    rocksdb_options_set_use_fsync(o, 1);
+    CheckCondition(1 == rocksdb_options_get_use_fsync(o));
+
+    rocksdb_options_set_WAL_ttl_seconds(o, 15);
+    CheckCondition(15 == rocksdb_options_get_WAL_ttl_seconds(o));
+
+    rocksdb_options_set_WAL_size_limit_MB(o, 16);
+    CheckCondition(16 == rocksdb_options_get_WAL_size_limit_MB(o));
+
+    rocksdb_options_set_manifest_preallocation_size(o, 17);
+    CheckCondition(17 == rocksdb_options_get_manifest_preallocation_size(o));
+
+    rocksdb_options_set_allow_mmap_reads(o, 1);
+    CheckCondition(1 == rocksdb_options_get_allow_mmap_reads(o));
+
+    rocksdb_options_set_allow_mmap_writes(o, 1);
+    CheckCondition(1 == rocksdb_options_get_allow_mmap_writes(o));
+
+    rocksdb_options_set_use_direct_reads(o, 1);
+    CheckCondition(1 == rocksdb_options_get_use_direct_reads(o));
+
+    rocksdb_options_set_use_direct_io_for_flush_and_compaction(o, 1);
+    CheckCondition(
+        1 == rocksdb_options_get_use_direct_io_for_flush_and_compaction(o));
+
+    rocksdb_options_set_is_fd_close_on_exec(o, 1);
+    CheckCondition(1 == rocksdb_options_get_is_fd_close_on_exec(o));
+
+    rocksdb_options_set_skip_log_error_on_recovery(o, 1);
+    CheckCondition(1 == rocksdb_options_get_skip_log_error_on_recovery(o));
+
+    rocksdb_options_set_stats_dump_period_sec(o, 18);
+    CheckCondition(18 == rocksdb_options_get_stats_dump_period_sec(o));
+
+    rocksdb_options_set_stats_persist_period_sec(o, 5);
+    CheckCondition(5 == rocksdb_options_get_stats_persist_period_sec(o));
+
+    rocksdb_options_set_advise_random_on_open(o, 1);
+    CheckCondition(1 == rocksdb_options_get_advise_random_on_open(o));
+
+    rocksdb_options_set_access_hint_on_compaction_start(o, 3);
+    CheckCondition(3 == rocksdb_options_get_access_hint_on_compaction_start(o));
+
+    rocksdb_options_set_use_adaptive_mutex(o, 1);
+    CheckCondition(1 == rocksdb_options_get_use_adaptive_mutex(o));
+
+    rocksdb_options_set_bytes_per_sync(o, 19);
+    CheckCondition(19 == rocksdb_options_get_bytes_per_sync(o));
+
+    rocksdb_options_set_wal_bytes_per_sync(o, 20);
+    CheckCondition(20 == rocksdb_options_get_wal_bytes_per_sync(o));
+
+    rocksdb_options_set_writable_file_max_buffer_size(o, 21);
+    CheckCondition(21 == rocksdb_options_get_writable_file_max_buffer_size(o));
+
+    rocksdb_options_set_allow_concurrent_memtable_write(o, 1);
+    CheckCondition(1 == rocksdb_options_get_allow_concurrent_memtable_write(o));
+
+    rocksdb_options_set_enable_write_thread_adaptive_yield(o, 1);
+    CheckCondition(1 ==
+                   rocksdb_options_get_enable_write_thread_adaptive_yield(o));
+
+    rocksdb_options_set_max_sequential_skip_in_iterations(o, 22);
+    CheckCondition(22 ==
+                   rocksdb_options_get_max_sequential_skip_in_iterations(o));
+
+    rocksdb_options_set_disable_auto_compactions(o, 1);
+    CheckCondition(1 == rocksdb_options_get_disable_auto_compactions(o));
+
+    rocksdb_options_set_optimize_filters_for_hits(o, 1);
+    CheckCondition(1 == rocksdb_options_get_optimize_filters_for_hits(o));
+
+    rocksdb_options_set_delete_obsolete_files_period_micros(o, 23);
+    CheckCondition(23 ==
+                   rocksdb_options_get_delete_obsolete_files_period_micros(o));
+
+    rocksdb_options_set_memtable_prefix_bloom_size_ratio(o, 2.0);
+    CheckCondition(2.0 ==
+                   rocksdb_options_get_memtable_prefix_bloom_size_ratio(o));
+
+    rocksdb_options_set_max_compaction_bytes(o, 24);
+    CheckCondition(24 == rocksdb_options_get_max_compaction_bytes(o));
+
+    rocksdb_options_set_memtable_huge_page_size(o, 25);
+    CheckCondition(25 == rocksdb_options_get_memtable_huge_page_size(o));
+
+    rocksdb_options_set_max_successive_merges(o, 26);
+    CheckCondition(26 == rocksdb_options_get_max_successive_merges(o));
+
+    rocksdb_options_set_bloom_locality(o, 27);
+    CheckCondition(27 == rocksdb_options_get_bloom_locality(o));
+
+    rocksdb_options_set_inplace_update_support(o, 1);
+    CheckCondition(1 == rocksdb_options_get_inplace_update_support(o));
+
+    rocksdb_options_set_inplace_update_num_locks(o, 28);
+    CheckCondition(28 == rocksdb_options_get_inplace_update_num_locks(o));
+
+    rocksdb_options_set_report_bg_io_stats(o, 1);
+    CheckCondition(1 == rocksdb_options_get_report_bg_io_stats(o));
+
+    rocksdb_options_set_wal_recovery_mode(o, 2);
+    CheckCondition(2 == rocksdb_options_get_wal_recovery_mode(o));
+
+    rocksdb_options_set_compression(o, 5);
+    CheckCondition(5 == rocksdb_options_get_compression(o));
+
+    rocksdb_options_set_bottommost_compression(o, 4);
+    CheckCondition(4 == rocksdb_options_get_bottommost_compression(o));
+
+    rocksdb_options_set_compaction_style(o, 2);
+    CheckCondition(2 == rocksdb_options_get_compaction_style(o));
+
+    rocksdb_options_set_atomic_flush(o, 1);
+    CheckCondition(1 == rocksdb_options_get_atomic_flush(o));
+
+    // Create a copy that should be equal to the original.
+    rocksdb_options_t* copy;
+    copy = rocksdb_options_create_copy(o);
+
+    CheckCondition(1 == rocksdb_options_get_allow_ingest_behind(copy));
+    CheckCondition(10 == rocksdb_options_get_compaction_readahead_size(copy));
+    CheckCondition(1 == rocksdb_options_get_create_if_missing(copy));
+    CheckCondition(1 ==
+                   rocksdb_options_get_create_missing_column_families(copy));
+    CheckCondition(1 == rocksdb_options_get_error_if_exists(copy));
+    CheckCondition(1 == rocksdb_options_get_paranoid_checks(copy));
+    CheckCondition(3 == rocksdb_options_get_info_log_level(copy));
+    CheckCondition(100 == rocksdb_options_get_write_buffer_size(copy));
+    CheckCondition(1000 == rocksdb_options_get_db_write_buffer_size(copy));
+    CheckCondition(21 == rocksdb_options_get_max_open_files(copy));
+    CheckCondition(5 == rocksdb_options_get_max_file_opening_threads(copy));
+    CheckCondition(400 == rocksdb_options_get_max_total_wal_size(copy));
+    CheckCondition(7 == rocksdb_options_get_num_levels(copy));
+    CheckCondition(
+        4 == rocksdb_options_get_level0_file_num_compaction_trigger(copy));
+    CheckCondition(6 ==
+                   rocksdb_options_get_level0_slowdown_writes_trigger(copy));
+    CheckCondition(8 == rocksdb_options_get_level0_stop_writes_trigger(copy));
+    CheckCondition(256 == rocksdb_options_get_target_file_size_base(copy));
+    CheckCondition(3 == rocksdb_options_get_target_file_size_multiplier(copy));
+    CheckCondition(1024 == rocksdb_options_get_max_bytes_for_level_base(copy));
+    CheckCondition(
+        1 == rocksdb_options_get_level_compaction_dynamic_level_bytes(copy));
+    CheckCondition(2.0 ==
+                   rocksdb_options_get_max_bytes_for_level_multiplier(copy));
+    CheckCondition(1 == rocksdb_options_get_skip_stats_update_on_db_open(copy));
+    CheckCondition(
+        1 == rocksdb_options_get_skip_checking_sst_file_sizes_on_db_open(copy));
+    CheckCondition(97 == rocksdb_options_get_max_write_buffer_number(copy));
+    CheckCondition(23 ==
+                   rocksdb_options_get_min_write_buffer_number_to_merge(copy));
+    CheckCondition(
+        64 == rocksdb_options_get_max_write_buffer_number_to_maintain(copy));
+    CheckCondition(50000 ==
+                   rocksdb_options_get_max_write_buffer_size_to_maintain(copy));
+    CheckCondition(1 == rocksdb_options_get_enable_pipelined_write(copy));
+    CheckCondition(1 == rocksdb_options_get_unordered_write(copy));
+    CheckCondition(123456 == rocksdb_options_get_max_subcompactions(copy));
+    CheckCondition(2 == rocksdb_options_get_max_background_jobs(copy));
+    CheckCondition(3 == rocksdb_options_get_max_background_compactions(copy));
+    CheckCondition(4 == rocksdb_options_get_base_background_compactions(copy));
+    CheckCondition(5 == rocksdb_options_get_max_background_flushes(copy));
+    CheckCondition(6 == rocksdb_options_get_max_log_file_size(copy));
+    CheckCondition(7 == rocksdb_options_get_log_file_time_to_roll(copy));
+    CheckCondition(8 == rocksdb_options_get_keep_log_file_num(copy));
+    CheckCondition(9 == rocksdb_options_get_recycle_log_file_num(copy));
+    CheckCondition(2.0 == rocksdb_options_get_soft_rate_limit(copy));
+    CheckCondition(4.0 == rocksdb_options_get_hard_rate_limit(copy));
+    CheckCondition(
+        10 == rocksdb_options_get_soft_pending_compaction_bytes_limit(copy));
+    CheckCondition(
+        11 == rocksdb_options_get_hard_pending_compaction_bytes_limit(copy));
+    CheckCondition(1 ==
+                   rocksdb_options_get_rate_limit_delay_max_milliseconds(copy));
+    CheckCondition(12 == rocksdb_options_get_max_manifest_file_size(copy));
+    CheckCondition(13 == rocksdb_options_get_table_cache_numshardbits(copy));
+    CheckCondition(14 == rocksdb_options_get_arena_block_size(copy));
+    CheckCondition(1 == rocksdb_options_get_use_fsync(copy));
+    CheckCondition(15 == rocksdb_options_get_WAL_ttl_seconds(copy));
+    CheckCondition(16 == rocksdb_options_get_WAL_size_limit_MB(copy));
+    CheckCondition(17 == rocksdb_options_get_manifest_preallocation_size(copy));
+    CheckCondition(1 == rocksdb_options_get_allow_mmap_reads(copy));
+    CheckCondition(1 == rocksdb_options_get_allow_mmap_writes(copy));
+    CheckCondition(1 == rocksdb_options_get_use_direct_reads(copy));
+    CheckCondition(
+        1 == rocksdb_options_get_use_direct_io_for_flush_and_compaction(copy));
+    CheckCondition(1 == rocksdb_options_get_is_fd_close_on_exec(copy));
+    CheckCondition(1 == rocksdb_options_get_skip_log_error_on_recovery(copy));
+    CheckCondition(18 == rocksdb_options_get_stats_dump_period_sec(copy));
+    CheckCondition(5 == rocksdb_options_get_stats_persist_period_sec(copy));
+    CheckCondition(1 == rocksdb_options_get_advise_random_on_open(copy));
+    CheckCondition(3 ==
+                   rocksdb_options_get_access_hint_on_compaction_start(copy));
+    CheckCondition(1 == rocksdb_options_get_use_adaptive_mutex(copy));
+    CheckCondition(19 == rocksdb_options_get_bytes_per_sync(copy));
+    CheckCondition(20 == rocksdb_options_get_wal_bytes_per_sync(copy));
+    CheckCondition(21 ==
+                   rocksdb_options_get_writable_file_max_buffer_size(copy));
+    CheckCondition(1 ==
+                   rocksdb_options_get_allow_concurrent_memtable_write(copy));
+    CheckCondition(
+        1 == rocksdb_options_get_enable_write_thread_adaptive_yield(copy));
+    CheckCondition(22 ==
+                   rocksdb_options_get_max_sequential_skip_in_iterations(copy));
+    CheckCondition(1 == rocksdb_options_get_disable_auto_compactions(copy));
+    CheckCondition(1 == rocksdb_options_get_optimize_filters_for_hits(copy));
+    CheckCondition(
+        23 == rocksdb_options_get_delete_obsolete_files_period_micros(copy));
+    CheckCondition(2.0 ==
+                   rocksdb_options_get_memtable_prefix_bloom_size_ratio(copy));
+    CheckCondition(24 == rocksdb_options_get_max_compaction_bytes(copy));
+    CheckCondition(25 == rocksdb_options_get_memtable_huge_page_size(copy));
+    CheckCondition(26 == rocksdb_options_get_max_successive_merges(copy));
+    CheckCondition(27 == rocksdb_options_get_bloom_locality(copy));
+    CheckCondition(1 == rocksdb_options_get_inplace_update_support(copy));
+    CheckCondition(28 == rocksdb_options_get_inplace_update_num_locks(copy));
+    CheckCondition(1 == rocksdb_options_get_report_bg_io_stats(copy));
+    CheckCondition(2 == rocksdb_options_get_wal_recovery_mode(copy));
+    CheckCondition(5 == rocksdb_options_get_compression(copy));
+    CheckCondition(4 == rocksdb_options_get_bottommost_compression(copy));
+    CheckCondition(2 == rocksdb_options_get_compaction_style(copy));
+    CheckCondition(1 == rocksdb_options_get_atomic_flush(copy));
+
+    // Copies should be independent.
+    rocksdb_options_set_allow_ingest_behind(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_allow_ingest_behind(copy));
+    CheckCondition(1 == rocksdb_options_get_allow_ingest_behind(o));
+
+    rocksdb_options_compaction_readahead_size(copy, 20);
+    CheckCondition(20 == rocksdb_options_get_compaction_readahead_size(copy));
+    CheckCondition(10 == rocksdb_options_get_compaction_readahead_size(o));
+
+    rocksdb_options_set_create_if_missing(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_create_if_missing(copy));
+    CheckCondition(1 == rocksdb_options_get_create_if_missing(o));
+
+    rocksdb_options_set_create_missing_column_families(copy, 0);
+    CheckCondition(0 ==
+                   rocksdb_options_get_create_missing_column_families(copy));
+    CheckCondition(1 == rocksdb_options_get_create_missing_column_families(o));
+
+    rocksdb_options_set_error_if_exists(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_error_if_exists(copy));
+    CheckCondition(1 == rocksdb_options_get_error_if_exists(o));
+
+    rocksdb_options_set_paranoid_checks(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_paranoid_checks(copy));
+    CheckCondition(1 == rocksdb_options_get_paranoid_checks(o));
+
+    rocksdb_options_set_info_log_level(copy, 2);
+    CheckCondition(2 == rocksdb_options_get_info_log_level(copy));
+    CheckCondition(3 == rocksdb_options_get_info_log_level(o));
+
+    rocksdb_options_set_write_buffer_size(copy, 200);
+    CheckCondition(200 == rocksdb_options_get_write_buffer_size(copy));
+    CheckCondition(100 == rocksdb_options_get_write_buffer_size(o));
+
+    rocksdb_options_set_db_write_buffer_size(copy, 2000);
+    CheckCondition(2000 == rocksdb_options_get_db_write_buffer_size(copy));
+    CheckCondition(1000 == rocksdb_options_get_db_write_buffer_size(o));
+
+    rocksdb_options_set_max_open_files(copy, 42);
+    CheckCondition(42 == rocksdb_options_get_max_open_files(copy));
+    CheckCondition(21 == rocksdb_options_get_max_open_files(o));
+
+    rocksdb_options_set_max_file_opening_threads(copy, 3);
+    CheckCondition(3 == rocksdb_options_get_max_file_opening_threads(copy));
+    CheckCondition(5 == rocksdb_options_get_max_file_opening_threads(o));
+
+    rocksdb_options_set_max_total_wal_size(copy, 4000);
+    CheckCondition(4000 == rocksdb_options_get_max_total_wal_size(copy));
+    CheckCondition(400 == rocksdb_options_get_max_total_wal_size(o));
+
+    rocksdb_options_set_num_levels(copy, 6);
+    CheckCondition(6 == rocksdb_options_get_num_levels(copy));
+    CheckCondition(7 == rocksdb_options_get_num_levels(o));
+
+    rocksdb_options_set_level0_file_num_compaction_trigger(copy, 14);
+    CheckCondition(
+        14 == rocksdb_options_get_level0_file_num_compaction_trigger(copy));
+    CheckCondition(4 ==
+                   rocksdb_options_get_level0_file_num_compaction_trigger(o));
+
+    rocksdb_options_set_level0_slowdown_writes_trigger(copy, 61);
+    CheckCondition(61 ==
+                   rocksdb_options_get_level0_slowdown_writes_trigger(copy));
+    CheckCondition(6 == rocksdb_options_get_level0_slowdown_writes_trigger(o));
+
+    rocksdb_options_set_level0_stop_writes_trigger(copy, 17);
+    CheckCondition(17 == rocksdb_options_get_level0_stop_writes_trigger(copy));
+    CheckCondition(8 == rocksdb_options_get_level0_stop_writes_trigger(o));
+
+    rocksdb_options_set_target_file_size_base(copy, 128);
+    CheckCondition(128 == rocksdb_options_get_target_file_size_base(copy));
+    CheckCondition(256 == rocksdb_options_get_target_file_size_base(o));
+
+    rocksdb_options_set_target_file_size_multiplier(copy, 13);
+    CheckCondition(13 == rocksdb_options_get_target_file_size_multiplier(copy));
+    CheckCondition(3 == rocksdb_options_get_target_file_size_multiplier(o));
+
+    rocksdb_options_set_max_bytes_for_level_base(copy, 900);
+    CheckCondition(900 == rocksdb_options_get_max_bytes_for_level_base(copy));
+    CheckCondition(1024 == rocksdb_options_get_max_bytes_for_level_base(o));
+
+    rocksdb_options_set_level_compaction_dynamic_level_bytes(copy, 0);
+    CheckCondition(
+        0 == rocksdb_options_get_level_compaction_dynamic_level_bytes(copy));
+    CheckCondition(1 ==
+                   rocksdb_options_get_level_compaction_dynamic_level_bytes(o));
+
+    rocksdb_options_set_max_bytes_for_level_multiplier(copy, 8.0);
+    CheckCondition(8.0 ==
+                   rocksdb_options_get_max_bytes_for_level_multiplier(copy));
+    CheckCondition(2.0 ==
+                   rocksdb_options_get_max_bytes_for_level_multiplier(o));
+
+    rocksdb_options_set_skip_stats_update_on_db_open(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_skip_stats_update_on_db_open(copy));
+    CheckCondition(1 == rocksdb_options_get_skip_stats_update_on_db_open(o));
+
+    rocksdb_options_set_skip_checking_sst_file_sizes_on_db_open(copy, 0);
+    CheckCondition(
+        0 == rocksdb_options_get_skip_checking_sst_file_sizes_on_db_open(copy));
+    CheckCondition(
+        1 == rocksdb_options_get_skip_checking_sst_file_sizes_on_db_open(o));
+
+    rocksdb_options_set_max_write_buffer_number(copy, 2000);
+    CheckCondition(2000 == rocksdb_options_get_max_write_buffer_number(copy));
+    CheckCondition(97 == rocksdb_options_get_max_write_buffer_number(o));
+
+    rocksdb_options_set_min_write_buffer_number_to_merge(copy, 146);
+    CheckCondition(146 ==
+                   rocksdb_options_get_min_write_buffer_number_to_merge(copy));
+    CheckCondition(23 ==
+                   rocksdb_options_get_min_write_buffer_number_to_merge(o));
+
+    rocksdb_options_set_max_write_buffer_number_to_maintain(copy, 128);
+    CheckCondition(
+        128 == rocksdb_options_get_max_write_buffer_number_to_maintain(copy));
+    CheckCondition(64 ==
+                   rocksdb_options_get_max_write_buffer_number_to_maintain(o));
+
+    rocksdb_options_set_max_write_buffer_size_to_maintain(copy, 9000);
+    CheckCondition(9000 ==
+                   rocksdb_options_get_max_write_buffer_size_to_maintain(copy));
+    CheckCondition(50000 ==
+                   rocksdb_options_get_max_write_buffer_size_to_maintain(o));
+
+    rocksdb_options_set_enable_pipelined_write(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_enable_pipelined_write(copy));
+    CheckCondition(1 == rocksdb_options_get_enable_pipelined_write(o));
+
+    rocksdb_options_set_unordered_write(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_unordered_write(copy));
+    CheckCondition(1 == rocksdb_options_get_unordered_write(o));
+
+    rocksdb_options_set_max_subcompactions(copy, 90001);
+    CheckCondition(90001 == rocksdb_options_get_max_subcompactions(copy));
+    CheckCondition(123456 == rocksdb_options_get_max_subcompactions(o));
+
+    rocksdb_options_set_max_background_jobs(copy, 12);
+    CheckCondition(12 == rocksdb_options_get_max_background_jobs(copy));
+    CheckCondition(2 == rocksdb_options_get_max_background_jobs(o));
+
+    rocksdb_options_set_max_background_compactions(copy, 13);
+    CheckCondition(13 == rocksdb_options_get_max_background_compactions(copy));
+    CheckCondition(3 == rocksdb_options_get_max_background_compactions(o));
+
+    rocksdb_options_set_base_background_compactions(copy, 14);
+    CheckCondition(14 == rocksdb_options_get_base_background_compactions(copy));
+    CheckCondition(4 == rocksdb_options_get_base_background_compactions(o));
+
+    rocksdb_options_set_max_background_flushes(copy, 15);
+    CheckCondition(15 == rocksdb_options_get_max_background_flushes(copy));
+    CheckCondition(5 == rocksdb_options_get_max_background_flushes(o));
+
+    rocksdb_options_set_max_log_file_size(copy, 16);
+    CheckCondition(16 == rocksdb_options_get_max_log_file_size(copy));
+    CheckCondition(6 == rocksdb_options_get_max_log_file_size(o));
+
+    rocksdb_options_set_log_file_time_to_roll(copy, 17);
+    CheckCondition(17 == rocksdb_options_get_log_file_time_to_roll(copy));
+    CheckCondition(7 == rocksdb_options_get_log_file_time_to_roll(o));
+
+    rocksdb_options_set_keep_log_file_num(copy, 18);
+    CheckCondition(18 == rocksdb_options_get_keep_log_file_num(copy));
+    CheckCondition(8 == rocksdb_options_get_keep_log_file_num(o));
+
+    rocksdb_options_set_recycle_log_file_num(copy, 19);
+    CheckCondition(19 == rocksdb_options_get_recycle_log_file_num(copy));
+    CheckCondition(9 == rocksdb_options_get_recycle_log_file_num(o));
+
+    rocksdb_options_set_soft_rate_limit(copy, 4.0);
+    CheckCondition(4.0 == rocksdb_options_get_soft_rate_limit(copy));
+    CheckCondition(2.0 == rocksdb_options_get_soft_rate_limit(o));
+
+    rocksdb_options_set_hard_rate_limit(copy, 2.0);
+    CheckCondition(2.0 == rocksdb_options_get_hard_rate_limit(copy));
+    CheckCondition(4.0 == rocksdb_options_get_hard_rate_limit(o));
+
+    rocksdb_options_set_soft_pending_compaction_bytes_limit(copy, 110);
+    CheckCondition(
+        110 == rocksdb_options_get_soft_pending_compaction_bytes_limit(copy));
+    CheckCondition(10 ==
+                   rocksdb_options_get_soft_pending_compaction_bytes_limit(o));
+
+    rocksdb_options_set_hard_pending_compaction_bytes_limit(copy, 111);
+    CheckCondition(
+        111 == rocksdb_options_get_hard_pending_compaction_bytes_limit(copy));
+    CheckCondition(11 ==
+                   rocksdb_options_get_hard_pending_compaction_bytes_limit(o));
+
+    rocksdb_options_set_rate_limit_delay_max_milliseconds(copy, 0);
+    CheckCondition(0 ==
+                   rocksdb_options_get_rate_limit_delay_max_milliseconds(copy));
+    CheckCondition(1 ==
+                   rocksdb_options_get_rate_limit_delay_max_milliseconds(o));
+
+    rocksdb_options_set_max_manifest_file_size(copy, 112);
+    CheckCondition(112 == rocksdb_options_get_max_manifest_file_size(copy));
+    CheckCondition(12 == rocksdb_options_get_max_manifest_file_size(o));
+
+    rocksdb_options_set_table_cache_numshardbits(copy, 113);
+    CheckCondition(113 == rocksdb_options_get_table_cache_numshardbits(copy));
+    CheckCondition(13 == rocksdb_options_get_table_cache_numshardbits(o));
+
+    rocksdb_options_set_arena_block_size(copy, 114);
+    CheckCondition(114 == rocksdb_options_get_arena_block_size(copy));
+    CheckCondition(14 == rocksdb_options_get_arena_block_size(o));
+
+    rocksdb_options_set_use_fsync(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_use_fsync(copy));
+    CheckCondition(1 == rocksdb_options_get_use_fsync(o));
+
+    rocksdb_options_set_WAL_ttl_seconds(copy, 115);
+    CheckCondition(115 == rocksdb_options_get_WAL_ttl_seconds(copy));
+    CheckCondition(15 == rocksdb_options_get_WAL_ttl_seconds(o));
+
+    rocksdb_options_set_WAL_size_limit_MB(copy, 116);
+    CheckCondition(116 == rocksdb_options_get_WAL_size_limit_MB(copy));
+    CheckCondition(16 == rocksdb_options_get_WAL_size_limit_MB(o));
+
+    rocksdb_options_set_manifest_preallocation_size(copy, 117);
+    CheckCondition(117 ==
+                   rocksdb_options_get_manifest_preallocation_size(copy));
+    CheckCondition(17 == rocksdb_options_get_manifest_preallocation_size(o));
+
+    rocksdb_options_set_allow_mmap_reads(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_allow_mmap_reads(copy));
+    CheckCondition(1 == rocksdb_options_get_allow_mmap_reads(o));
+
+    rocksdb_options_set_allow_mmap_writes(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_allow_mmap_writes(copy));
+    CheckCondition(1 == rocksdb_options_get_allow_mmap_writes(o));
+
+    rocksdb_options_set_use_direct_reads(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_use_direct_reads(copy));
+    CheckCondition(1 == rocksdb_options_get_use_direct_reads(o));
+
+    rocksdb_options_set_use_direct_io_for_flush_and_compaction(copy, 0);
+    CheckCondition(
+        0 == rocksdb_options_get_use_direct_io_for_flush_and_compaction(copy));
+    CheckCondition(
+        1 == rocksdb_options_get_use_direct_io_for_flush_and_compaction(o));
+
+    rocksdb_options_set_is_fd_close_on_exec(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_is_fd_close_on_exec(copy));
+    CheckCondition(1 == rocksdb_options_get_is_fd_close_on_exec(o));
+
+    rocksdb_options_set_skip_log_error_on_recovery(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_skip_log_error_on_recovery(copy));
+    CheckCondition(1 == rocksdb_options_get_skip_log_error_on_recovery(o));
+
+    rocksdb_options_set_stats_dump_period_sec(copy, 218);
+    CheckCondition(218 == rocksdb_options_get_stats_dump_period_sec(copy));
+    CheckCondition(18 == rocksdb_options_get_stats_dump_period_sec(o));
+
+    rocksdb_options_set_stats_persist_period_sec(copy, 600);
+    CheckCondition(600 == rocksdb_options_get_stats_persist_period_sec(copy));
+    CheckCondition(5 == rocksdb_options_get_stats_persist_period_sec(o));
+
+    rocksdb_options_set_advise_random_on_open(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_advise_random_on_open(copy));
+    CheckCondition(1 == rocksdb_options_get_advise_random_on_open(o));
+
+    rocksdb_options_set_access_hint_on_compaction_start(copy, 2);
+    CheckCondition(2 ==
+                   rocksdb_options_get_access_hint_on_compaction_start(copy));
+    CheckCondition(3 == rocksdb_options_get_access_hint_on_compaction_start(o));
+
+    rocksdb_options_set_use_adaptive_mutex(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_use_adaptive_mutex(copy));
+    CheckCondition(1 == rocksdb_options_get_use_adaptive_mutex(o));
+
+    rocksdb_options_set_bytes_per_sync(copy, 219);
+    CheckCondition(219 == rocksdb_options_get_bytes_per_sync(copy));
+    CheckCondition(19 == rocksdb_options_get_bytes_per_sync(o));
+
+    rocksdb_options_set_wal_bytes_per_sync(copy, 120);
+    CheckCondition(120 == rocksdb_options_get_wal_bytes_per_sync(copy));
+    CheckCondition(20 == rocksdb_options_get_wal_bytes_per_sync(o));
+
+    rocksdb_options_set_writable_file_max_buffer_size(copy, 121);
+    CheckCondition(121 ==
+                   rocksdb_options_get_writable_file_max_buffer_size(copy));
+    CheckCondition(21 == rocksdb_options_get_writable_file_max_buffer_size(o));
+
+    rocksdb_options_set_allow_concurrent_memtable_write(copy, 0);
+    CheckCondition(0 ==
+                   rocksdb_options_get_allow_concurrent_memtable_write(copy));
+    CheckCondition(1 == rocksdb_options_get_allow_concurrent_memtable_write(o));
+
+    rocksdb_options_set_enable_write_thread_adaptive_yield(copy, 0);
+    CheckCondition(
+        0 == rocksdb_options_get_enable_write_thread_adaptive_yield(copy));
+    CheckCondition(1 ==
+                   rocksdb_options_get_enable_write_thread_adaptive_yield(o));
+
+    rocksdb_options_set_max_sequential_skip_in_iterations(copy, 122);
+    CheckCondition(122 ==
+                   rocksdb_options_get_max_sequential_skip_in_iterations(copy));
+    CheckCondition(22 ==
+                   rocksdb_options_get_max_sequential_skip_in_iterations(o));
+
+    rocksdb_options_set_disable_auto_compactions(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_disable_auto_compactions(copy));
+    CheckCondition(1 == rocksdb_options_get_disable_auto_compactions(o));
+
+    rocksdb_options_set_optimize_filters_for_hits(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_optimize_filters_for_hits(copy));
+    CheckCondition(1 == rocksdb_options_get_optimize_filters_for_hits(o));
+
+    rocksdb_options_set_delete_obsolete_files_period_micros(copy, 123);
+    CheckCondition(
+        123 == rocksdb_options_get_delete_obsolete_files_period_micros(copy));
+    CheckCondition(23 ==
+                   rocksdb_options_get_delete_obsolete_files_period_micros(o));
+
+    rocksdb_options_set_memtable_prefix_bloom_size_ratio(copy, 4.0);
+    CheckCondition(4.0 ==
+                   rocksdb_options_get_memtable_prefix_bloom_size_ratio(copy));
+    CheckCondition(2.0 ==
+                   rocksdb_options_get_memtable_prefix_bloom_size_ratio(o));
+
+    rocksdb_options_set_max_compaction_bytes(copy, 124);
+    CheckCondition(124 == rocksdb_options_get_max_compaction_bytes(copy));
+    CheckCondition(24 == rocksdb_options_get_max_compaction_bytes(o));
+
+    rocksdb_options_set_memtable_huge_page_size(copy, 125);
+    CheckCondition(125 == rocksdb_options_get_memtable_huge_page_size(copy));
+    CheckCondition(25 == rocksdb_options_get_memtable_huge_page_size(o));
+
+    rocksdb_options_set_max_successive_merges(copy, 126);
+    CheckCondition(126 == rocksdb_options_get_max_successive_merges(copy));
+    CheckCondition(26 == rocksdb_options_get_max_successive_merges(o));
+
+    rocksdb_options_set_bloom_locality(copy, 127);
+    CheckCondition(127 == rocksdb_options_get_bloom_locality(copy));
+    CheckCondition(27 == rocksdb_options_get_bloom_locality(o));
+
+    rocksdb_options_set_inplace_update_support(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_inplace_update_support(copy));
+    CheckCondition(1 == rocksdb_options_get_inplace_update_support(o));
+
+    rocksdb_options_set_inplace_update_num_locks(copy, 128);
+    CheckCondition(128 == rocksdb_options_get_inplace_update_num_locks(copy));
+    CheckCondition(28 == rocksdb_options_get_inplace_update_num_locks(o));
+
+    rocksdb_options_set_report_bg_io_stats(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_report_bg_io_stats(copy));
+    CheckCondition(1 == rocksdb_options_get_report_bg_io_stats(o));
+
+    rocksdb_options_set_wal_recovery_mode(copy, 1);
+    CheckCondition(1 == rocksdb_options_get_wal_recovery_mode(copy));
+    CheckCondition(2 == rocksdb_options_get_wal_recovery_mode(o));
+
+    rocksdb_options_set_compression(copy, 4);
+    CheckCondition(4 == rocksdb_options_get_compression(copy));
+    CheckCondition(5 == rocksdb_options_get_compression(o));
+
+    rocksdb_options_set_bottommost_compression(copy, 3);
+    CheckCondition(3 == rocksdb_options_get_bottommost_compression(copy));
+    CheckCondition(4 == rocksdb_options_get_bottommost_compression(o));
+
+    rocksdb_options_set_compaction_style(copy, 1);
+    CheckCondition(1 == rocksdb_options_get_compaction_style(copy));
+    CheckCondition(2 == rocksdb_options_get_compaction_style(o));
+
+    rocksdb_options_set_atomic_flush(copy, 0);
+    CheckCondition(0 == rocksdb_options_get_atomic_flush(copy));
+    CheckCondition(1 == rocksdb_options_get_atomic_flush(o));
+
+    rocksdb_options_destroy(copy);
+    rocksdb_options_destroy(o);
+  }
+
  StartPhase("iterate_upper_bound");
  {
    // Create new empty database
@@ -1840,6 +2614,9 @@ int main(int argc, char** argv) {
    CheckNoError(err);
  }

+  StartPhase("cancel_all_background_work");
+  rocksdb_cancel_all_background_work(db, 1);
+
  StartPhase("cleanup");
  rocksdb_close(db);
  rocksdb_options_destroy(options);
@@ -34,6 +34,7 @@
 #include "table/block_based/block_based_table_factory.h"
 #include "table/merging_iterator.h"
 #include "util/autovector.h"
+#include "util/cast_util.h"
 #include "util/compression.h"

 namespace ROCKSDB_NAMESPACE {
@@ -346,7 +347,7 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
  }

  bool is_block_based_table =
-      (result.table_factory->Name() == BlockBasedTableFactory().Name());
+      (result.table_factory->Name() == BlockBasedTableFactory::kName);

  const uint64_t kAdjustedTtl = 30 * 24 * 60 * 60;
  if (result.ttl == kDefaultTtl) {
@@ -1046,13 +1047,14 @@ bool ColumnFamilyData::NeedsCompaction() const {
 }

 Compaction* ColumnFamilyData::PickCompaction(
-    const MutableCFOptions& mutable_options, LogBuffer* log_buffer) {
+    const MutableCFOptions& mutable_options,
+    const MutableDBOptions& mutable_db_options, LogBuffer* log_buffer) {
  SequenceNumber earliest_mem_seqno =
      std::min(mem_->GetEarliestSequenceNumber(),
               imm_.current()->GetEarliestSequenceNumber(false));
  auto* result = compaction_picker_->PickCompaction(
-      GetName(), mutable_options, current_->storage_info(), log_buffer,
-      earliest_mem_seqno);
+      GetName(), mutable_options, mutable_db_options, current_->storage_info(),
+      log_buffer, earliest_mem_seqno);
  if (result != nullptr) {
    result->SetInputVersion(current_);
  }
@@ -1122,14 +1124,16 @@ const int ColumnFamilyData::kCompactAllLevels = -1;
 const int ColumnFamilyData::kCompactToBaseLevel = -2;

 Compaction* ColumnFamilyData::CompactRange(
-    const MutableCFOptions& mutable_cf_options, int input_level,
+    const MutableCFOptions& mutable_cf_options,
+    const MutableDBOptions& mutable_db_options, int input_level,
    int output_level, const CompactRangeOptions& compact_range_options,
    const InternalKey* begin, const InternalKey* end,
    InternalKey** compaction_end, bool* conflict,
    uint64_t max_file_num_to_ignore) {
  auto* result = compaction_picker_->CompactRange(
-      GetName(), mutable_cf_options, current_->storage_info(), input_level,
-      output_level, compact_range_options, begin, end, compaction_end, conflict,
+      GetName(), mutable_cf_options, mutable_db_options,
+      current_->storage_info(), input_level, output_level,
+      compact_range_options, begin, end, compaction_end, conflict,
      max_file_num_to_ignore);
  if (result != nullptr) {
    result->SetInputVersion(current_);
@@ -1299,7 +1303,7 @@ Status ColumnFamilyData::ValidateOptions(
  }

  if (cf_options.ttl > 0 && cf_options.ttl != kDefaultTtl) {
-    if (cf_options.table_factory->Name() != BlockBasedTableFactory().Name()) {
+    if (cf_options.table_factory->Name() != BlockBasedTableFactory::kName) {
      return Status::NotSupported(
          "TTL is only supported in Block-Based Table format. ");
    }
@@ -1307,7 +1311,7 @@ Status ColumnFamilyData::ValidateOptions(

  if (cf_options.periodic_compaction_seconds > 0 &&
      cf_options.periodic_compaction_seconds != kDefaultPeriodicCompSecs) {
-    if (cf_options.table_factory->Name() != BlockBasedTableFactory().Name()) {
+    if (cf_options.table_factory->Name() != BlockBasedTableFactory::kName) {
      return Status::NotSupported(
          "Periodic Compaction is only supported in "
          "Block-Based Table format. ");
@@ -1546,7 +1550,7 @@ ColumnFamilyHandle* ColumnFamilyMemTablesImpl::GetColumnFamilyHandle() {
 uint32_t GetColumnFamilyID(ColumnFamilyHandle* column_family) {
  uint32_t column_family_id = 0;
  if (column_family != nullptr) {
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+    auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
    column_family_id = cfh->GetID();
  }
  return column_family_id;
@@ -387,6 +387,7 @@ class ColumnFamilyData {
  bool NeedsCompaction() const;
  // REQUIRES: DB mutex held
  Compaction* PickCompaction(const MutableCFOptions& mutable_options,
+                             const MutableDBOptions& mutable_db_options,
                             LogBuffer* log_buffer);

  // Check if the passed range overlap with any running compactions.
@@ -412,6 +413,7 @@ class ColumnFamilyData {
  static const int kCompactToBaseLevel;
  // REQUIRES: DB mutex held
  Compaction* CompactRange(const MutableCFOptions& mutable_cf_options,
+                           const MutableDBOptions& mutable_db_options,
                           int input_level, int output_level,
                           const CompactRangeOptions& compact_range_options,
                           const InternalKey* begin, const InternalKey* end,
@@ -22,26 +22,18 @@
 #include "rocksdb/env.h"
 #include "rocksdb/iterator.h"
 #include "rocksdb/utilities/object_registry.h"
-#include "test_util/fault_injection_test_env.h"
 #include "test_util/sync_point.h"
 #include "test_util/testharness.h"
 #include "test_util/testutil.h"
 #include "util/coding.h"
 #include "util/string_util.h"
+#include "utilities/fault_injection_env.h"
 #include "utilities/merge_operators.h"

 namespace ROCKSDB_NAMESPACE {

 static const int kValueSize = 1000;

-namespace {
-std::string RandomString(Random* rnd, int len) {
-  std::string r;
-  test::RandomString(rnd, len, &r);
-  return r;
-}
-}  // anonymous namespace
-
 // counts how many operations were performed
 class EnvCounter : public EnvWrapper {
 public:
@@ -109,11 +101,11 @@ class ColumnFamilyTestBase : public testing::Test {
      // preserves the implementation that was in place when all of the
      // magic values in this file were picked.
      *storage = std::string(kValueSize, ' ');
-      return Slice(*storage);
    } else {
      Random r(k);
-      return test::RandomString(&r, kValueSize, storage);
+      *storage = r.RandomString(kValueSize);
    }
+    return Slice(*storage);
  }

  void Build(int base, int n, int flush_every = 0) {
@@ -122,7 +114,7 @@ class ColumnFamilyTestBase : public testing::Test {

    for (int i = 0; i < n; i++) {
      if (flush_every != 0 && i != 0 && i % flush_every == 0) {
-        DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+        DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
        dbi->TEST_FlushMemTable();
      }

@@ -227,7 +219,7 @@ class ColumnFamilyTestBase : public testing::Test {
    Open({"default"});
  }

-  DBImpl* dbfull() { return reinterpret_cast<DBImpl*>(db_); }
+  DBImpl* dbfull() { return static_cast_with_check<DBImpl>(db_); }

  int GetProperty(int cf, std::string property) {
    std::string value;
@@ -329,11 +321,11 @@ class ColumnFamilyTestBase : public testing::Test {
      // 10 bytes for key, rest is value
      if (!save) {
        ASSERT_OK(Put(cf, test::RandomKey(&rnd_, 11),
-                      RandomString(&rnd_, key_value_size - 10)));
+                      rnd_.RandomString(key_value_size - 10)));
      } else {
        std::string key = test::RandomKey(&rnd_, 11);
        keys_[cf].insert(key);
-        ASSERT_OK(Put(cf, key, RandomString(&rnd_, key_value_size - 10)));
+        ASSERT_OK(Put(cf, key, rnd_.RandomString(key_value_size - 10)));
      }
    }
    db_->FlushWAL(false);
@@ -570,7 +562,7 @@ TEST_P(ColumnFamilyTest, DontReuseColumnFamilyID) {
    Open();
    CreateColumnFamilies({"one", "two", "three"});
    for (size_t i = 0; i < handles_.size(); ++i) {
-      auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(handles_[i]);
+      auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(handles_[i]);
      ASSERT_EQ(i, cfh->GetID());
    }
    if (iter == 1) {
@@ -586,7 +578,7 @@ TEST_P(ColumnFamilyTest, DontReuseColumnFamilyID) {
    CreateColumnFamilies({"three2"});
    // ID 3 that was used for dropped column family "three" should not be
    // reused
-    auto cfh3 = reinterpret_cast<ColumnFamilyHandleImpl*>(handles_[3]);
+    auto cfh3 = static_cast_with_check<ColumnFamilyHandleImpl>(handles_[3]);
    ASSERT_EQ(4U, cfh3->GetID());
    Close();
    Destroy();
@@ -2432,7 +2424,10 @@ TEST_P(ColumnFamilyTest, FlushAndDropRaceCondition) {

  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task,
                 Env::Priority::HIGH);
-
+  // Make sure the task is sleeping. Otherwise, it might start to execute
+  // after sleeping_task.WaitUntilDone() and cause TSAN warning.
+  sleeping_task.WaitUntilSleeping();
+  
  // 1MB should create ~10 files for each CF
  int kKeysNum = 10000;
  PutRandomData(1, kKeysNum, 100);
@@ -2446,6 +2441,9 @@ TEST_P(ColumnFamilyTest, FlushAndDropRaceCondition) {
  // now we sleep again. this is just so we're certain that flush job finished
  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task,
                 Env::Priority::HIGH);
+  // Make sure the task is sleeping. Otherwise, it might start to execute
+  // after sleeping_task.WaitUntilDone() and cause TSAN warning.
+  sleeping_task.WaitUntilSleeping();
  sleeping_task.WakeUp();
  sleeping_task.WaitUntilDone();

@@ -2995,6 +2993,9 @@ TEST_P(ColumnFamilyTest, FlushCloseWALFiles) {
  test::SleepingBackgroundTask sleeping_task;
  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task,
                 Env::Priority::HIGH);
+  // Make sure the task is sleeping. Otherwise, it might start to execute
+  // after sleeping_task.WaitUntilDone() and cause TSAN warning.
+  sleeping_task.WaitUntilSleeping();

  WriteOptions wo;
  wo.sync = true;
@@ -3040,6 +3041,9 @@ TEST_P(ColumnFamilyTest, IteratorCloseWALFile1) {
  test::SleepingBackgroundTask sleeping_task;
  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task,
                 Env::Priority::HIGH);
+  // Make sure the task is sleeping. Otherwise, it might start to execute
+  // after sleeping_task.WaitUntilDone() and cause TSAN warning.
+  sleeping_task.WaitUntilSleeping();

  WriteOptions wo;
  wo.sync = true;
@@ -3354,7 +3358,7 @@ TEST_P(ColumnFamilyTest, MultipleCFPathsTest) {

  // Re-open and verify the keys.
  Reopen({ColumnFamilyOptions(), cf_opt1, cf_opt2});
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  for (int cf = 1; cf != 3; ++cf) {
    ReadOptions read_options;
    read_options.readahead_size = 0;
@@ -16,6 +16,7 @@
 #include "rocksdb/env.h"
 #include "test_util/sync_point.h"
 #include "test_util/testharness.h"
+#include "util/cast_util.h"
 #include "util/string_util.h"

 namespace ROCKSDB_NAMESPACE {
@@ -148,7 +149,7 @@ TEST_F(CompactFilesTest, ObsoleteFiles) {

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

  // verify all compaction input files are deleted
  for (auto fname : l0_files) {
@@ -183,13 +184,13 @@ TEST_F(CompactFilesTest, NotCutOutputOnLevel0) {
  for (int i = 0; i < 500; ++i) {
    db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
  }
-  reinterpret_cast<DBImpl*>(db)->TEST_WaitForFlushMemTable();
+  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)));
  }
-  reinterpret_cast<DBImpl*>(db)->TEST_WaitForFlushMemTable();
+  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));
@@ -383,7 +384,7 @@ TEST_F(CompactFilesTest, GetCompactionJobInfo) {
  for (int i = 0; i < 500; ++i) {
    db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
  }
-  reinterpret_cast<DBImpl*>(db)->TEST_WaitForFlushMemTable();
+  static_cast_with_check<DBImpl>(db)->TEST_WaitForFlushMemTable();
  auto l0_files_1 = collector->GetFlushedFiles();
  CompactionOptions co;
  co.compression = CompressionType::kLZ4Compression;
@@ -5,9 +5,11 @@

 #ifndef ROCKSDB_LITE
 #include "db/compacted_db_impl.h"
+
 #include "db/db_impl/db_impl.h"
 #include "db/version_set.h"
 #include "table/get_context.h"
+#include "util/cast_util.h"

 namespace ROCKSDB_NAMESPACE {

@@ -90,8 +92,8 @@ Status CompactedDBImpl::Init(const Options& options) {
                            ColumnFamilyOptions(options));
  Status s = Recover({cf}, true /* read only */, false, true);
  if (s.ok()) {
-    cfd_ = reinterpret_cast<ColumnFamilyHandleImpl*>(
-              DefaultColumnFamily())->cfd();
+    cfd_ = static_cast_with_check<ColumnFamilyHandleImpl>(DefaultColumnFamily())
+               ->cfd();
    cfd_->InstallSuperVersion(&sv_context, &mutex_);
  }
  mutex_.Unlock();
@@ -7,12 +7,14 @@
 // 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/compaction/compaction.h"
+
 #include <cinttypes>
 #include <vector>

 #include "db/column_family.h"
-#include "db/compaction/compaction.h"
 #include "rocksdb/compaction_filter.h"
+#include "rocksdb/sst_partitioner.h"
 #include "test_util/sync_point.h"
 #include "util/string_util.h"

@@ -205,6 +207,7 @@ bool Compaction::IsFullCompaction(
 Compaction::Compaction(VersionStorageInfo* vstorage,
                       const ImmutableCFOptions& _immutable_cf_options,
                       const MutableCFOptions& _mutable_cf_options,
+                       const MutableDBOptions& _mutable_db_options,
                       std::vector<CompactionInputFiles> _inputs,
                       int _output_level, uint64_t _target_file_size,
                       uint64_t _max_compaction_bytes, uint32_t _output_path_id,
@@ -243,7 +246,7 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
    compaction_reason_ = CompactionReason::kManualCompaction;
  }
  if (max_subcompactions_ == 0) {
-    max_subcompactions_ = immutable_cf_options_.max_subcompactions;
+    max_subcompactions_ = _mutable_db_options.max_subcompactions;
  }
  if (!bottommost_level_) {
    // Currently we only enable dictionary compression during compaction to the
@@ -328,6 +331,8 @@ bool Compaction::IsTrivialMove() const {

  // assert inputs_.size() == 1

+  std::unique_ptr<SstPartitioner> partitioner = CreateSstPartitioner();
+
  for (const auto& file : inputs_.front().files) {
    std::vector<FileMetaData*> file_grand_parents;
    if (output_level_ + 1 >= number_levels_) {
@@ -340,6 +345,13 @@ bool Compaction::IsTrivialMove() const {
    if (compaction_size > max_compaction_bytes_) {
      return false;
    }
+
+    if (partitioner.get() != nullptr) {
+      if (!partitioner->CanDoTrivialMove(file->smallest.user_key(),
+                                         file->largest.user_key())) {
+        return false;
+      }
+    }
  }

  return true;
@@ -525,6 +537,21 @@ std::unique_ptr<CompactionFilter> Compaction::CreateCompactionFilter() const {
      context);
 }

+std::unique_ptr<SstPartitioner> Compaction::CreateSstPartitioner() const {
+  if (!immutable_cf_options_.sst_partitioner_factory) {
+    return nullptr;
+  }
+
+  SstPartitioner::Context context;
+  context.is_full_compaction = is_full_compaction_;
+  context.is_manual_compaction = is_manual_compaction_;
+  context.output_level = output_level_;
+  context.smallest_user_key = smallest_user_key_;
+  context.largest_user_key = largest_user_key_;
+  return immutable_cf_options_.sst_partitioner_factory->CreatePartitioner(
+      context);
+}
+
 bool Compaction::IsOutputLevelEmpty() const {
  return inputs_.back().level != output_level_ || inputs_.back().empty();
 }
@@ -11,6 +11,7 @@
 #include "db/version_set.h"
 #include "memory/arena.h"
 #include "options/cf_options.h"
+#include "rocksdb/sst_partitioner.h"
 #include "util/autovector.h"

 namespace ROCKSDB_NAMESPACE {
@@ -71,6 +72,7 @@ class Compaction {
  Compaction(VersionStorageInfo* input_version,
             const ImmutableCFOptions& immutable_cf_options,
             const MutableCFOptions& mutable_cf_options,
+             const MutableDBOptions& mutable_db_options,
             std::vector<CompactionInputFiles> inputs, int output_level,
             uint64_t target_file_size, uint64_t max_compaction_bytes,
             uint32_t output_path_id, CompressionType compression,
@@ -255,6 +257,9 @@ class Compaction {
  // Create a CompactionFilter from compaction_filter_factory
  std::unique_ptr<CompactionFilter> CreateCompactionFilter() const;

+  // Create a SstPartitioner from sst_partitioner_factory
+  std::unique_ptr<SstPartitioner> CreateSstPartitioner() const;
+
  // Is the input level corresponding to output_level_ empty?
  bool IsOutputLevelEmpty() const;

@@ -182,7 +182,7 @@ void CompactionIterator::Next() {
  PrepareOutput();
 }

-void CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
+bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
                                              Slice* skip_until) {
  if (compaction_filter_ != nullptr &&
      (ikey_.type == kTypeValue || ikey_.type == kTypeBlobIndex)) {
@@ -225,14 +225,31 @@ void CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
      value_.clear();
      iter_stats_.num_record_drop_user++;
    } else if (filter == CompactionFilter::Decision::kChangeValue) {
+      if (ikey_.type == kTypeBlobIndex) {
+        // value transfer from blob file to inlined data
+        ikey_.type = kTypeValue;
+        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
+      }
      value_ = compaction_filter_value_;
    } else if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil) {
      *need_skip = true;
      compaction_filter_skip_until_.ConvertFromUserKey(kMaxSequenceNumber,
                                                       kValueTypeForSeek);
      *skip_until = compaction_filter_skip_until_.Encode();
+    } else if (filter == CompactionFilter::Decision::kChangeBlobIndex) {
+      if (ikey_.type == kTypeValue) {
+        // value transfer from inlined data to blob file
+        ikey_.type = kTypeBlobIndex;
+        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
+      }
+      value_ = compaction_filter_value_;
+    } else if (filter == CompactionFilter::Decision::kIOError) {
+      status_ =
+          Status::IOError("Failed to access blob during compaction filter");
+      return false;
    }
  }
+  return true;
 }

 void CompactionIterator::NextFromInput() {
@@ -246,6 +263,7 @@ void CompactionIterator::NextFromInput() {
    iter_stats_.num_input_records++;

    if (!ParseInternalKey(key_, &ikey_)) {
+      iter_stats_.num_input_corrupt_records++;
      // If `expect_valid_internal_key_` is false, return the corrupted key
      // and let the caller decide what to do with it.
      // TODO(noetzli): We should have a more elegant solution for this.
@@ -258,7 +276,6 @@ void CompactionIterator::NextFromInput() {
      has_current_user_key_ = false;
      current_user_key_sequence_ = kMaxSequenceNumber;
      current_user_key_snapshot_ = 0;
-      iter_stats_.num_input_corrupt_records++;
      valid_ = true;
      break;
    }
@@ -295,8 +312,9 @@ void CompactionIterator::NextFromInput() {

      // Apply the compaction filter to the first committed version of the user
      // key.
-      if (current_key_committed_) {
-        InvokeFilterIfNeeded(&need_skip, &skip_until);
+      if (current_key_committed_ &&
+          !InvokeFilterIfNeeded(&need_skip, &skip_until)) {
+        break;
      }
    } else {
      // Update the current key to reflect the new sequence number/type without
@@ -316,8 +334,9 @@ void CompactionIterator::NextFromInput() {
        current_key_committed_ = KeyCommitted(ikey_.sequence);
        // Apply the compaction filter to the first committed version of the
        // user key.
-        if (current_key_committed_) {
-          InvokeFilterIfNeeded(&need_skip, &skip_until);
+        if (current_key_committed_ &&
+            !InvokeFilterIfNeeded(&need_skip, &skip_until)) {
+          break;
        }
      }
    }
@@ -121,7 +121,8 @@ class CompactionIterator {
  void PrepareOutput();

  // Invoke compaction filter if needed.
-  void InvokeFilterIfNeeded(bool* need_skip, Slice* skip_until);
+  // Return true on success, false on failures (e.g.: kIOError).
+  bool InvokeFilterIfNeeded(bool* need_skip, Slice* skip_until);

  // Given a sequence number, return the sequence number of the
  // earliest snapshot that this sequence number is visible in.
@@ -7,6 +7,8 @@
 // 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/compaction/compaction_job.h"
+
 #include <algorithm>
 #include <cinttypes>
 #include <functional>
@@ -19,7 +21,6 @@
 #include <vector>

 #include "db/builder.h"
-#include "db/compaction/compaction_job.h"
 #include "db/db_impl/db_impl.h"
 #include "db/db_iter.h"
 #include "db/dbformat.h"
@@ -45,6 +46,7 @@
 #include "port/port.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
+#include "rocksdb/sst_partitioner.h"
 #include "rocksdb/statistics.h"
 #include "rocksdb/status.h"
 #include "rocksdb/table.h"
@@ -54,6 +56,7 @@
 #include "table/table_builder.h"
 #include "test_util/sync_point.h"
 #include "util/coding.h"
+#include "util/hash.h"
 #include "util/mutexlock.h"
 #include "util/random.h"
 #include "util/stop_watch.h"
@@ -116,10 +119,14 @@ struct CompactionJob::SubcompactionState {
  // The return status of this subcompaction
  Status status;

+  // The return IO Status of this subcompaction
+  IOStatus io_status;
+
  // Files produced by this subcompaction
  struct Output {
    FileMetaData meta;
    bool finished;
+    uint64_t paranoid_hash;
    std::shared_ptr<const TableProperties> table_properties;
  };

@@ -129,7 +136,7 @@ struct CompactionJob::SubcompactionState {
  std::unique_ptr<TableBuilder> builder;
  Output* current_output() {
    if (outputs.empty()) {
-      // This subcompaction's outptut could be empty if compaction was aborted
+      // This subcompaction's output could be empty if compaction was aborted
      // before this subcompaction had a chance to generate any output files.
      // When subcompactions are executed sequentially this is more likely and
      // will be particulalry likely for the later subcompactions to be empty.
@@ -179,6 +186,7 @@ struct CompactionJob::SubcompactionState {
    start = std::move(o.start);
    end = std::move(o.end);
    status = std::move(o.status);
+    io_status = std::move(o.io_status);
    outputs = std::move(o.outputs);
    outfile = std::move(o.outfile);
    builder = std::move(o.builder);
@@ -198,6 +206,21 @@ struct CompactionJob::SubcompactionState {

  SubcompactionState& operator=(const SubcompactionState&) = delete;

+  // Adds the key and value to the builder
+  // If paranoid is true, adds the key-value to the paranoid hash
+  void AddToBuilder(const Slice& key, const Slice& value, bool paranoid) {
+    auto curr = current_output();
+    assert(builder != nullptr);
+    assert(curr != nullptr);
+    if (paranoid) {
+      // Generate a rolling 64-bit hash of the key and values
+      curr->paranoid_hash = Hash64(key.data(), key.size(), curr->paranoid_hash);
+      curr->paranoid_hash =
+          Hash64(value.data(), value.size(), curr->paranoid_hash);
+    }
+    builder->Add(key, value);
+  }
+
  // Returns true iff we should stop building the current output
  // before processing "internal_key".
  bool ShouldStopBefore(const Slice& internal_key, uint64_t curr_file_size) {
@@ -305,12 +328,15 @@ CompactionJob::CompactionJob(
    const SnapshotChecker* snapshot_checker, std::shared_ptr<Cache> table_cache,
    EventLogger* event_logger, bool paranoid_file_checks, bool measure_io_stats,
    const std::string& dbname, CompactionJobStats* compaction_job_stats,
-    Env::Priority thread_pri, const std::atomic<bool>* manual_compaction_paused)
+    Env::Priority thread_pri, const std::atomic<bool>* manual_compaction_paused,
+    const std::string& db_id, const std::string& db_session_id)
    : job_id_(job_id),
      compact_(new CompactionState(compaction)),
      compaction_job_stats_(compaction_job_stats),
      compaction_stats_(compaction->compaction_reason(), 1),
      dbname_(dbname),
+      db_id_(db_id),
+      db_session_id_(db_session_id),
      db_options_(db_options),
      file_options_(file_options),
      env_(db_options.env),
@@ -542,7 +568,7 @@ void CompactionJob::GenSubcompactionBoundaries() {
    // Greedily add ranges to the subcompaction until the sum of the ranges'
    // sizes becomes >= the expected mean size of a subcompaction
    sum = 0;
-    for (size_t i = 0; i < ranges.size() - 1; i++) {
+    for (size_t i = 0; i + 1 < ranges.size(); i++) {
      sum += ranges[i].size;
      if (subcompactions == 1) {
        // If there's only one left to schedule then it goes to the end so no
@@ -606,38 +632,42 @@ Status CompactionJob::Run() {

  // Check if any thread encountered an error during execution
  Status status;
+  IOStatus io_s;
  for (const auto& state : compact_->sub_compact_states) {
    if (!state.status.ok()) {
      status = state.status;
+      io_s = state.io_status;
      break;
    }
  }
-
-  IOStatus io_s;
+  if (io_status_.ok()) {
+    io_status_ = io_s;
+  }
  if (status.ok() && output_directory_) {
    io_s = output_directory_->Fsync(IOOptions(), nullptr);
  }
-  if (!io_s.ok()) {
+  if (io_status_.ok()) {
    io_status_ = io_s;
+  }
+  if (status.ok()) {
    status = io_s;
  }
-
  if (status.ok()) {
    thread_pool.clear();
-    std::vector<const FileMetaData*> files_meta;
+    std::vector<const CompactionJob::SubcompactionState::Output*> files_output;
    for (const auto& state : compact_->sub_compact_states) {
      for (const auto& output : state.outputs) {
-        files_meta.emplace_back(&output.meta);
+        files_output.emplace_back(&output);
      }
    }
    ColumnFamilyData* cfd = compact_->compaction->column_family_data();
    auto prefix_extractor =
        compact_->compaction->mutable_cf_options()->prefix_extractor.get();
-    std::atomic<size_t> next_file_meta_idx(0);
+    std::atomic<size_t> next_file_idx(0);
    auto verify_table = [&](Status& output_status) {
      while (true) {
-        size_t file_idx = next_file_meta_idx.fetch_add(1);
-        if (file_idx >= files_meta.size()) {
+        size_t file_idx = next_file_idx.fetch_add(1);
+        if (file_idx >= files_output.size()) {
          break;
        }
        // Verify that the table is usable
@@ -648,20 +678,31 @@ Status CompactionJob::Run() {
        // to cache it here for further user reads
        InternalIterator* iter = cfd->table_cache()->NewIterator(
            ReadOptions(), file_options_, cfd->internal_comparator(),
-            *files_meta[file_idx], /*range_del_agg=*/nullptr, prefix_extractor,
+            files_output[file_idx]->meta, /*range_del_agg=*/nullptr,
+            prefix_extractor,
            /*table_reader_ptr=*/nullptr,
            cfd->internal_stats()->GetFileReadHist(
                compact_->compaction->output_level()),
            TableReaderCaller::kCompactionRefill, /*arena=*/nullptr,
            /*skip_filters=*/false, compact_->compaction->output_level(),
+            MaxFileSizeForL0MetaPin(
+                *compact_->compaction->mutable_cf_options()),
            /*smallest_compaction_key=*/nullptr,
            /*largest_compaction_key=*/nullptr,
            /*allow_unprepared_value=*/false);
        auto s = iter->status();

        if (s.ok() && paranoid_file_checks_) {
-          for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {}
+          uint64_t hash = 0;
+          for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
+            // Generate a rolling 64-bit hash of the key and values, using the
+            hash = Hash64(iter->key().data(), iter->key().size(), hash);
+            hash = Hash64(iter->value().data(), iter->value().size(), hash);
+          }
          s = iter->status();
+          if (s.ok() && hash != files_output[file_idx]->paranoid_hash) {
+            s = Status::Corruption("Paraniod checksums do not match");
+          }
        }

        delete iter;
@@ -719,7 +760,6 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
  cfd->internal_stats()->AddCompactionStats(
      compact_->compaction->output_level(), thread_pri_, compaction_stats_);

-  versions_->SetIOStatusOK();
  if (status.ok()) {
    status = InstallCompactionResults(mutable_cf_options);
  }
@@ -894,9 +934,10 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  sub_compact->c_iter.reset(new CompactionIterator(
      input.get(), cfd->user_comparator(), &merge, versions_->LastSequence(),
      &existing_snapshots_, earliest_write_conflict_snapshot_,
-      snapshot_checker_, env_, ShouldReportDetailedTime(env_, stats_), false,
-      &range_del_agg, sub_compact->compaction, compaction_filter,
-      shutting_down_, preserve_deletes_seqnum_, manual_compaction_paused_,
+      snapshot_checker_, env_, ShouldReportDetailedTime(env_, stats_),
+      /*expect_valid_internal_key=*/true, &range_del_agg,
+      sub_compact->compaction, compaction_filter, shutting_down_,
+      preserve_deletes_seqnum_, manual_compaction_paused_,
      db_options_.info_log));
  auto c_iter = sub_compact->c_iter.get();
  c_iter->SeekToFirst();
@@ -909,6 +950,12 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  }
  const auto& c_iter_stats = c_iter->iter_stats();

+  std::unique_ptr<SstPartitioner> partitioner =
+      sub_compact->compaction->output_level() == 0
+          ? nullptr
+          : sub_compact->compaction->CreateSstPartitioner();
+  std::string last_key_for_partitioner;
+
  while (status.ok() && !cfd->IsDropped() && c_iter->Valid()) {
    // Invariant: c_iter.status() is guaranteed to be OK if c_iter->Valid()
    // returns true.
@@ -935,9 +982,8 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
        break;
      }
    }
-    assert(sub_compact->builder != nullptr);
-    assert(sub_compact->current_output() != nullptr);
-    sub_compact->builder->Add(key, value);
+    sub_compact->AddToBuilder(key, value, paranoid_file_checks_);
+
    sub_compact->current_output_file_size =
        sub_compact->builder->EstimatedFileSize();
    const ParsedInternalKey& ikey = c_iter->ikey();
@@ -967,20 +1013,29 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
        "CompactionJob::Run():PausingManualCompaction:2",
        reinterpret_cast<void*>(
            const_cast<std::atomic<bool>*>(manual_compaction_paused_)));
+    if (partitioner.get()) {
+      last_key_for_partitioner.assign(c_iter->user_key().data_,
+                                      c_iter->user_key().size_);
+    }
    c_iter->Next();
    if (c_iter->status().IsManualCompactionPaused()) {
      break;
    }
-    if (!output_file_ended && c_iter->Valid() &&
-        sub_compact->compaction->output_level() != 0 &&
-        sub_compact->ShouldStopBefore(c_iter->key(),
-                                      sub_compact->current_output_file_size) &&
-        sub_compact->builder != nullptr) {
-      // (2) this key belongs to the next file. For historical reasons, the
-      // iterator status after advancing will be given to
-      // FinishCompactionOutputFile().
-      input_status = input->status();
-      output_file_ended = true;
+    if (!output_file_ended && c_iter->Valid()) {
+      if (((partitioner.get() &&
+            partitioner->ShouldPartition(PartitionerRequest(
+                last_key_for_partitioner, c_iter->user_key(),
+                sub_compact->current_output_file_size)) == kRequired) ||
+           (sub_compact->compaction->output_level() != 0 &&
+            sub_compact->ShouldStopBefore(
+                c_iter->key(), sub_compact->current_output_file_size))) &&
+          sub_compact->builder != nullptr) {
+        // (2) this key belongs to the next file. For historical reasons, the
+        // iterator status after advancing will be given to
+        // FinishCompactionOutputFile().
+        input_status = input->status();
+        output_file_ended = true;
+      }
    }
    if (output_file_ended) {
      const Slice* next_key = nullptr;
@@ -1205,6 +1260,7 @@ Status CompactionJob::FinishCompactionOutputFile(
    } else {
      it->SeekToFirst();
    }
+    TEST_SYNC_POINT("CompactionJob::FinishCompactionOutputFile1");
    for (; it->Valid(); it->Next()) {
      auto tombstone = it->Tombstone();
      if (upper_bound != nullptr) {
@@ -1232,6 +1288,7 @@ Status CompactionJob::FinishCompactionOutputFile(
      auto kv = tombstone.Serialize();
      assert(lower_bound == nullptr ||
             ucmp->Compare(*lower_bound, kv.second) < 0);
+      // Range tombstone is not supported by output validator yet.
      sub_compact->builder->Add(kv.first.Encode(), kv.second);
      InternalKey smallest_candidate = std::move(kv.first);
      if (lower_bound != nullptr &&
@@ -1305,9 +1362,9 @@ Status CompactionJob::FinishCompactionOutputFile(
  } else {
    sub_compact->builder->Abandon();
  }
-  if (!sub_compact->builder->io_status().ok()) {
-    io_status_ = sub_compact->builder->io_status();
-    s = io_status_;
+  IOStatus io_s = sub_compact->builder->io_status();
+  if (s.ok()) {
+    s = io_s;
  }
  const uint64_t current_bytes = sub_compact->builder->FileSize();
  if (s.ok()) {
@@ -1317,24 +1374,25 @@ 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);
  }
-  if (io_s.ok()) {
+  if (s.ok() && io_s.ok()) {
    io_s = sub_compact->outfile->Close();
  }
-  if (io_s.ok()) {
+  if (s.ok() && io_s.ok()) {
    // Add the checksum information to file metadata.
    meta->file_checksum = sub_compact->outfile->GetFileChecksum();
    meta->file_checksum_func_name =
        sub_compact->outfile->GetFileChecksumFuncName();
  }
-  if (!io_s.ok()) {
-    io_status_ = io_s;
+  if (s.ok()) {
    s = io_s;
  }
+  if (sub_compact->io_status.ok()) {
+    sub_compact->io_status = io_s;
+  }
  sub_compact->outfile.reset();

  TableProperties tp;
@@ -1483,7 +1541,12 @@ Status CompactionJob::OpenCompactionOutputFile(
  TEST_SYNC_POINT_CALLBACK("CompactionJob::OpenCompactionOutputFile",
                           &syncpoint_arg);
 #endif
-  Status s = NewWritableFile(fs_, fname, &writable_file, file_options_);
+  Status s;
+  IOStatus io_s = NewWritableFile(fs_, fname, &writable_file, file_options_);
+  s = io_s;
+  if (sub_compact->io_status.ok()) {
+    sub_compact->io_status = io_s;
+  }
  if (!s.ok()) {
    ROCKS_LOG_ERROR(
        db_options_.info_log,
@@ -1523,6 +1586,7 @@ Status CompactionJob::OpenCompactionOutputFile(
    out.meta.oldest_ancester_time = oldest_ancester_time;
    out.meta.file_creation_time = current_time;
    out.finished = false;
+    out.paranoid_hash = 0;
    sub_compact->outputs.push_back(out);
  }

@@ -1552,7 +1616,8 @@ Status CompactionJob::OpenCompactionOutputFile(
      sub_compact->compaction->output_compression_opts(),
      sub_compact->compaction->output_level(), skip_filters,
      oldest_ancester_time, 0 /* oldest_key_time */,
-      sub_compact->compaction->max_output_file_size(), current_time));
+      sub_compact->compaction->max_output_file_size(), current_time, db_id_,
+      db_session_id_));
  LogFlush(db_options_.info_log);
  return s;
 }
@@ -78,7 +78,9 @@ class CompactionJob {
                const std::string& dbname,
                CompactionJobStats* compaction_job_stats,
                Env::Priority thread_pri,
-                const std::atomic<bool>* manual_compaction_paused = nullptr);
+                const std::atomic<bool>* manual_compaction_paused = nullptr,
+                const std::string& db_id = "",
+                const std::string& db_session_id = "");

  ~CompactionJob();

@@ -152,6 +154,8 @@ class CompactionJob {

  // DBImpl state
  const std::string& dbname_;
+  const std::string db_id_;
+  const std::string db_session_id_;
  const ImmutableDBOptions& db_options_;
  const FileOptions file_options_;

@@ -52,6 +52,7 @@
 #include "test_util/sync_point.h"
 #include "test_util/testharness.h"
 #include "test_util/testutil.h"
+#include "util/cast_util.h"
 #include "util/compression.h"
 #include "util/hash.h"
 #include "util/mutexlock.h"
@@ -126,9 +127,7 @@ class CompactionJobStatsTest : public testing::Test,
  static void SetUpTestCase() {}
  static void TearDownTestCase() {}

-  DBImpl* dbfull() {
-    return reinterpret_cast<DBImpl*>(db_);
-  }
+  DBImpl* dbfull() { return static_cast_with_check<DBImpl>(db_); }

  void CreateColumnFamilies(const std::vector<std::string>& cfs,
                            const Options& options) {
@@ -797,7 +796,7 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {
    }

    ASSERT_OK(Flush(1));
-    reinterpret_cast<DBImpl*>(db_)->TEST_WaitForCompact();
+    static_cast_with_check<DBImpl>(db_)->TEST_WaitForCompact();

    stats_checker->set_verify_next_comp_io_stats(true);
    std::atomic<bool> first_prepare_write(true);
@@ -1012,7 +1011,7 @@ TEST_P(CompactionJobStatsTest, UniversalCompactionTest) {
        &rnd, start_key, start_key + key_base - 1,
        kKeySize, kValueSize, key_interval,
        compression_ratio, 1);
-    reinterpret_cast<DBImpl*>(db_)->TEST_WaitForCompact();
+    static_cast_with_check<DBImpl>(db_)->TEST_WaitForCompact();
  }
  ASSERT_EQ(stats_checker->NumberOfUnverifiedStats(), 0U);
 }
@@ -76,6 +76,7 @@ class CompactionJobTest : public testing::Test {
        dbname_(test::PerThreadDBPath("compaction_job_test")),
        db_options_(),
        mutable_cf_options_(cf_options_),
+        mutable_db_options_(),
        table_cache_(NewLRUCache(50000, 16)),
        write_buffer_manager_(db_options_.db_write_buffer_size),
        versions_(new VersionSet(dbname_, &db_options_, env_options_,
@@ -316,9 +317,10 @@ class CompactionJobTest : public testing::Test {

    Compaction compaction(
        cfd->current()->storage_info(), *cfd->ioptions(),
-        *cfd->GetLatestMutableCFOptions(), compaction_input_files, output_level,
-        1024 * 1024, 10 * 1024 * 1024, 0, kNoCompression,
-        cfd->GetLatestMutableCFOptions()->compression_opts, 0, {}, true);
+        *cfd->GetLatestMutableCFOptions(), mutable_db_options_,
+        compaction_input_files, output_level, 1024 * 1024, 10 * 1024 * 1024, 0,
+        kNoCompression, cfd->GetLatestMutableCFOptions()->compression_opts, 0,
+        {}, true);
    compaction.SetInputVersion(cfd->current());

    LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, db_options_.info_log.get());
@@ -370,6 +372,7 @@ class CompactionJobTest : public testing::Test {
  ImmutableDBOptions db_options_;
  ColumnFamilyOptions cf_options_;
  MutableCFOptions mutable_cf_options_;
+  MutableDBOptions mutable_db_options_;
  std::shared_ptr<Cache> table_cache_;
  WriteController write_controller_;
  WriteBufferManager write_buffer_manager_;
@@ -395,7 +398,7 @@ TEST_F(CompactionJobTest, Simple) {
  RunCompaction({ files }, expected_results);
 }

-TEST_F(CompactionJobTest, SimpleCorrupted) {
+TEST_F(CompactionJobTest, DISABLED_SimpleCorrupted) {
  NewDB();

  auto expected_results = CreateTwoFiles(true);
@@ -989,7 +992,7 @@ TEST_F(CompactionJobTest, MultiSingleDelete) {
 // single deletion and the (single) deletion gets removed while the corrupt key
 // gets written out. TODO(noetzli): We probably want a better way to treat
 // corrupt keys.
-TEST_F(CompactionJobTest, CorruptionAfterDeletion) {
+TEST_F(CompactionJobTest, DISABLED_CorruptionAfterDeletion) {
  NewDB();

  auto file1 =
@@ -332,7 +332,7 @@ Compaction* CompactionPicker::CompactFiles(
    const CompactionOptions& compact_options,
    const std::vector<CompactionInputFiles>& input_files, int output_level,
    VersionStorageInfo* vstorage, const MutableCFOptions& mutable_cf_options,
-    uint32_t output_path_id) {
+    const MutableDBOptions& mutable_db_options, uint32_t output_path_id) {
  assert(input_files.size());
  // This compaction output should not overlap with a running compaction as
  // `SanitizeCompactionInputFiles` should've checked earlier and db mutex
@@ -356,8 +356,8 @@ Compaction* CompactionPicker::CompactFiles(
    compression_type = compact_options.compression;
  }
  auto c = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, input_files, output_level,
-      compact_options.output_file_size_limit,
+      vstorage, ioptions_, mutable_cf_options, mutable_db_options, input_files,
+      output_level, compact_options.output_file_size_limit,
      mutable_cf_options.max_compaction_bytes, output_path_id, compression_type,
      GetCompressionOptions(mutable_cf_options, vstorage, output_level),
      compact_options.max_subcompactions,
@@ -563,7 +563,8 @@ void CompactionPicker::GetGrandparents(

 Compaction* CompactionPicker::CompactRange(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, int input_level, int output_level,
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    int input_level, int output_level,
    const CompactRangeOptions& compact_range_options, const InternalKey* begin,
    const InternalKey* end, InternalKey** compaction_end, bool* manual_conflict,
    uint64_t max_file_num_to_ignore) {
@@ -626,8 +627,8 @@ Compaction* CompactionPicker::CompactRange(
    }

    Compaction* c = new Compaction(
-        vstorage, ioptions_, mutable_cf_options, std::move(inputs),
-        output_level,
+        vstorage, ioptions_, mutable_cf_options, mutable_db_options,
+        std::move(inputs), output_level,
        MaxFileSizeForLevel(mutable_cf_options, output_level,
                            ioptions_.compaction_style),
        /* max_compaction_bytes */ LLONG_MAX,
@@ -778,8 +779,8 @@ Compaction* CompactionPicker::CompactRange(
  std::vector<FileMetaData*> grandparents;
  GetGrandparents(vstorage, inputs, output_level_inputs, &grandparents);
  Compaction* compaction = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, std::move(compaction_inputs),
-      output_level,
+      vstorage, ioptions_, mutable_cf_options, mutable_db_options,
+      std::move(compaction_inputs), output_level,
      MaxFileSizeForLevel(mutable_cf_options, output_level,
                          ioptions_.compaction_style, vstorage->base_level(),
                          ioptions_.level_compaction_dynamic_level_bytes),
@@ -1085,6 +1086,8 @@ void CompactionPicker::PickFilesMarkedForCompaction(
  Random64 rnd(/* seed */ reinterpret_cast<uint64_t>(vstorage));
  size_t random_file_index = static_cast<size_t>(rnd.Uniform(
      static_cast<uint64_t>(vstorage->FilesMarkedForCompaction().size())));
+  TEST_SYNC_POINT_CALLBACK("CompactionPicker::PickFilesMarkedForCompaction",
+                           &random_file_index);

  if (continuation(vstorage->FilesMarkedForCompaction()[random_file_index])) {
    // found the compaction!
@@ -56,7 +56,8 @@ class CompactionPicker {
  // describes the compaction.  Caller should delete the result.
  virtual Compaction* PickCompaction(
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* vstorage, LogBuffer* log_buffer,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+      LogBuffer* log_buffer,
      SequenceNumber earliest_memtable_seqno = kMaxSequenceNumber) = 0;

  // Return a compaction object for compacting the range [begin,end] in
@@ -72,7 +73,8 @@ class CompactionPicker {
  // *compaction_end should point to valid InternalKey!
  virtual Compaction* CompactRange(
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* vstorage, int input_level, int output_level,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+      int input_level, int output_level,
      const CompactRangeOptions& compact_range_options,
      const InternalKey* begin, const InternalKey* end,
      InternalKey** compaction_end, bool* manual_conflict,
@@ -113,6 +115,7 @@ class CompactionPicker {
                           const std::vector<CompactionInputFiles>& input_files,
                           int output_level, VersionStorageInfo* vstorage,
                           const MutableCFOptions& mutable_cf_options,
+                           const MutableDBOptions& mutable_db_options,
                           uint32_t output_path_id);

  // Converts a set of compaction input file numbers into
@@ -250,6 +253,7 @@ class NullCompactionPicker : public CompactionPicker {
  Compaction* PickCompaction(
      const std::string& /*cf_name*/,
      const MutableCFOptions& /*mutable_cf_options*/,
+      const MutableDBOptions& /*mutable_db_options*/,
      VersionStorageInfo* /*vstorage*/, LogBuffer* /* log_buffer */,
      SequenceNumber /* earliest_memtable_seqno */) override {
    return nullptr;
@@ -258,6 +262,7 @@ class NullCompactionPicker : public CompactionPicker {
  // Always return "nullptr"
  Compaction* CompactRange(const std::string& /*cf_name*/,
                           const MutableCFOptions& /*mutable_cf_options*/,
+                           const MutableDBOptions& /*mutable_db_options*/,
                           VersionStorageInfo* /*vstorage*/,
                           int /*input_level*/, int /*output_level*/,
                           const CompactRangeOptions& /*compact_range_options*/,
@@ -36,7 +36,8 @@ bool FIFOCompactionPicker::NeedsCompaction(

 Compaction* FIFOCompactionPicker::PickTTLCompaction(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    LogBuffer* log_buffer) {
  assert(mutable_cf_options.ttl > 0);

  const int kLevel0 = 0;
@@ -108,8 +109,9 @@ Compaction* FIFOCompactionPicker::PickTTLCompaction(
  }

  Compaction* c = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
-      kNoCompression, mutable_cf_options.compression_opts,
+      vstorage, ioptions_, mutable_cf_options, mutable_db_options,
+      std::move(inputs), 0, 0, 0, 0, kNoCompression,
+      mutable_cf_options.compression_opts,
      /* max_subcompactions */ 0, {}, /* is manual */ false,
      vstorage->CompactionScore(0),
      /* is deletion compaction */ true, CompactionReason::kFIFOTtl);
@@ -118,7 +120,8 @@ Compaction* FIFOCompactionPicker::PickTTLCompaction(

 Compaction* FIFOCompactionPicker::PickSizeCompaction(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    LogBuffer* log_buffer) {
  const int kLevel0 = 0;
  const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
  uint64_t total_size = GetTotalFilesSize(level_files);
@@ -147,8 +150,8 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(
              max_compact_bytes_per_del_file,
              mutable_cf_options.max_compaction_bytes, &comp_inputs)) {
        Compaction* c = new Compaction(
-            vstorage, ioptions_, mutable_cf_options, {comp_inputs}, 0,
-            16 * 1024 * 1024 /* output file size limit */,
+            vstorage, ioptions_, mutable_cf_options, mutable_db_options,
+            {comp_inputs}, 0, 16 * 1024 * 1024 /* output file size limit */,
            0 /* max compaction bytes, not applicable */,
            0 /* output path ID */, mutable_cf_options.compression,
            mutable_cf_options.compression_opts, 0 /* max_subcompactions */, {},
@@ -198,8 +201,9 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(
  }

  Compaction* c = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
-      kNoCompression, mutable_cf_options.compression_opts,
+      vstorage, ioptions_, mutable_cf_options, mutable_db_options,
+      std::move(inputs), 0, 0, 0, 0, kNoCompression,
+      mutable_cf_options.compression_opts,
      /* max_subcompactions */ 0, {}, /* is manual */ false,
      vstorage->CompactionScore(0),
      /* is deletion compaction */ true, CompactionReason::kFIFOMaxSize);
@@ -208,16 +212,18 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(

 Compaction* FIFOCompactionPicker::PickCompaction(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer,
-    SequenceNumber /*earliest_memtable_seqno*/) {
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    LogBuffer* log_buffer, SequenceNumber /*earliest_memtable_seqno*/) {
  assert(vstorage->num_levels() == 1);

  Compaction* c = nullptr;
  if (mutable_cf_options.ttl > 0) {
-    c = PickTTLCompaction(cf_name, mutable_cf_options, vstorage, log_buffer);
+    c = PickTTLCompaction(cf_name, mutable_cf_options, mutable_db_options,
+                          vstorage, log_buffer);
  }
  if (c == nullptr) {
-    c = PickSizeCompaction(cf_name, mutable_cf_options, vstorage, log_buffer);
+    c = PickSizeCompaction(cf_name, mutable_cf_options, mutable_db_options,
+                           vstorage, log_buffer);
  }
  RegisterCompaction(c);
  return c;
@@ -225,7 +231,8 @@ Compaction* FIFOCompactionPicker::PickCompaction(

 Compaction* FIFOCompactionPicker::CompactRange(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, int input_level, int output_level,
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    int input_level, int output_level,
    const CompactRangeOptions& /*compact_range_options*/,
    const InternalKey* /*begin*/, const InternalKey* /*end*/,
    InternalKey** compaction_end, bool* /*manual_conflict*/,
@@ -238,8 +245,8 @@ Compaction* FIFOCompactionPicker::CompactRange(
  assert(output_level == 0);
  *compaction_end = nullptr;
  LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, ioptions_.info_log);
-  Compaction* c =
-      PickCompaction(cf_name, mutable_cf_options, vstorage, &log_buffer);
+  Compaction* c = PickCompaction(cf_name, mutable_cf_options,
+                                 mutable_db_options, vstorage, &log_buffer);
  log_buffer.FlushBufferToLog();
  return c;
 }
@@ -21,12 +21,14 @@ class FIFOCompactionPicker : public CompactionPicker {

  virtual Compaction* PickCompaction(
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* version, LogBuffer* log_buffer,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* version,
+      LogBuffer* log_buffer,
      SequenceNumber earliest_memtable_seqno = kMaxSequenceNumber) override;

  virtual Compaction* CompactRange(
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* vstorage, int input_level, int output_level,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+      int input_level, int output_level,
      const CompactRangeOptions& compact_range_options,
      const InternalKey* begin, const InternalKey* end,
      InternalKey** compaction_end, bool* manual_conflict,
@@ -41,11 +43,13 @@ class FIFOCompactionPicker : public CompactionPicker {
 private:
  Compaction* PickTTLCompaction(const std::string& cf_name,
                                const MutableCFOptions& mutable_cf_options,
+                                const MutableDBOptions& mutable_db_options,
                                VersionStorageInfo* version,
                                LogBuffer* log_buffer);

  Compaction* PickSizeCompaction(const std::string& cf_name,
                                 const MutableCFOptions& mutable_cf_options,
+                                 const MutableDBOptions& mutable_db_options,
                                 VersionStorageInfo* version,
                                 LogBuffer* log_buffer);
 };
@@ -49,14 +49,16 @@ class LevelCompactionBuilder {
                         CompactionPicker* compaction_picker,
                         LogBuffer* log_buffer,
                         const MutableCFOptions& mutable_cf_options,
-                         const ImmutableCFOptions& ioptions)
+                         const ImmutableCFOptions& ioptions,
+                         const MutableDBOptions& mutable_db_options)
      : cf_name_(cf_name),
        vstorage_(vstorage),
        earliest_mem_seqno_(earliest_mem_seqno),
        compaction_picker_(compaction_picker),
        log_buffer_(log_buffer),
        mutable_cf_options_(mutable_cf_options),
-        ioptions_(ioptions) {}
+        ioptions_(ioptions),
+        mutable_db_options_(mutable_db_options) {}

  // Pick and return a compaction.
  Compaction* PickCompaction();
@@ -120,6 +122,7 @@ class LevelCompactionBuilder {

  const MutableCFOptions& mutable_cf_options_;
  const ImmutableCFOptions& ioptions_;
+  const MutableDBOptions& mutable_db_options_;
  // Pick a path ID to place a newly generated file, with its level
  static uint32_t GetPathId(const ImmutableCFOptions& ioptions,
                            const MutableCFOptions& mutable_cf_options,
@@ -323,8 +326,8 @@ Compaction* LevelCompactionBuilder::PickCompaction() {

 Compaction* LevelCompactionBuilder::GetCompaction() {
  auto c = new Compaction(
-      vstorage_, ioptions_, mutable_cf_options_, std::move(compaction_inputs_),
-      output_level_,
+      vstorage_, ioptions_, mutable_cf_options_, mutable_db_options_,
+      std::move(compaction_inputs_), output_level_,
      MaxFileSizeForLevel(mutable_cf_options_, output_level_,
                          ioptions_.compaction_style, vstorage_->base_level(),
                          ioptions_.level_compaction_dynamic_level_bytes),
@@ -497,10 +500,11 @@ bool LevelCompactionBuilder::PickIntraL0Compaction() {

 Compaction* LevelCompactionPicker::PickCompaction(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer,
-    SequenceNumber earliest_mem_seqno) {
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    LogBuffer* log_buffer, SequenceNumber earliest_mem_seqno) {
  LevelCompactionBuilder builder(cf_name, vstorage, earliest_mem_seqno, this,
-                                 log_buffer, mutable_cf_options, ioptions_);
+                                 log_buffer, mutable_cf_options, ioptions_,
+                                 mutable_db_options);
  return builder.PickCompaction();
 }
 }  // namespace ROCKSDB_NAMESPACE
@@ -22,7 +22,8 @@ class LevelCompactionPicker : public CompactionPicker {
      : CompactionPicker(ioptions, icmp) {}
  virtual Compaction* PickCompaction(
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* vstorage, LogBuffer* log_buffer,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+      LogBuffer* log_buffer,
      SequenceNumber earliest_memtable_seqno = kMaxSequenceNumber) override;

  virtual bool NeedsCompaction(
@@ -33,6 +33,7 @@ class CompactionPickerTest : public testing::Test {
  Options options_;
  ImmutableCFOptions ioptions_;
  MutableCFOptions mutable_cf_options_;
+  MutableDBOptions mutable_db_options_;
  LevelCompactionPicker level_compaction_picker;
  std::string cf_name_;
  CountingLogger logger_;
@@ -52,6 +53,7 @@ class CompactionPickerTest : public testing::Test {
        icmp_(ucmp_),
        ioptions_(options_),
        mutable_cf_options_(options_),
+        mutable_db_options_(),
        level_compaction_picker(ioptions_, &icmp_),
        cf_name_("dummy"),
        log_buffer_(InfoLogLevel::INFO_LEVEL, &logger_),
@@ -78,8 +80,17 @@ class CompactionPickerTest : public testing::Test {
    vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
  }

+  // Create a new VersionStorageInfo object so we can add mode files and then
+  // merge it with the existing VersionStorageInfo
+  void AddVersionStorage() {
+    temp_vstorage_.reset(new VersionStorageInfo(
+        &icmp_, ucmp_, options_.num_levels, ioptions_.compaction_style,
+        vstorage_.get(), false));
+  }
+
  void DeleteVersionStorage() {
    vstorage_.reset();
+    temp_vstorage_.reset();
    files_.clear();
    file_map_.clear();
    input_files_.clear();
@@ -88,18 +99,24 @@ class CompactionPickerTest : public testing::Test {
  void Add(int level, uint32_t file_number, const char* smallest,
           const char* largest, uint64_t file_size = 1, uint32_t path_id = 0,
           SequenceNumber smallest_seq = 100, SequenceNumber largest_seq = 100,
-           size_t compensated_file_size = 0) {
-    assert(level < vstorage_->num_levels());
+           size_t compensated_file_size = 0, bool marked_for_compact = false) {
+    VersionStorageInfo* vstorage;
+    if (temp_vstorage_) {
+      vstorage = temp_vstorage_.get();
+    } else {
+      vstorage = vstorage_.get();
+    }
+    assert(level < vstorage->num_levels());
    FileMetaData* f = new FileMetaData(
        file_number, path_id, file_size,
        InternalKey(smallest, smallest_seq, kTypeValue),
        InternalKey(largest, largest_seq, kTypeValue), smallest_seq,
-        largest_seq, /* marked_for_compact */ false, kInvalidBlobFileNumber,
+        largest_seq, marked_for_compact, kInvalidBlobFileNumber,
        kUnknownOldestAncesterTime, kUnknownFileCreationTime,
        kUnknownFileChecksum, kUnknownFileChecksumFuncName);
    f->compensated_file_size =
        (compensated_file_size != 0) ? compensated_file_size : file_size;
-    vstorage_->AddFile(level, f);
+    vstorage->AddFile(level, f);
    files_.emplace_back(f);
    file_map_.insert({file_number, {f, level}});
  }
@@ -122,6 +139,12 @@ class CompactionPickerTest : public testing::Test {
  }

  void UpdateVersionStorageInfo() {
+    if (temp_vstorage_) {
+      VersionBuilder builder(FileOptions(), &ioptions_, nullptr,
+                             vstorage_.get(), nullptr);
+      builder.SaveTo(temp_vstorage_.get());
+      vstorage_ = std::move(temp_vstorage_);
+    }
    vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
    vstorage_->UpdateFilesByCompactionPri(ioptions_.compaction_pri);
    vstorage_->UpdateNumNonEmptyLevels();
@@ -132,13 +155,36 @@ class CompactionPickerTest : public testing::Test {
    vstorage_->ComputeFilesMarkedForCompaction();
    vstorage_->SetFinalized();
  }
+  void AddFileToVersionStorage(int level, uint32_t file_number,
+                               const char* smallest, const char* largest,
+                               uint64_t file_size = 1, uint32_t path_id = 0,
+                               SequenceNumber smallest_seq = 100,
+                               SequenceNumber largest_seq = 100,
+                               size_t compensated_file_size = 0,
+                               bool marked_for_compact = false) {
+    VersionStorageInfo* base_vstorage = vstorage_.release();
+    vstorage_.reset(new VersionStorageInfo(&icmp_, ucmp_, options_.num_levels,
+                                           kCompactionStyleUniversal,
+                                           base_vstorage, false));
+    Add(level, file_number, smallest, largest, file_size, path_id, smallest_seq,
+        largest_seq, compensated_file_size, marked_for_compact);
+
+    VersionBuilder builder(FileOptions(), &ioptions_, nullptr, base_vstorage,
+                           nullptr);
+    builder.SaveTo(vstorage_.get());
+    UpdateVersionStorageInfo();
+  }
+
+ private:
+  std::unique_ptr<VersionStorageInfo> temp_vstorage_;
 };

 TEST_F(CompactionPickerTest, Empty) {
  NewVersionStorage(6, kCompactionStyleLevel);
  UpdateVersionStorageInfo();
  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() == nullptr);
 }

@@ -149,7 +195,8 @@ TEST_F(CompactionPickerTest, Single) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() == nullptr);
 }

@@ -162,7 +209,8 @@ TEST_F(CompactionPickerTest, Level0Trigger) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_files(0));
  ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
@@ -175,7 +223,8 @@ TEST_F(CompactionPickerTest, Level1Trigger) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  ASSERT_EQ(66U, compaction->input(0, 0)->fd.GetNumber());
@@ -193,7 +242,8 @@ TEST_F(CompactionPickerTest, Level1Trigger2) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  ASSERT_EQ(2U, compaction->num_input_files(1));
@@ -224,7 +274,8 @@ TEST_F(CompactionPickerTest, LevelMaxScore) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  ASSERT_EQ(7U, compaction->input(0, 0)->fd.GetNumber());
@@ -271,7 +322,8 @@ TEST_F(CompactionPickerTest, Level0TriggerDynamic) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_files(0));
  ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
@@ -295,7 +347,8 @@ TEST_F(CompactionPickerTest, Level0TriggerDynamic2) {
  ASSERT_EQ(vstorage_->base_level(), num_levels - 2);

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_files(0));
  ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
@@ -320,7 +373,8 @@ TEST_F(CompactionPickerTest, Level0TriggerDynamic3) {
  ASSERT_EQ(vstorage_->base_level(), num_levels - 3);

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_files(0));
  ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
@@ -349,7 +403,8 @@ TEST_F(CompactionPickerTest, Level0TriggerDynamic4) {
  ASSERT_EQ(vstorage_->base_level(), num_levels - 3);

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_files(0));
  ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
@@ -371,8 +426,8 @@ TEST_F(CompactionPickerTest, LevelTriggerDynamic4) {
  mutable_cf_options_.max_bytes_for_level_multiplier = 10;
  NewVersionStorage(num_levels, kCompactionStyleLevel);
  Add(0, 1U, "150", "200");
-  Add(num_levels - 1, 3U, "200", "250", 300U);
-  Add(num_levels - 1, 4U, "300", "350", 3000U);
+  Add(num_levels - 1, 2U, "200", "250", 300U);
+  Add(num_levels - 1, 3U, "300", "350", 3000U);
  Add(num_levels - 1, 4U, "400", "450", 3U);
  Add(num_levels - 2, 5U, "150", "180", 300U);
  Add(num_levels - 2, 6U, "181", "350", 500U);
@@ -381,7 +436,8 @@ TEST_F(CompactionPickerTest, LevelTriggerDynamic4) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  ASSERT_EQ(5U, compaction->input(0, 0)->fd.GetNumber());
@@ -438,7 +494,8 @@ TEST_F(CompactionPickerTest, CompactionUniversalIngestBehindReservedLevel) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));

  // output level should be the one above the bottom-most
  ASSERT_EQ(1, compaction->output_level());
@@ -472,7 +529,8 @@ TEST_F(CompactionPickerTest, CannotTrivialMoveUniversal) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));

  ASSERT_TRUE(!compaction->is_trivial_move());
 }
@@ -498,7 +556,8 @@ TEST_F(CompactionPickerTest, AllowsTrivialMoveUniversal) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));

  ASSERT_TRUE(compaction->is_trivial_move());
 }
@@ -526,7 +585,8 @@ TEST_F(CompactionPickerTest, UniversalPeriodicCompaction1) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));

  ASSERT_TRUE(compaction);
  ASSERT_EQ(4, compaction->output_level());
@@ -556,7 +616,8 @@ TEST_F(CompactionPickerTest, UniversalPeriodicCompaction2) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));

  ASSERT_FALSE(compaction);
 }
@@ -582,7 +643,8 @@ TEST_F(CompactionPickerTest, UniversalPeriodicCompaction3) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));

  ASSERT_FALSE(compaction);
 }
@@ -612,7 +674,8 @@ TEST_F(CompactionPickerTest, UniversalPeriodicCompaction4) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
  ASSERT_TRUE(!compaction ||
              compaction->start_level() != compaction->output_level());
 }
@@ -632,7 +695,8 @@ TEST_F(CompactionPickerTest, UniversalPeriodicCompaction5) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
  ASSERT_TRUE(compaction);
  ASSERT_EQ(0, compaction->start_level());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -656,7 +720,8 @@ TEST_F(CompactionPickerTest, UniversalPeriodicCompaction6) {

  std::unique_ptr<Compaction> compaction(
      universal_compaction_picker.PickCompaction(
-          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
  ASSERT_TRUE(compaction);
  ASSERT_EQ(4, compaction->start_level());
  ASSERT_EQ(2U, compaction->num_input_files(0));
@@ -716,7 +781,8 @@ TEST_F(CompactionPickerTest, CompactionPriMinOverlapping1) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  // Pick file 8 because it overlaps with 0 files on level 3.
@@ -739,7 +805,7 @@ TEST_F(CompactionPickerTest, CompactionPriMinOverlapping2) {
  Add(2, 8U, "201", "300",
      60000000U);  // Overlaps with file 28, 29, total size 521M

-  Add(3, 26U, "100", "110", 261000000U);
+  Add(3, 25U, "100", "110", 261000000U);
  Add(3, 26U, "150", "170", 261000000U);
  Add(3, 27U, "171", "179", 260000000U);
  Add(3, 28U, "191", "220", 260000000U);
@@ -748,7 +814,8 @@ TEST_F(CompactionPickerTest, CompactionPriMinOverlapping2) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  // Picking file 7 because overlapping ratio is the biggest.
@@ -775,7 +842,8 @@ TEST_F(CompactionPickerTest, CompactionPriMinOverlapping3) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  // Picking file 8 because overlapping ratio is the biggest.
@@ -804,7 +872,8 @@ TEST_F(CompactionPickerTest, CompactionPriMinOverlapping4) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_files(0));
  // Picking file 8 because overlapping ratio is the biggest.
@@ -831,7 +900,8 @@ TEST_F(CompactionPickerTest, ParentIndexResetBug) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
 }

 // This test checks ExpandWhileOverlapping() by having overlapping user keys
@@ -848,7 +918,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-              cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_levels());
  ASSERT_EQ(2U, compaction->num_input_files(0));
@@ -867,7 +938,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys2) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-              cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(2U, compaction->num_input_files(0));
@@ -894,7 +966,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys3) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-              cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(5U, compaction->num_input_files(0));
@@ -924,7 +997,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys4) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -947,7 +1021,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys5) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() == nullptr);
 }

@@ -968,7 +1043,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys6) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -988,7 +1064,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys7) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_GE(1U, compaction->num_input_files(0));
@@ -1016,7 +1093,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys8) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(3U, compaction->num_input_files(0));
@@ -1048,7 +1126,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys9) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(5U, compaction->num_input_files(0));
@@ -1088,7 +1167,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys10) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -1126,7 +1206,8 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys11) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -1163,7 +1244,8 @@ TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri1) {
  ASSERT_EQ(0, vstorage_->CompactionScoreLevel(0));
  ASSERT_EQ(1, vstorage_->CompactionScoreLevel(1));
  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() == nullptr);
 }

@@ -1193,7 +1275,8 @@ TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri2) {
  ASSERT_EQ(0, vstorage_->CompactionScoreLevel(0));
  ASSERT_EQ(1, vstorage_->CompactionScoreLevel(1));
  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
 }

@@ -1226,7 +1309,8 @@ TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri3) {
  ASSERT_EQ(1, vstorage_->CompactionScoreLevel(0));
  ASSERT_EQ(0, vstorage_->CompactionScoreLevel(1));
  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
 }

@@ -1255,7 +1339,7 @@ TEST_F(CompactionPickerTest, EstimateCompactionBytesNeeded1) {
  // Size ratio L4/L3 is 9.9
  // After merge from L3, L4 size is 1000900
  Add(4, 11U, "400", "500", 999900);
-  Add(5, 11U, "400", "500", 8007200);
+  Add(5, 12U, "400", "500", 8007200);

  UpdateVersionStorageInfo();

@@ -1520,7 +1604,8 @@ TEST_F(CompactionPickerTest, MaxCompactionBytesHit) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -1544,7 +1629,8 @@ TEST_F(CompactionPickerTest, MaxCompactionBytesNotHit) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(2U, compaction->num_input_levels());
  ASSERT_EQ(3U, compaction->num_input_files(0));
@@ -1568,16 +1654,43 @@ TEST_F(CompactionPickerTest, IsTrivialMoveOn) {

  Add(3, 5U, "120", "130", 7000U);
  Add(3, 6U, "170", "180", 7000U);
-  Add(3, 5U, "220", "230", 7000U);
-  Add(3, 5U, "270", "280", 7000U);
+  Add(3, 7U, "220", "230", 7000U);
+  Add(3, 8U, "270", "280", 7000U);
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-    cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_TRUE(compaction->IsTrivialMove());
 }

+TEST_F(CompactionPickerTest, IsTrivialMoveOffSstPartitioned) {
+  mutable_cf_options_.max_bytes_for_level_base = 10000u;
+  mutable_cf_options_.max_compaction_bytes = 10001u;
+  ioptions_.level_compaction_dynamic_level_bytes = false;
+  ioptions_.sst_partitioner_factory = NewSstPartitionerFixedPrefixFactory(1);
+  NewVersionStorage(6, kCompactionStyleLevel);
+  // A compaction should be triggered and pick file 2
+  Add(1, 1U, "100", "150", 3000U);
+  Add(1, 2U, "151", "200", 3001U);
+  Add(1, 3U, "201", "250", 3000U);
+  Add(1, 4U, "251", "300", 3000U);
+
+  Add(3, 5U, "120", "130", 7000U);
+  Add(3, 6U, "170", "180", 7000U);
+  Add(3, 7U, "220", "230", 7000U);
+  Add(3, 8U, "270", "280", 7000U);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
+  ASSERT_TRUE(compaction.get() != nullptr);
+  // No trivial move, because partitioning is applied
+  ASSERT_TRUE(!compaction->IsTrivialMove());
+}
+
 TEST_F(CompactionPickerTest, IsTrivialMoveOff) {
  mutable_cf_options_.max_bytes_for_level_base = 1000000u;
  mutable_cf_options_.max_compaction_bytes = 10000u;
@@ -1594,7 +1707,8 @@ TEST_F(CompactionPickerTest, IsTrivialMoveOff) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-    cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_FALSE(compaction->IsTrivialMove());
 }
@@ -1619,7 +1733,8 @@ TEST_F(CompactionPickerTest, CacheNextCompactionIndex) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -1628,7 +1743,8 @@ TEST_F(CompactionPickerTest, CacheNextCompactionIndex) {
  ASSERT_EQ(2, vstorage_->NextCompactionIndex(1 /* level */));

  compaction.reset(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_levels());
  ASSERT_EQ(1U, compaction->num_input_files(0));
@@ -1637,7 +1753,8 @@ TEST_F(CompactionPickerTest, CacheNextCompactionIndex) {
  ASSERT_EQ(3, vstorage_->NextCompactionIndex(1 /* level */));

  compaction.reset(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() == nullptr);
  ASSERT_EQ(4, vstorage_->NextCompactionIndex(1 /* level */));
 }
@@ -1662,7 +1779,8 @@ TEST_F(CompactionPickerTest, IntraL0MaxCompactionBytesNotHit) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_levels());
  ASSERT_EQ(5U, compaction->num_input_files(0));
@@ -1692,7 +1810,8 @@ TEST_F(CompactionPickerTest, IntraL0MaxCompactionBytesHit) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_levels());
  ASSERT_EQ(4U, compaction->num_input_files(0));
@@ -1724,7 +1843,8 @@ TEST_F(CompactionPickerTest, IntraL0ForEarliestSeqno) {
  UpdateVersionStorageInfo();

  std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
-      cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_, 107));
+      cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+      &log_buffer_, 107));
  ASSERT_TRUE(compaction.get() != nullptr);
  ASSERT_EQ(1U, compaction->num_input_levels());
  ASSERT_EQ(4U, compaction->num_input_files(0));
@@ -1733,6 +1853,171 @@ TEST_F(CompactionPickerTest, IntraL0ForEarliestSeqno) {
  ASSERT_EQ(0, compaction->output_level());
 }

+#ifndef ROCKSDB_LITE
+TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap) {
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+
+  // This test covers the case where a "regular" universal compaction is
+  // scheduled first, followed by a delete triggered compaction. The latter
+  // should fail
+  NewVersionStorage(5, kCompactionStyleUniversal);
+
+  Add(0, 1U, "150", "200", kFileSize, 0, 500, 550);
+  Add(0, 2U, "201", "250", 2 * kFileSize, 0, 401, 450);
+  Add(0, 4U, "260", "300", 4 * kFileSize, 0, 260, 300);
+  Add(3, 5U, "010", "080", 8 * kFileSize, 0, 200, 251);
+  Add(4, 3U, "301", "350", 8 * kFileSize, 0, 101, 150);
+  Add(4, 6U, "501", "750", 8 * kFileSize, 0, 101, 150);
+
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
+
+  ASSERT_TRUE(compaction);
+  // Validate that its a compaction to reduce sorted runs
+  ASSERT_EQ(CompactionReason::kUniversalSortedRunNum,
+            compaction->compaction_reason());
+  ASSERT_EQ(0, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(2U, compaction->num_input_files(0));
+
+  AddVersionStorage();
+  // Simulate a flush and mark the file for compaction
+  Add(0, 7U, "150", "200", kFileSize, 0, 551, 600, 0, true);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction2(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
+  ASSERT_FALSE(compaction2);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap2) {
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+
+  // This test covers the case where a delete triggered compaction is
+  // scheduled first, followed by a "regular" compaction. The latter
+  // should fail
+  NewVersionStorage(5, kCompactionStyleUniversal);
+
+  // Mark file number 4 for compaction
+  Add(0, 4U, "260", "300", 4 * kFileSize, 0, 260, 300, 0, true);
+  Add(3, 5U, "240", "290", 8 * kFileSize, 0, 201, 250);
+  Add(4, 3U, "301", "350", 8 * kFileSize, 0, 101, 150);
+  Add(4, 6U, "501", "750", 8 * kFileSize, 0, 101, 150);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
+
+  ASSERT_TRUE(compaction);
+  // Validate that its a delete triggered compaction
+  ASSERT_EQ(CompactionReason::kFilesMarkedForCompaction,
+            compaction->compaction_reason());
+  ASSERT_EQ(3, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(1U, compaction->num_input_files(0));
+  ASSERT_EQ(1U, compaction->num_input_files(1));
+
+  AddVersionStorage();
+  Add(0, 1U, "150", "200", kFileSize, 0, 500, 550);
+  Add(0, 2U, "201", "250", 2 * kFileSize, 0, 401, 450);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction2(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          &log_buffer_));
+  ASSERT_FALSE(compaction2);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedCompactionStartOutputOverlap) {
+  // The case where universal periodic compaction can be picked
+  // with some newer files being compacted.
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+
+  bool input_level_overlap = false;
+  bool output_level_overlap = false;
+  // Let's mark 2 files in 2 different levels for compaction. The
+  // compaction picker will randomly pick one, so use the sync point to
+  // ensure a deterministic order. Loop until both cases are covered
+  size_t random_index = 0;
+  SyncPoint::GetInstance()->SetCallBack(
+      "CompactionPicker::PickFilesMarkedForCompaction", [&](void* arg) {
+        size_t* index = static_cast<size_t*>(arg);
+        *index = random_index;
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+  while (!input_level_overlap || !output_level_overlap) {
+    // Ensure that the L0 file gets picked first
+    random_index = !input_level_overlap ? 0 : 1;
+    UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+    NewVersionStorage(5, kCompactionStyleUniversal);
+
+    Add(0, 1U, "260", "300", 4 * kFileSize, 0, 260, 300, 0, true);
+    Add(3, 2U, "010", "020", 2 * kFileSize, 0, 201, 248);
+    Add(3, 3U, "250", "270", 2 * kFileSize, 0, 202, 249);
+    Add(3, 4U, "290", "310", 2 * kFileSize, 0, 203, 250);
+    Add(3, 5U, "310", "320", 2 * kFileSize, 0, 204, 251, 0, true);
+    Add(4, 6U, "301", "350", 8 * kFileSize, 0, 101, 150);
+    Add(4, 7U, "501", "750", 8 * kFileSize, 0, 101, 150);
+    UpdateVersionStorageInfo();
+
+    std::unique_ptr<Compaction> compaction(
+        universal_compaction_picker.PickCompaction(
+            cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+            &log_buffer_));
+
+    ASSERT_TRUE(compaction);
+    // Validate that its a delete triggered compaction
+    ASSERT_EQ(CompactionReason::kFilesMarkedForCompaction,
+              compaction->compaction_reason());
+    ASSERT_TRUE(compaction->start_level() == 0 ||
+                compaction->start_level() == 3);
+    if (compaction->start_level() == 0) {
+      // The L0 file was picked. The next compaction will detect an
+      // overlap on its input level
+      input_level_overlap = true;
+      ASSERT_EQ(3, compaction->output_level());
+      ASSERT_EQ(1U, compaction->num_input_files(0));
+      ASSERT_EQ(3U, compaction->num_input_files(1));
+    } else {
+      // The level 3 file was picked. The next compaction will pick
+      // the L0 file and will detect overlap when adding output
+      // level inputs
+      output_level_overlap = true;
+      ASSERT_EQ(4, compaction->output_level());
+      ASSERT_EQ(2U, compaction->num_input_files(0));
+      ASSERT_EQ(1U, compaction->num_input_files(1));
+    }
+
+    vstorage_->ComputeCompactionScore(ioptions_, mutable_cf_options_);
+    // After recomputing the compaction score, only one marked file will remain
+    random_index = 0;
+    std::unique_ptr<Compaction> compaction2(
+        universal_compaction_picker.PickCompaction(
+            cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+            &log_buffer_));
+    ASSERT_FALSE(compaction2);
+    DeleteVersionStorage();
+  }
+}
+#endif  // ROCKSDB_LITE
+
 }  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
@@ -31,17 +31,16 @@ namespace {
 // PickCompaction().
 class UniversalCompactionBuilder {
 public:
-  UniversalCompactionBuilder(const ImmutableCFOptions& ioptions,
-                             const InternalKeyComparator* icmp,
-                             const std::string& cf_name,
-                             const MutableCFOptions& mutable_cf_options,
-                             VersionStorageInfo* vstorage,
-                             UniversalCompactionPicker* picker,
-                             LogBuffer* log_buffer)
+  UniversalCompactionBuilder(
+      const ImmutableCFOptions& ioptions, const InternalKeyComparator* icmp,
+      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+      UniversalCompactionPicker* picker, LogBuffer* log_buffer)
      : ioptions_(ioptions),
        icmp_(icmp),
        cf_name_(cf_name),
        mutable_cf_options_(mutable_cf_options),
+        mutable_db_options_(mutable_db_options),
        vstorage_(vstorage),
        picker_(picker),
        log_buffer_(log_buffer) {}
@@ -115,13 +114,13 @@ class UniversalCompactionBuilder {
  std::vector<SortedRun> sorted_runs_;
  const std::string& cf_name_;
  const MutableCFOptions& mutable_cf_options_;
+  const MutableDBOptions& mutable_db_options_;
  VersionStorageInfo* vstorage_;
  UniversalCompactionPicker* picker_;
  LogBuffer* log_buffer_;

  static std::vector<SortedRun> CalculateSortedRuns(
-      const VersionStorageInfo& vstorage, const ImmutableCFOptions& ioptions,
-      const MutableCFOptions& mutable_cf_options);
+      const VersionStorageInfo& vstorage);

  // Pick a path ID to place a newly generated file, with its estimated file
  // size.
@@ -278,11 +277,11 @@ bool UniversalCompactionPicker::NeedsCompaction(

 Compaction* UniversalCompactionPicker::PickCompaction(
    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer,
-    SequenceNumber /* earliest_memtable_seqno */) {
+    const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+    LogBuffer* log_buffer, SequenceNumber /* earliest_memtable_seqno */) {
  UniversalCompactionBuilder builder(ioptions_, icmp_, cf_name,
-                                     mutable_cf_options, vstorage, this,
-                                     log_buffer);
+                                     mutable_cf_options, mutable_db_options,
+                                     vstorage, this, log_buffer);
  return builder.PickCompaction();
 }

@@ -325,8 +324,7 @@ void UniversalCompactionBuilder::SortedRun::DumpSizeInfo(

 std::vector<UniversalCompactionBuilder::SortedRun>
 UniversalCompactionBuilder::CalculateSortedRuns(
-    const VersionStorageInfo& vstorage, const ImmutableCFOptions& /*ioptions*/,
-    const MutableCFOptions& mutable_cf_options) {
+    const VersionStorageInfo& vstorage) {
  std::vector<UniversalCompactionBuilder::SortedRun> ret;
  for (FileMetaData* f : vstorage.LevelFiles(0)) {
    ret.emplace_back(0, f, f->fd.GetFileSize(), f->compensated_file_size,
@@ -336,27 +334,16 @@ UniversalCompactionBuilder::CalculateSortedRuns(
    uint64_t total_compensated_size = 0U;
    uint64_t total_size = 0U;
    bool being_compacted = false;
-    bool is_first = true;
    for (FileMetaData* f : vstorage.LevelFiles(level)) {
      total_compensated_size += f->compensated_file_size;
      total_size += f->fd.GetFileSize();
-      if (mutable_cf_options.compaction_options_universal.allow_trivial_move ==
-          true) {
-        if (f->being_compacted) {
-          being_compacted = f->being_compacted;
-        }
-      } else {
-        // Compaction always includes all files for a non-zero level, so for a
-        // non-zero level, all the files should share the same being_compacted
-        // value.
-        // This assumption is only valid when
-        // mutable_cf_options.compaction_options_universal.allow_trivial_move
-        // is false
-        assert(is_first || f->being_compacted == being_compacted);
-      }
-      if (is_first) {
+      // Size amp, read amp and periodic compactions always include all files
+      // for a non-zero level. However, a delete triggered compaction and
+      // a trivial move might pick a subset of files in a sorted run. So
+      // always check all files in a sorted run and mark the entire run as
+      // being compacted if one or more files are being compacted
+      if (f->being_compacted) {
        being_compacted = f->being_compacted;
-        is_first = false;
      }
    }
    if (total_compensated_size > 0) {
@@ -372,8 +359,7 @@ UniversalCompactionBuilder::CalculateSortedRuns(
 Compaction* UniversalCompactionBuilder::PickCompaction() {
  const int kLevel0 = 0;
  score_ = vstorage_->CompactionScore(kLevel0);
-  sorted_runs_ =
-      CalculateSortedRuns(*vstorage_, ioptions_, mutable_cf_options_);
+  sorted_runs_ = CalculateSortedRuns(*vstorage_);

  if (sorted_runs_.size() == 0 ||
      (vstorage_->FilesMarkedForPeriodicCompaction().empty() &&
@@ -744,8 +730,8 @@ Compaction* UniversalCompactionBuilder::PickCompactionToReduceSortedRuns(
    compaction_reason = CompactionReason::kUniversalSortedRunNum;
  }
  return new Compaction(
-      vstorage_, ioptions_, mutable_cf_options_, std::move(inputs),
-      output_level,
+      vstorage_, ioptions_, mutable_cf_options_, mutable_db_options_,
+      std::move(inputs), output_level,
      MaxFileSizeForLevel(mutable_cf_options_, output_level,
                          kCompactionStyleUniversal),
      LLONG_MAX, path_id,
@@ -779,7 +765,7 @@ Compaction* UniversalCompactionBuilder::PickCompactionToReduceSizeAmp() {
  }

  // Skip files that are already being compacted
-  for (size_t loop = 0; loop < sorted_runs_.size() - 1; loop++) {
+  for (size_t loop = 0; loop + 1 < sorted_runs_.size(); loop++) {
    sr = &sorted_runs_[loop];
    if (!sr->being_compacted) {
      start_index = loop;  // Consider this as the first candidate.
@@ -807,7 +793,7 @@ Compaction* UniversalCompactionBuilder::PickCompactionToReduceSizeAmp() {
  }

  // keep adding up all the remaining files
-  for (size_t loop = start_index; loop < sorted_runs_.size() - 1; loop++) {
+  for (size_t loop = start_index; loop + 1 < sorted_runs_.size(); loop++) {
    sr = &sorted_runs_[loop];
    if (sr->being_compacted) {
      char file_num_buf[kFormatFileNumberBufSize];
@@ -855,6 +841,7 @@ Compaction* UniversalCompactionBuilder::PickDeleteTriggeredCompaction() {
  std::vector<CompactionInputFiles> inputs;

  if (vstorage_->num_levels() == 1) {
+#if defined(ENABLE_SINGLE_LEVEL_DTC)
    // This is single level universal. Since we're basically trying to reclaim
    // space by processing files marked for compaction due to high tombstone
    // density, let's do the same thing as compaction to reduce size amp which
@@ -877,6 +864,11 @@ Compaction* UniversalCompactionBuilder::PickDeleteTriggeredCompaction() {
      return nullptr;
    }
    inputs.push_back(start_level_inputs);
+#else
+    // Disable due to a known race condition.
+    // TODO: Reenable once the race condition is fixed
+    return nullptr;
+#endif  // ENABLE_SINGLE_LEVEL_DTC
  } else {
    int start_level;

@@ -952,8 +944,8 @@ Compaction* UniversalCompactionBuilder::PickDeleteTriggeredCompaction() {
  uint32_t path_id =
      GetPathId(ioptions_, mutable_cf_options_, estimated_total_size);
  return new Compaction(
-      vstorage_, ioptions_, mutable_cf_options_, std::move(inputs),
-      output_level,
+      vstorage_, ioptions_, mutable_cf_options_, mutable_db_options_,
+      std::move(inputs), output_level,
      MaxFileSizeForLevel(mutable_cf_options_, output_level,
                          kCompactionStyleUniversal),
      /* max_grandparent_overlap_bytes */ LLONG_MAX, path_id,
@@ -1022,14 +1014,14 @@ Compaction* UniversalCompactionBuilder::PickCompactionToOldest(
  // compaction_options_universal.compression_size_percent,
  // because we always compact all the files, so always compress.
  return new Compaction(
-      vstorage_, ioptions_, mutable_cf_options_, std::move(inputs),
-      output_level,
+      vstorage_, ioptions_, mutable_cf_options_, mutable_db_options_,
+      std::move(inputs), output_level,
      MaxFileSizeForLevel(mutable_cf_options_, output_level,
                          kCompactionStyleUniversal),
      LLONG_MAX, path_id,
-      GetCompressionType(ioptions_, vstorage_, mutable_cf_options_, start_level,
-                         1, true /* enable_compression */),
-      GetCompressionOptions(mutable_cf_options_, vstorage_, start_level,
+      GetCompressionType(ioptions_, vstorage_, mutable_cf_options_,
+                         output_level, 1, true /* enable_compression */),
+      GetCompressionOptions(mutable_cf_options_, vstorage_, output_level,
                            true /* enable_compression */),
      /* max_subcompactions */ 0, /* grandparents */ {}, /* is manual */ false,
      score_, false /* deletion_compaction */, compaction_reason);
@@ -20,7 +20,8 @@ class UniversalCompactionPicker : public CompactionPicker {
      : CompactionPicker(ioptions, icmp) {}
  virtual Compaction* PickCompaction(
      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* vstorage, LogBuffer* log_buffer,
+      const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
+      LogBuffer* log_buffer,
      SequenceNumber earliest_memtable_seqno = kMaxSequenceNumber) override;
  virtual int MaxOutputLevel() const override { return NumberLevels() - 1; }

@@ -0,0 +1,44 @@
+//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under both the GPLv2 (found in the
+//  COPYING file in the root directory) and Apache 2.0 License
+//  (found in the LICENSE.Apache file in the root directory).
+//
+
+#include "rocksdb/sst_partitioner.h"
+
+#include <algorithm>
+
+namespace ROCKSDB_NAMESPACE {
+
+PartitionerResult SstPartitionerFixedPrefix::ShouldPartition(
+    const PartitionerRequest& request) {
+  Slice last_key_fixed(*request.prev_user_key);
+  if (last_key_fixed.size() > len_) {
+    last_key_fixed.size_ = len_;
+  }
+  Slice current_key_fixed(*request.current_user_key);
+  if (current_key_fixed.size() > len_) {
+    current_key_fixed.size_ = len_;
+  }
+  return last_key_fixed.compare(current_key_fixed) != 0 ? kRequired
+                                                        : kNotRequired;
+}
+
+bool SstPartitionerFixedPrefix::CanDoTrivialMove(
+    const Slice& smallest_user_key, const Slice& largest_user_key) {
+  return ShouldPartition(PartitionerRequest(smallest_user_key, largest_user_key,
+                                            0)) == kNotRequired;
+}
+
+std::unique_ptr<SstPartitioner>
+SstPartitionerFixedPrefixFactory::CreatePartitioner(
+    const SstPartitioner::Context& /* context */) const {
+  return std::unique_ptr<SstPartitioner>(new SstPartitionerFixedPrefix(len_));
+}
+
+std::shared_ptr<SstPartitionerFactory> NewSstPartitionerFixedPrefixFactory(
+    size_t prefix_len) {
+  return std::make_shared<SstPartitionerFixedPrefixFactory>(prefix_len);
+}
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -13,6 +13,7 @@
 #include "test_util/testutil.h"
 #include "util/hash.h"
 #include "util/kv_map.h"
+#include "util/random.h"
 #include "util/string_util.h"
 #include "utilities/merge_operators.h"

@@ -342,12 +343,12 @@ TEST_P(ComparatorDBTest, SimpleSuffixReverseComparator) {
    std::vector<std::string> source_prefixes;
    // Randomly generate 5 prefixes
    for (int i = 0; i < 5; i++) {
-      source_prefixes.push_back(test::RandomHumanReadableString(&rnd, 8));
+      source_prefixes.push_back(rnd.HumanReadableString(8));
    }
    for (int j = 0; j < 20; j++) {
      int prefix_index = rnd.Uniform(static_cast<int>(source_prefixes.size()));
      std::string key = source_prefixes[prefix_index] +
-                        test::RandomHumanReadableString(&rnd, rnd.Uniform(8));
+                        rnd.HumanReadableString(rnd.Uniform(8));
      source_strings.push_back(key);
    }

@@ -61,7 +61,8 @@ Status VerifySstFileChecksum(const Options& options,
  s = ioptions.table_factory->NewTableReader(
      TableReaderOptions(ioptions, options.prefix_extractor.get(), env_options,
                         internal_comparator, false /* skip_filters */,
-                         !kImmortal, -1 /* level */),
+                         !kImmortal, false /* force_direct_prefetch */,
+                         -1 /* level */),
      std::move(file_reader), file_size, &table_reader,
      false /* prefetch_index_and_filter_in_cache */);
  if (!s.ok()) {
@@ -9,13 +9,13 @@

 #ifndef ROCKSDB_LITE

-#include "rocksdb/db.h"
-
 #include <errno.h>
 #include <fcntl.h>
 #include <sys/stat.h>
 #include <sys/types.h>
+
 #include <cinttypes>
+
 #include "db/db_impl/db_impl.h"
 #include "db/db_test_util.h"
 #include "db/log_format.h"
@@ -24,13 +24,16 @@
 #include "file/filename.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/convenience.h"
+#include "rocksdb/db.h"
 #include "rocksdb/env.h"
 #include "rocksdb/table.h"
 #include "rocksdb/write_batch.h"
 #include "table/block_based/block_based_table_builder.h"
 #include "table/meta_blocks.h"
+#include "table/mock_table.h"
 #include "test_util/testharness.h"
 #include "test_util/testutil.h"
+#include "util/random.h"
 #include "util/string_util.h"

 namespace ROCKSDB_NAMESPACE {
@@ -101,22 +104,24 @@ class CorruptionTest : public testing::Test {
    ASSERT_OK(::ROCKSDB_NAMESPACE::RepairDB(dbname_, options_));
  }

-  void Build(int n, int flush_every = 0) {
+  void Build(int n, int start, int flush_every) {
    std::string key_space, value_space;
    WriteBatch batch;
    for (int i = 0; i < n; i++) {
      if (flush_every != 0 && i != 0 && i % flush_every == 0) {
-        DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+        DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
        dbi->TEST_FlushMemTable();
      }
      //if ((i % 100) == 0) fprintf(stderr, "@ %d of %d\n", i, n);
-      Slice key = Key(i, &key_space);
+      Slice key = Key(i + start, &key_space);
      batch.Clear();
      batch.Put(key, Value(i, &value_space));
      ASSERT_OK(db_->Write(WriteOptions(), &batch));
    }
  }

+  void Build(int n, int flush_every = 0) { Build(n, 0, flush_every); }
+
  void Check(int min_expected, int max_expected) {
    uint64_t next_expected = 0;
    uint64_t missed = 0;
@@ -157,42 +162,6 @@ class CorruptionTest : public testing::Test {
    ASSERT_GE(max_expected, correct);
  }

-  void CorruptFile(const std::string& fname, int offset, int bytes_to_corrupt) {
-    struct stat sbuf;
-    if (stat(fname.c_str(), &sbuf) != 0) {
-      const char* msg = strerror(errno);
-      FAIL() << fname << ": " << msg;
-    }
-
-    if (offset < 0) {
-      // Relative to end of file; make it absolute
-      if (-offset > sbuf.st_size) {
-        offset = 0;
-      } else {
-        offset = static_cast<int>(sbuf.st_size + offset);
-      }
-    }
-    if (offset > sbuf.st_size) {
-      offset = static_cast<int>(sbuf.st_size);
-    }
-    if (offset + bytes_to_corrupt > sbuf.st_size) {
-      bytes_to_corrupt = static_cast<int>(sbuf.st_size - offset);
-    }
-
-    // Do it
-    std::string contents;
-    Status s = ReadFileToString(Env::Default(), fname, &contents);
-    ASSERT_TRUE(s.ok()) << s.ToString();
-    for (int i = 0; i < bytes_to_corrupt; i++) {
-      contents[i + offset] ^= 0x80;
-    }
-    s = WriteStringToFile(Env::Default(), contents, fname);
-    ASSERT_TRUE(s.ok()) << s.ToString();
-    Options options;
-    EnvOptions env_options;
-    ASSERT_NOK(VerifySstFileChecksum(options, env_options, fname));
-  }
-
  void Corrupt(FileType filetype, int offset, int bytes_to_corrupt) {
    // Pick file to corrupt
    std::vector<std::string> filenames;
@@ -211,7 +180,7 @@ class CorruptionTest : public testing::Test {
    }
    ASSERT_TRUE(!fname.empty()) << filetype;

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

  // corrupts exactly one file at level `level`. if no file found at level,
@@ -221,7 +190,7 @@ class CorruptionTest : public testing::Test {
    db_->GetLiveFilesMetaData(&metadata);
    for (const auto& m : metadata) {
      if (m.level == level) {
-        CorruptFile(dbname_ + "/" + m.name, offset, bytes_to_corrupt);
+        test::CorruptFile(dbname_ + "/" + m.name, offset, bytes_to_corrupt);
        return;
      }
    }
@@ -255,11 +224,11 @@ class CorruptionTest : public testing::Test {
      // preserves the implementation that was in place when all of the
      // magic values in this file were picked.
      *storage = std::string(kValueSize, ' ');
-      return Slice(*storage);
    } else {
      Random r(k);
-      return test::RandomString(&r, kValueSize, storage);
+      *storage = r.RandomString(kValueSize);
    }
+    return Slice(*storage);
  }
 };

@@ -317,7 +286,7 @@ TEST_F(CorruptionTest, NewFileErrorDuringWrite) {

 TEST_F(CorruptionTest, TableFile) {
  Build(100);
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  dbi->TEST_FlushMemTable();
  dbi->TEST_CompactRange(0, nullptr, nullptr);
  dbi->TEST_CompactRange(1, nullptr, nullptr);
@@ -340,7 +309,7 @@ TEST_F(CorruptionTest, VerifyChecksumReadahead) {
  Reopen(&options);

  Build(10000);
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  dbi->TEST_FlushMemTable();
  dbi->TEST_CompactRange(0, nullptr, nullptr);
  dbi->TEST_CompactRange(1, nullptr, nullptr);
@@ -387,14 +356,14 @@ TEST_F(CorruptionTest, TableFileIndexData) {
  Reopen(&options);
  // build 2 tables, flush at 5000
  Build(10000, 5000);
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  dbi->TEST_FlushMemTable();

  // corrupt an index block of an entire file
  Corrupt(kTableFile, -2000, 500);
  options.paranoid_checks = false;
  Reopen(&options);
-  dbi = reinterpret_cast<DBImpl*>(db_);
+  dbi = static_cast_with_check<DBImpl>(db_);
  // one full file may be readable, since only one was corrupted
  // the other file should be fully non-readable, since index was corrupted
  Check(0, 5000);
@@ -434,7 +403,7 @@ TEST_F(CorruptionTest, SequenceNumberRecovery) {

 TEST_F(CorruptionTest, CorruptedDescriptor) {
  ASSERT_OK(db_->Put(WriteOptions(), "foo", "hello"));
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  dbi->TEST_FlushMemTable();
  dbi->TEST_CompactRange(0, nullptr, nullptr);

@@ -453,7 +422,7 @@ TEST_F(CorruptionTest, CompactionInputError) {
  Options options;
  Reopen(&options);
  Build(10);
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  dbi->TEST_FlushMemTable();
  dbi->TEST_CompactRange(0, nullptr, nullptr);
  dbi->TEST_CompactRange(1, nullptr, nullptr);
@@ -475,7 +444,7 @@ TEST_F(CorruptionTest, CompactionInputErrorParanoid) {
  options.write_buffer_size = 131072;
  options.max_write_buffer_number = 2;
  Reopen(&options);
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);

  // Fill levels >= 1
  for (int level = 1; level < dbi->NumberLevels(); level++) {
@@ -490,7 +459,7 @@ TEST_F(CorruptionTest, CompactionInputErrorParanoid) {

  Reopen(&options);

-  dbi = reinterpret_cast<DBImpl*>(db_);
+  dbi = static_cast_with_check<DBImpl>(db_);
  Build(10);
  dbi->TEST_FlushMemTable();
  dbi->TEST_WaitForCompact();
@@ -517,7 +486,7 @@ TEST_F(CorruptionTest, CompactionInputErrorParanoid) {

 TEST_F(CorruptionTest, UnrelatedKeys) {
  Build(10);
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+  DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
  dbi->TEST_FlushMemTable();
  Corrupt(kTableFile, 100, 1);
  ASSERT_NOK(dbi->VerifyChecksum());
@@ -556,7 +525,7 @@ TEST_F(CorruptionTest, RangeDeletionCorrupted) {
      ImmutableCFOptions(options_), kRangeDelBlock, &range_del_handle));

  ASSERT_OK(TryReopen());
-  CorruptFile(filename, static_cast<int>(range_del_handle.offset()), 1);
+  test::CorruptFile(filename, static_cast<int>(range_del_handle.offset()), 1);
  ASSERT_TRUE(TryReopen().IsCorruption());
 }

@@ -568,7 +537,7 @@ TEST_F(CorruptionTest, FileSystemStateCorrupted) {
    Reopen(&options);
    Build(10);
    ASSERT_OK(db_->Flush(FlushOptions()));
-    DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+    DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
    std::vector<LiveFileMetaData> metadata;
    dbi->GetLiveFilesMetaData(&metadata);
    ASSERT_GT(metadata.size(), size_t(0));
@@ -594,6 +563,155 @@ TEST_F(CorruptionTest, FileSystemStateCorrupted) {
  }
 }

+static const auto& corruption_modes = {mock::MockTableFactory::kCorruptNone,
+                                       mock::MockTableFactory::kCorruptKey,
+                                       mock::MockTableFactory::kCorruptValue};
+
+TEST_F(CorruptionTest, ParaniodFileChecksOnFlush) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  Status s;
+  for (const auto& mode : corruption_modes) {
+    delete db_;
+    s = DestroyDB(dbname_, options);
+    std::shared_ptr<mock::MockTableFactory> mock =
+        std::make_shared<mock::MockTableFactory>();
+    options.table_factory = mock;
+    mock->SetCorruptionMode(mode);
+    ASSERT_OK(DB::Open(options, dbname_, &db_));
+    Build(10);
+    s = db_->Flush(FlushOptions());
+    if (mode == mock::MockTableFactory::kCorruptNone) {
+      ASSERT_OK(s);
+    } else {
+      ASSERT_NOK(s);
+    }
+  }
+}
+
+TEST_F(CorruptionTest, ParaniodFileChecksOnCompact) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  Status s;
+  for (const auto& mode : corruption_modes) {
+    delete db_;
+    s = DestroyDB(dbname_, options);
+    std::shared_ptr<mock::MockTableFactory> mock =
+        std::make_shared<mock::MockTableFactory>();
+    options.table_factory = mock;
+    ASSERT_OK(DB::Open(options, dbname_, &db_));
+    Build(100, 2);
+    // ASSERT_OK(db_->Flush(FlushOptions()));
+    DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
+    ASSERT_OK(dbi->TEST_FlushMemTable());
+    mock->SetCorruptionMode(mode);
+    s = dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true);
+    if (mode == mock::MockTableFactory::kCorruptNone) {
+      ASSERT_OK(s);
+    } else {
+      ASSERT_NOK(s);
+    }
+  }
+}
+
+TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeFirst) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  for (bool do_flush : {true, false}) {
+    delete db_;
+    db_ = nullptr;
+    ASSERT_OK(DestroyDB(dbname_, options));
+    ASSERT_OK(DB::Open(options, dbname_, &db_));
+    std::string start, end;
+    assert(db_ != nullptr);
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(3, &start), Key(7, &end)));
+    auto snap = db_->GetSnapshot();
+    ASSERT_NE(snap, nullptr);
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(8, &start), Key(9, &end)));
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(2, &start), Key(5, &end)));
+    Build(10);
+    if (do_flush) {
+      ASSERT_OK(db_->Flush(FlushOptions()));
+    } else {
+      DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
+      ASSERT_OK(dbi->TEST_FlushMemTable());
+      ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
+    }
+    db_->ReleaseSnapshot(snap);
+  }
+}
+
+TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRange) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  for (bool do_flush : {true, false}) {
+    delete db_;
+    db_ = nullptr;
+    ASSERT_OK(DestroyDB(dbname_, options));
+    ASSERT_OK(DB::Open(options, dbname_, &db_));
+    assert(db_ != nullptr);
+    Build(10, 0, 0);
+    std::string start, end;
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(5, &start), Key(15, &end)));
+    auto snap = db_->GetSnapshot();
+    ASSERT_NE(snap, nullptr);
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(8, &start), Key(9, &end)));
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(12, &start), Key(17, &end)));
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(2, &start), Key(4, &end)));
+    Build(10, 10, 0);
+    if (do_flush) {
+      ASSERT_OK(db_->Flush(FlushOptions()));
+    } else {
+      DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
+      ASSERT_OK(dbi->TEST_FlushMemTable());
+      ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
+    }
+    db_->ReleaseSnapshot(snap);
+  }
+}
+
+TEST_F(CorruptionTest, ParanoidFileChecksWithDeleteRangeLast) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  for (bool do_flush : {true, false}) {
+    delete db_;
+    db_ = nullptr;
+    ASSERT_OK(DestroyDB(dbname_, options));
+    ASSERT_OK(DB::Open(options, dbname_, &db_));
+    assert(db_ != nullptr);
+    std::string start, end;
+    Build(10);
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(3, &start), Key(7, &end)));
+    auto snap = db_->GetSnapshot();
+    ASSERT_NE(snap, nullptr);
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(6, &start), Key(8, &end)));
+    ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
+                               Key(2, &start), Key(5, &end)));
+    if (do_flush) {
+      ASSERT_OK(db_->Flush(FlushOptions()));
+    } else {
+      DBImpl* dbi = static_cast_with_check<DBImpl>(db_);
+      ASSERT_OK(dbi->TEST_FlushMemTable());
+      ASSERT_OK(dbi->TEST_CompactRange(0, nullptr, nullptr, nullptr, true));
+    }
+    db_->ReleaseSnapshot(snap);
+  }
+}
+
 }  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
@@ -13,6 +13,7 @@
 #include "table/meta_blocks.h"
 #include "test_util/testharness.h"
 #include "test_util/testutil.h"
+#include "util/cast_util.h"
 #include "util/string_util.h"

 namespace ROCKSDB_NAMESPACE {
@@ -46,9 +47,7 @@ class CuckooTableDBTest : public testing::Test {
    return options;
  }

-  DBImpl* dbfull() {
-    return reinterpret_cast<DBImpl*>(db_);
-  }
+  DBImpl* dbfull() { return static_cast_with_check<DBImpl>(db_); }

  // The following util methods are copied from plain_table_db_test.
  void Reopen(Options* options = nullptr) {
@@ -7,10 +7,12 @@
 // 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 <cstdlib>
+
 #include "cache/lru_cache.h"
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
 #include "util/compression.h"
+#include "util/random.h"

 namespace ROCKSDB_NAMESPACE {

@@ -187,7 +189,7 @@ TEST_F(DBBlockCacheTest, TestWithoutCompressedBlockCache) {
  Iterator* iter = nullptr;

  // Load blocks into cache.
-  for (size_t i = 0; i < kNumBlocks - 1; i++) {
+  for (size_t i = 0; i + 1 < kNumBlocks; i++) {
    iter = db_->NewIterator(read_options);
    iter->Seek(ToString(i));
    ASSERT_OK(iter->status());
@@ -209,12 +211,12 @@ TEST_F(DBBlockCacheTest, TestWithoutCompressedBlockCache) {
  iter = nullptr;

  // Release iterators and access cache again.
-  for (size_t i = 0; i < kNumBlocks - 1; i++) {
+  for (size_t i = 0; i + 1 < kNumBlocks; i++) {
    iterators[i].reset();
    CheckCacheCounters(options, 0, 0, 0, 0);
  }
  ASSERT_EQ(0, cache->GetPinnedUsage());
-  for (size_t i = 0; i < kNumBlocks - 1; i++) {
+  for (size_t i = 0; i + 1 < kNumBlocks; i++) {
    iter = db_->NewIterator(read_options);
    iter->Seek(ToString(i));
    ASSERT_OK(iter->status());
@@ -243,7 +245,7 @@ TEST_F(DBBlockCacheTest, TestWithCompressedBlockCache) {
  Iterator* iter = nullptr;

  // Load blocks into cache.
-  for (size_t i = 0; i < kNumBlocks - 1; i++) {
+  for (size_t i = 0; i + 1 < kNumBlocks; i++) {
    iter = db_->NewIterator(read_options);
    iter->Seek(ToString(i));
    ASSERT_OK(iter->status());
@@ -764,7 +766,7 @@ TEST_F(DBBlockCacheTest, CompressedCache) {
    std::string str;
    for (int i = 0; i < num_iter; i++) {
      if (i % 4 == 0) {  // high compression ratio
-        str = RandomString(&rnd, 1000);
+        str = rnd.RandomString(1000);
      }
      values.push_back(str);
      ASSERT_OK(Put(1, Key(i), values[i]));
@@ -851,7 +853,7 @@ TEST_F(DBBlockCacheTest, CacheCompressionDict) {
    for (int i = 0; i < kNumFiles; ++i) {
      ASSERT_EQ(i, NumTableFilesAtLevel(0, 0));
      for (int j = 0; j < kNumEntriesPerFile; ++j) {
-        std::string value = RandomString(&rnd, kNumBytesPerEntry);
+        std::string value = rnd.RandomString(kNumBytesPerEntry);
        ASSERT_OK(Put(Key(j * kNumFiles + i), value.c_str()));
      }
      ASSERT_OK(Flush());
@@ -1063,12 +1063,21 @@ TEST_P(DBBloomFilterTestVaryPrefixAndFormatVer, PartitionedMultiGet) {
  bbto.partition_filters = true;
  bbto.index_type = BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch;
  bbto.whole_key_filtering = !use_prefix_;
-  bbto.metadata_block_size = 128;
+  if (use_prefix_) {  // (not related to prefix, just alternating between)
+    // Make sure code appropriately deals with metadata block size setting
+    // that is "too small" (smaller than minimum size for filter builder)
+    bbto.metadata_block_size = 63;
+  } else {
+    // Make sure the test will work even on platforms with large minimum
+    // filter size, due to large cache line size.
+    // (Largest cache line size + 10+% overhead.)
+    bbto.metadata_block_size = 290;
+  }
  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
  DestroyAndReopen(options);
  ReadOptions ropts;

-  constexpr uint32_t N = 10000;
+  constexpr uint32_t N = 12000;
  // Add N/2 evens
  for (uint32_t i = 0; i < N; i += 2) {
    ASSERT_OK(Put(UKey(i), UKey(i)));
@@ -14,9 +14,10 @@
 #include "rocksdb/experimental.h"
 #include "rocksdb/sst_file_writer.h"
 #include "rocksdb/utilities/convenience.h"
-#include "test_util/fault_injection_test_env.h"
 #include "test_util/sync_point.h"
 #include "util/concurrent_task_limiter_impl.h"
+#include "util/random.h"
+#include "utilities/fault_injection_env.h"

 namespace ROCKSDB_NAMESPACE {

@@ -295,7 +296,7 @@ TEST_P(DBCompactionTestWithParam, CompactionDeletionTrigger) {
    const int kTestSize = kCDTKeysPerBuffer * 1024;
    std::vector<std::string> values;
    for (int k = 0; k < kTestSize; ++k) {
-      values.push_back(RandomString(&rnd, kCDTValueSize));
+      values.push_back(rnd.RandomString(kCDTValueSize));
      ASSERT_OK(Put(Key(k), values[k]));
    }
    dbfull()->TEST_WaitForFlushMemTable();
@@ -343,7 +344,7 @@ TEST_P(DBCompactionTestWithParam, CompactionsPreserveDeletes) {
    const int kTestSize = kCDTKeysPerBuffer;
    std::vector<std::string> values;
    for (int k = 0; k < kTestSize; ++k) {
-      values.push_back(RandomString(&rnd, kCDTValueSize));
+      values.push_back(rnd.RandomString(kCDTValueSize));
      ASSERT_OK(Put(Key(k), values[k]));
    }

@@ -408,7 +409,7 @@ TEST_F(DBCompactionTest, SkipStatsUpdateTest) {
  const int kTestSize = kCDTKeysPerBuffer * 512;
  std::vector<std::string> values;
  for (int k = 0; k < kTestSize; ++k) {
-    values.push_back(RandomString(&rnd, kCDTValueSize));
+    values.push_back(rnd.RandomString(kCDTValueSize));
    ASSERT_OK(Put(Key(k), values[k]));
  }

@@ -555,7 +556,7 @@ TEST_P(DBCompactionTestWithParam, CompactionDeletionTriggerReopen) {
    const int kTestSize = kCDTKeysPerBuffer * 512;
    std::vector<std::string> values;
    for (int k = 0; k < kTestSize; ++k) {
-      values.push_back(RandomString(&rnd, kCDTValueSize));
+      values.push_back(rnd.RandomString(kCDTValueSize));
      ASSERT_OK(Put(Key(k), values[k]));
    }
    dbfull()->TEST_WaitForFlushMemTable();
@@ -673,7 +674,7 @@ TEST_F(DBCompactionTest, DisableStatsUpdateReopen) {
    const int kTestSize = kCDTKeysPerBuffer * 512;
    std::vector<std::string> values;
    for (int k = 0; k < kTestSize; ++k) {
-      values.push_back(RandomString(&rnd, kCDTValueSize));
+      values.push_back(rnd.RandomString(kCDTValueSize));
      ASSERT_OK(Put(Key(k), values[k]));
    }
    dbfull()->TEST_WaitForFlushMemTable();
@@ -736,7 +737,7 @@ TEST_P(DBCompactionTestWithParam, CompactionTrigger) {
    std::vector<std::string> values;
    // Write 100KB (100 values, each 1K)
    for (int i = 0; i < kNumKeysPerFile; i++) {
-      values.push_back(RandomString(&rnd, 990));
+      values.push_back(rnd.RandomString(990));
      ASSERT_OK(Put(1, Key(i), values[i]));
    }
    // put extra key to trigger flush
@@ -748,7 +749,7 @@ TEST_P(DBCompactionTestWithParam, CompactionTrigger) {
  // generate one more file in level-0, and should trigger level-0 compaction
  std::vector<std::string> values;
  for (int i = 0; i < kNumKeysPerFile; i++) {
-    values.push_back(RandomString(&rnd, 990));
+    values.push_back(rnd.RandomString(990));
    ASSERT_OK(Put(1, Key(i), values[i]));
  }
  // put extra key to trigger flush
@@ -867,7 +868,7 @@ TEST_P(DBCompactionTestWithParam, CompactionsGenerateMultipleFiles) {
  ASSERT_EQ(NumTableFilesAtLevel(0, 1), 0);
  std::vector<std::string> values;
  for (int i = 0; i < 80; i++) {
-    values.push_back(RandomString(&rnd, 100000));
+    values.push_back(rnd.RandomString(100000));
    ASSERT_OK(Put(1, Key(i), values[i]));
  }

@@ -976,6 +977,60 @@ TEST_F(DBCompactionTest, UserKeyCrossFile2) {
  ASSERT_EQ("NOT_FOUND", Get("3"));
 }

+TEST_F(DBCompactionTest, CompactionSstPartitioner) {
+  Options options = CurrentOptions();
+  options.compaction_style = kCompactionStyleLevel;
+  options.level0_file_num_compaction_trigger = 3;
+  std::shared_ptr<SstPartitionerFactory> factory(
+      NewSstPartitionerFixedPrefixFactory(4));
+  options.sst_partitioner_factory = factory;
+
+  DestroyAndReopen(options);
+
+  // create first file and flush to l0
+  Put("aaaa1", "A");
+  Put("bbbb1", "B");
+  Flush();
+  dbfull()->TEST_WaitForFlushMemTable();
+
+  Put("aaaa1", "A2");
+  Flush();
+  dbfull()->TEST_WaitForFlushMemTable();
+
+  // move both files down to l1
+  dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+
+  std::vector<LiveFileMetaData> files;
+  dbfull()->GetLiveFilesMetaData(&files);
+  ASSERT_EQ(2, files.size());
+  ASSERT_EQ("A2", Get("aaaa1"));
+  ASSERT_EQ("B", Get("bbbb1"));
+}
+
+TEST_F(DBCompactionTest, CompactionSstPartitionerNonTrivial) {
+  Options options = CurrentOptions();
+  options.compaction_style = kCompactionStyleLevel;
+  options.level0_file_num_compaction_trigger = 1;
+  std::shared_ptr<SstPartitionerFactory> factory(
+      NewSstPartitionerFixedPrefixFactory(4));
+  options.sst_partitioner_factory = factory;
+
+  DestroyAndReopen(options);
+
+  // create first file and flush to l0
+  Put("aaaa1", "A");
+  Put("bbbb1", "B");
+  Flush();
+  dbfull()->TEST_WaitForFlushMemTable();
+  dbfull()->TEST_WaitForCompact(true);
+
+  std::vector<LiveFileMetaData> files;
+  dbfull()->GetLiveFilesMetaData(&files);
+  ASSERT_EQ(2, files.size());
+  ASSERT_EQ("A", Get("aaaa1"));
+  ASSERT_EQ("B", Get("bbbb1"));
+}
+
 TEST_F(DBCompactionTest, ZeroSeqIdCompaction) {
  Options options = CurrentOptions();
  options.compaction_style = kCompactionStyleLevel;
@@ -1105,7 +1160,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveOneFile) {
  Random rnd(301);
  std::vector<std::string> values;
  for (int i = 0; i < num_keys; i++) {
-    values.push_back(RandomString(&rnd, value_size));
+    values.push_back(rnd.RandomString(value_size));
    ASSERT_OK(Put(Key(i), values[i]));
  }

@@ -1177,7 +1232,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveNonOverlappingFiles) {
  std::map<int32_t, std::string> values;
  for (size_t i = 0; i < ranges.size(); i++) {
    for (int32_t j = ranges[i].first; j <= ranges[i].second; j++) {
-      values[j] = RandomString(&rnd, value_size);
+      values[j] = rnd.RandomString(value_size);
      ASSERT_OK(Put(Key(j), values[j]));
    }
    ASSERT_OK(Flush());
@@ -1223,7 +1278,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveNonOverlappingFiles) {
  };
  for (size_t i = 0; i < ranges.size(); i++) {
    for (int32_t j = ranges[i].first; j <= ranges[i].second; j++) {
-      values[j] = RandomString(&rnd, value_size);
+      values[j] = rnd.RandomString(value_size);
      ASSERT_OK(Put(Key(j), values[j]));
    }
    ASSERT_OK(Flush());
@@ -1268,14 +1323,14 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveTargetLevel) {

  // file 1 [0 => 300]
  for (int32_t i = 0; i <= 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

  // file 2 [600 => 700]
  for (int32_t i = 600; i <= 700; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1349,14 +1404,14 @@ TEST_P(DBCompactionTestWithParam, ManualCompactionPartial) {

  // file 1 [0 => 100]
  for (int32_t i = 0; i < 100; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

  // file 2 [100 => 300]
  for (int32_t i = 100; i < 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1377,7 +1432,7 @@ TEST_P(DBCompactionTestWithParam, ManualCompactionPartial) {

  // file 3 [ 0 => 200]
  for (int32_t i = 0; i < 200; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1409,21 +1464,21 @@ TEST_P(DBCompactionTestWithParam, ManualCompactionPartial) {
  TEST_SYNC_POINT("DBCompaction::ManualPartial:1");
  // file 4 [300 => 400)
  for (int32_t i = 300; i <= 400; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

  // file 5 [400 => 500)
  for (int32_t i = 400; i <= 500; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

  // file 6 [500 => 600)
  for (int32_t i = 500; i <= 600; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  // Second non-trivial compaction is triggered
@@ -1491,14 +1546,14 @@ TEST_F(DBCompactionTest, DISABLED_ManualPartialFill) {

  // file 1 [0 => 100]
  for (int32_t i = 0; i < 100; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

  // file 2 [100 => 300]
  for (int32_t i = 100; i < 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1517,7 +1572,7 @@ TEST_F(DBCompactionTest, DISABLED_ManualPartialFill) {

  // file 3 [ 0 => 200]
  for (int32_t i = 0; i < 200; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1549,7 +1604,7 @@ TEST_F(DBCompactionTest, DISABLED_ManualPartialFill) {
        ASSERT_OK(Flush());
        dbfull()->TEST_WaitForFlushMemTable();
      }
-      values[j] = RandomString(&rnd, value_size);
+      values[j] = rnd.RandomString(value_size);
      ASSERT_OK(Put(Key(j), values[j]));
    }
  }
@@ -1620,14 +1675,14 @@ TEST_F(DBCompactionTest, DeleteFileRange) {

  // file 1 [0 => 100]
  for (int32_t i = 0; i < 100; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

  // file 2 [100 => 300]
  for (int32_t i = 100; i < 300; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1643,7 +1698,7 @@ TEST_F(DBCompactionTest, DeleteFileRange) {

  // file 3 [ 0 => 200]
  for (int32_t i = 0; i < 200; i++) {
-    values[i] = RandomString(&rnd, value_size);
+    values[i] = rnd.RandomString(value_size);
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1655,7 +1710,7 @@ TEST_F(DBCompactionTest, DeleteFileRange) {
        ASSERT_OK(Flush());
        dbfull()->TEST_WaitForFlushMemTable();
      }
-      values[j] = RandomString(&rnd, value_size);
+      values[j] = rnd.RandomString(value_size);
      ASSERT_OK(Put(Key(j), values[j]));
    }
  }
@@ -1742,7 +1797,7 @@ TEST_F(DBCompactionTest, DeleteFilesInRanges) {
  for (auto i = 0; i < 10; i++) {
    for (auto j = 0; j < 100; j++) {
      auto k = i * 100 + j;
-      values[k] = RandomString(&rnd, value_size);
+      values[k] = rnd.RandomString(value_size);
      ASSERT_OK(Put(Key(k), values[k]));
    }
    ASSERT_OK(Flush());
@@ -1874,7 +1929,7 @@ TEST_F(DBCompactionTest, DeleteFileRangeFileEndpointsOverlapBug) {
  // would cause `1 -> vals[0]` (an older key) to reappear.
  std::string vals[kNumL0Files];
  for (int i = 0; i < kNumL0Files; ++i) {
-    vals[i] = RandomString(&rnd, kValSize);
+    vals[i] = rnd.RandomString(kValSize);
    Put(Key(i), vals[i]);
    Put(Key(i + 1), vals[i]);
    Flush();
@@ -1916,7 +1971,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveToLastLevelWithFiles) {
  std::vector<std::string> values;
  // File with keys [ 0 => 99 ]
  for (int i = 0; i < 100; i++) {
-    values.push_back(RandomString(&rnd, value_size));
+    values.push_back(rnd.RandomString(value_size));
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -1934,7 +1989,7 @@ TEST_P(DBCompactionTestWithParam, TrivialMoveToLastLevelWithFiles) {

  // File with keys [ 100 => 199 ]
  for (int i = 100; i < 200; i++) {
-    values.push_back(RandomString(&rnd, value_size));
+    values.push_back(rnd.RandomString(value_size));
    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -2329,7 +2384,7 @@ TEST_P(DBCompactionTestWithParam, ConvertCompactionStyle) {

  for (int i = 0; i <= max_key_level_insert; i++) {
    // each value is 10K
-    ASSERT_OK(Put(1, Key(i), RandomString(&rnd, 10000)));
+    ASSERT_OK(Put(1, Key(i), rnd.RandomString(10000)));
  }
  ASSERT_OK(Flush(1));
  dbfull()->TEST_WaitForCompact();
@@ -2387,7 +2442,7 @@ TEST_P(DBCompactionTestWithParam, ConvertCompactionStyle) {
  ReopenWithColumnFamilies({"default", "pikachu"}, options);

  for (int i = max_key_level_insert / 2; i <= max_key_universal_insert; i++) {
-    ASSERT_OK(Put(1, Key(i), RandomString(&rnd, 10000)));
+    ASSERT_OK(Put(1, Key(i), rnd.RandomString(10000)));
  }
  dbfull()->Flush(FlushOptions());
  ASSERT_OK(Flush(1));
@@ -2682,7 +2737,7 @@ TEST_P(DBCompactionTestWithParam, DISABLED_CompactFilesOnLevelCompaction) {

  Random rnd(301);
  for (int key = 64 * kEntriesPerBuffer; key >= 0; --key) {
-    ASSERT_OK(Put(1, ToString(key), RandomString(&rnd, kTestValueSize)));
+    ASSERT_OK(Put(1, ToString(key), rnd.RandomString(kTestValueSize)));
  }
  dbfull()->TEST_WaitForFlushMemTable(handles_[1]);
  dbfull()->TEST_WaitForCompact();
@@ -2758,8 +2813,8 @@ TEST_P(DBCompactionTestWithParam, PartialCompactionFailure) {
  std::vector<std::string> keys;
  std::vector<std::string> values;
  for (int k = 0; k < kNumInsertedKeys; ++k) {
-    keys.emplace_back(RandomString(&rnd, kKeySize));
-    values.emplace_back(RandomString(&rnd, kKvSize - kKeySize));
+    keys.emplace_back(rnd.RandomString(kKeySize));
+    values.emplace_back(rnd.RandomString(kKvSize - kKeySize));
    ASSERT_OK(Put(Slice(keys[k]), Slice(values[k])));
    dbfull()->TEST_WaitForFlushMemTable();
  }
@@ -2825,7 +2880,7 @@ TEST_P(DBCompactionTestWithParam, DeleteMovedFileAfterCompaction) {
    for (int i = 0; i < 2; ++i) {
      // Create 1MB sst file
      for (int j = 0; j < 100; ++j) {
-        ASSERT_OK(Put(Key(i * 50 + j), RandomString(&rnd, 10 * 1024)));
+        ASSERT_OK(Put(Key(i * 50 + j), rnd.RandomString(10 * 1024)));
      }
      ASSERT_OK(Flush());
    }
@@ -2860,7 +2915,7 @@ TEST_P(DBCompactionTestWithParam, DeleteMovedFileAfterCompaction) {
    for (int i = 0; i < 2; ++i) {
      // Create 1MB sst file
      for (int j = 0; j < 100; ++j) {
-        ASSERT_OK(Put(Key(i * 50 + j + 100), RandomString(&rnd, 10 * 1024)));
+        ASSERT_OK(Put(Key(i * 50 + j + 100), rnd.RandomString(10 * 1024)));
      }
      ASSERT_OK(Flush());
    }
@@ -3118,7 +3173,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
  std::vector<std::string> values;
  // File with keys [ 0 => 99 ]
  for (int i = 0; i < 100; i++) {
-    values.push_back(RandomString(&rnd, value_size));
+    values.push_back(rnd.RandomString(value_size));
    ASSERT_OK(Put(ShortKey(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -3135,7 +3190,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {

  // File with keys [ 100 => 199 ]
  for (int i = 100; i < 200; i++) {
-    values.push_back(RandomString(&rnd, value_size));
+    values.push_back(rnd.RandomString(value_size));
    ASSERT_OK(Put(ShortKey(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -3153,7 +3208,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {

  // File with keys [ 200 => 299 ]
  for (int i = 200; i < 300; i++) {
-    values.push_back(RandomString(&rnd, value_size));
+    values.push_back(rnd.RandomString(value_size));
    ASSERT_OK(Put(ShortKey(i), values[i]));
  }
  ASSERT_OK(Flush());
@@ -3184,11 +3239,20 @@ TEST_P(DBCompactionTestWithParam, IntraL0Compaction) {
  options.level0_file_num_compaction_trigger = 5;
  options.max_background_compactions = 2;
  options.max_subcompactions = max_subcompactions_;
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
+  options.write_buffer_size = 2 << 20;  // 2MB
+
+  BlockBasedTableOptions table_options;
+  table_options.block_cache = NewLRUCache(64 << 20);  // 64MB
+  table_options.cache_index_and_filter_blocks = true;
+  table_options.pin_l0_filter_and_index_blocks_in_cache = true;
+  options.table_factory.reset(new BlockBasedTableFactory(table_options));
+
  DestroyAndReopen(options);

  const size_t kValueSize = 1 << 20;
  Random rnd(301);
-  std::string value(RandomString(&rnd, kValueSize));
+  std::string value(rnd.RandomString(kValueSize));

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"LevelCompactionPicker::PickCompactionBySize:0",
@@ -3228,6 +3292,16 @@ TEST_P(DBCompactionTestWithParam, IntraL0Compaction) {
  for (int i = 0; i < 2; ++i) {
    ASSERT_GE(level_to_files[0][i].fd.file_size, 1 << 21);
  }
+
+  // The index/filter in the file produced by intra-L0 should not be pinned.
+  // That means clearing unref'd entries in block cache and re-accessing the
+  // file produced by intra-L0 should bump the index block miss count.
+  uint64_t prev_index_misses =
+      TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS);
+  table_options.block_cache->EraseUnRefEntries();
+  ASSERT_EQ("", Get(Key(0)));
+  ASSERT_EQ(prev_index_misses + 1,
+            TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS));
 }

 TEST_P(DBCompactionTestWithParam, IntraL0CompactionDoesNotObsoleteDeletions) {
@@ -3242,7 +3316,7 @@ TEST_P(DBCompactionTestWithParam, IntraL0CompactionDoesNotObsoleteDeletions) {

  const size_t kValueSize = 1 << 20;
  Random rnd(301);
-  std::string value(RandomString(&rnd, kValueSize));
+  std::string value(rnd.RandomString(kValueSize));

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"LevelCompactionPicker::PickCompactionBySize:0",
@@ -3461,7 +3535,7 @@ TEST_F(DBCompactionTest, CompactBottomLevelFilesWithDeletions) {
  for (int i = 0; i < kNumLevelFiles; ++i) {
    for (int j = 0; j < kNumKeysPerFile; ++j) {
      ASSERT_OK(
-          Put(Key(i * kNumKeysPerFile + j), RandomString(&rnd, kValueSize)));
+          Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
    }
    if (i == kNumLevelFiles - 1) {
      snapshot = db_->GetSnapshot();
@@ -3533,7 +3607,7 @@ TEST_F(DBCompactionTest, LevelCompactExpiredTtlFiles) {
  for (int i = 0; i < kNumLevelFiles; ++i) {
    for (int j = 0; j < kNumKeysPerFile; ++j) {
      ASSERT_OK(
-          Put(Key(i * kNumKeysPerFile + j), RandomString(&rnd, kValueSize)));
+          Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
    }
    Flush();
  }
@@ -3579,7 +3653,7 @@ TEST_F(DBCompactionTest, LevelCompactExpiredTtlFiles) {
  for (int i = 0; i < kNumLevelFiles; ++i) {
    for (int j = 0; j < kNumKeysPerFile; ++j) {
      ASSERT_OK(
-          Put(Key(i * kNumKeysPerFile + j), RandomString(&rnd, kValueSize)));
+          Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
    }
    Flush();
  }
@@ -3674,7 +3748,7 @@ TEST_F(DBCompactionTest, LevelTtlCascadingCompactions) {
      // Add two L6 files with key ranges: [1 .. 100], [101 .. 200].
      Random rnd(301);
      for (int i = 1; i <= 100; ++i) {
-        ASSERT_OK(Put(Key(i), RandomString(&rnd, kValueSize)));
+        ASSERT_OK(Put(Key(i), rnd.RandomString(kValueSize)));
      }
      Flush();
      // Get the first file's creation time. This will be the oldest file in the
@@ -3687,7 +3761,7 @@ TEST_F(DBCompactionTest, LevelTtlCascadingCompactions) {
      // Add 1 hour and do another flush.
      env_->addon_time_.fetch_add(1 * 60 * 60);
      for (int i = 101; i <= 200; ++i) {
-        ASSERT_OK(Put(Key(i), RandomString(&rnd, kValueSize)));
+        ASSERT_OK(Put(Key(i), rnd.RandomString(kValueSize)));
      }
      Flush();
      MoveFilesToLevel(6);
@@ -3696,12 +3770,12 @@ TEST_F(DBCompactionTest, LevelTtlCascadingCompactions) {
      env_->addon_time_.fetch_add(1 * 60 * 60);
      // Add two L4 files with key ranges: [1 .. 50], [51 .. 150].
      for (int i = 1; i <= 50; ++i) {
-        ASSERT_OK(Put(Key(i), RandomString(&rnd, kValueSize)));
+        ASSERT_OK(Put(Key(i), rnd.RandomString(kValueSize)));
      }
      Flush();
      env_->addon_time_.fetch_add(1 * 60 * 60);
      for (int i = 51; i <= 150; ++i) {
-        ASSERT_OK(Put(Key(i), RandomString(&rnd, kValueSize)));
+        ASSERT_OK(Put(Key(i), rnd.RandomString(kValueSize)));
      }
      Flush();
      MoveFilesToLevel(4);
@@ -3710,7 +3784,7 @@ TEST_F(DBCompactionTest, LevelTtlCascadingCompactions) {
      env_->addon_time_.fetch_add(1 * 60 * 60);
      // Add one L1 file with key range: [26, 75].
      for (int i = 26; i <= 75; ++i) {
-        ASSERT_OK(Put(Key(i), RandomString(&rnd, kValueSize)));
+        ASSERT_OK(Put(Key(i), rnd.RandomString(kValueSize)));
      }
      Flush();
      dbfull()->TEST_WaitForCompact();
@@ -3821,8 +3895,8 @@ TEST_F(DBCompactionTest, LevelPeriodicCompaction) {
      Random rnd(301);
      for (int i = 0; i < kNumLevelFiles; ++i) {
        for (int j = 0; j < kNumKeysPerFile; ++j) {
-          ASSERT_OK(Put(Key(i * kNumKeysPerFile + j),
-                        RandomString(&rnd, kValueSize)));
+          ASSERT_OK(
+              Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
        }
        Flush();
      }
@@ -3916,7 +3990,7 @@ TEST_F(DBCompactionTest, LevelPeriodicCompactionWithOldDB) {
  for (int i = 0; i < kNumFiles; ++i) {
    for (int j = 0; j < kNumKeysPerFile; ++j) {
      ASSERT_OK(
-          Put(Key(i * kNumKeysPerFile + j), RandomString(&rnd, kValueSize)));
+          Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
    }
    Flush();
    // Move the first two files to L2.
@@ -3979,7 +4053,7 @@ TEST_F(DBCompactionTest, LevelPeriodicAndTtlCompaction) {
  for (int i = 0; i < kNumLevelFiles; ++i) {
    for (int j = 0; j < kNumKeysPerFile; ++j) {
      ASSERT_OK(
-          Put(Key(i * kNumKeysPerFile + j), RandomString(&rnd, kValueSize)));
+          Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
    }
    Flush();
  }
@@ -4090,7 +4164,7 @@ TEST_F(DBCompactionTest, LevelPeriodicCompactionWithCompactionFilters) {
    for (int i = 0; i < kNumLevelFiles; ++i) {
      for (int j = 0; j < kNumKeysPerFile; ++j) {
        ASSERT_OK(
-            Put(Key(i * kNumKeysPerFile + j), RandomString(&rnd, kValueSize)));
+            Put(Key(i * kNumKeysPerFile + j), rnd.RandomString(kValueSize)));
      }
      Flush();
    }
@@ -4150,7 +4224,7 @@ TEST_F(DBCompactionTest, CompactRangeDelayedByL0FileCount) {
    Random rnd(301);
    for (int j = 0; j < kNumL0FilesLimit - 1; ++j) {
      for (int k = 0; k < 2; ++k) {
-        ASSERT_OK(Put(Key(k), RandomString(&rnd, 1024)));
+        ASSERT_OK(Put(Key(k), rnd.RandomString(1024)));
      }
      Flush();
    }
@@ -4204,7 +4278,7 @@ TEST_F(DBCompactionTest, CompactRangeDelayedByImmMemTableCount) {

    Random rnd(301);
    for (int j = 0; j < kNumImmMemTableLimit - 1; ++j) {
-      ASSERT_OK(Put(Key(0), RandomString(&rnd, 1024)));
+      ASSERT_OK(Put(Key(0), rnd.RandomString(1024)));
      FlushOptions flush_opts;
      flush_opts.wait = false;
      flush_opts.allow_write_stall = true;
@@ -4252,7 +4326,7 @@ TEST_F(DBCompactionTest, CompactRangeShutdownWhileDelayed) {
    Random rnd(301);
    for (int j = 0; j < kNumL0FilesLimit - 1; ++j) {
      for (int k = 0; k < 2; ++k) {
-        ASSERT_OK(Put(1, Key(k), RandomString(&rnd, 1024)));
+        ASSERT_OK(Put(1, Key(k), rnd.RandomString(1024)));
      }
      Flush(1);
    }
@@ -4312,7 +4386,7 @@ TEST_F(DBCompactionTest, CompactRangeSkipFlushAfterDelay) {
  flush_opts.allow_write_stall = true;
  for (int i = 0; i < kNumL0FilesLimit - 1; ++i) {
    for (int j = 0; j < 2; ++j) {
-      ASSERT_OK(Put(Key(j), RandomString(&rnd, 1024)));
+      ASSERT_OK(Put(Key(j), rnd.RandomString(1024)));
    }
    dbfull()->Flush(flush_opts);
  }
@@ -4323,9 +4397,9 @@ TEST_F(DBCompactionTest, CompactRangeSkipFlushAfterDelay) {
  });

  TEST_SYNC_POINT("DBCompactionTest::CompactRangeSkipFlushAfterDelay:PreFlush");
-  Put(ToString(0), RandomString(&rnd, 1024));
+  Put(ToString(0), rnd.RandomString(1024));
  dbfull()->Flush(flush_opts);
-  Put(ToString(0), RandomString(&rnd, 1024));
+  Put(ToString(0), rnd.RandomString(1024));
  TEST_SYNC_POINT("DBCompactionTest::CompactRangeSkipFlushAfterDelay:PostFlush");
  manual_compaction_thread.join();

@@ -4765,7 +4839,7 @@ TEST_P(CompactionPriTest, Test) {
  RandomShuffle(std::begin(keys), std::end(keys), rnd.Next());

  for (int i = 0; i < kNKeys; i++) {
-    ASSERT_OK(Put(Key(keys[i]), RandomString(&rnd, 102)));
+    ASSERT_OK(Put(Key(keys[i]), rnd.RandomString(102)));
  }

  dbfull()->TEST_WaitForCompact();
@@ -4807,7 +4881,7 @@ TEST_F(DBCompactionTest, PartialManualCompaction) {
  Random rnd(301);
  for (auto i = 0; i < 8; ++i) {
    for (auto j = 0; j < 10; ++j) {
-      Merge("foo", RandomString(&rnd, 1024));
+      Merge("foo", rnd.RandomString(1024));
    }
    Flush();
  }
@@ -4839,8 +4913,8 @@ TEST_F(DBCompactionTest, ManualCompactionFailsInReadOnlyMode) {
  Random rnd(301);
  for (int i = 0; i < kNumL0Files; ++i) {
    // Make sure files are overlapping in key-range to prevent trivial move.
-    Put("key1", RandomString(&rnd, 1024));
-    Put("key2", RandomString(&rnd, 1024));
+    Put("key1", rnd.RandomString(1024));
+    Put("key2", rnd.RandomString(1024));
    Flush();
  }
  ASSERT_EQ(kNumL0Files, NumTableFilesAtLevel(0));
@@ -4849,7 +4923,7 @@ TEST_F(DBCompactionTest, ManualCompactionFailsInReadOnlyMode) {
  mock_env->SetFilesystemActive(false);
  // Make sure this is outside `CompactRange`'s range so that it doesn't fail
  // early trying to flush memtable.
-  ASSERT_NOK(Put("key3", RandomString(&rnd, 1024)));
+  ASSERT_NOK(Put("key3", rnd.RandomString(1024)));

  // In the bug scenario, the first manual compaction would fail and forget to
  // unregister itself, causing the second one to hang forever due to conflict
@@ -4888,7 +4962,7 @@ TEST_F(DBCompactionTest, ManualCompactionBottomLevelOptimized) {
  for (auto i = 0; i < 8; ++i) {
    for (auto j = 0; j < 10; ++j) {
      ASSERT_OK(
-          Put("foo" + std::to_string(i * 10 + j), RandomString(&rnd, 1024)));
+          Put("foo" + std::to_string(i * 10 + j), rnd.RandomString(1024)));
    }
    Flush();
  }
@@ -4898,7 +4972,7 @@ TEST_F(DBCompactionTest, ManualCompactionBottomLevelOptimized) {
  for (auto i = 0; i < 8; ++i) {
    for (auto j = 0; j < 10; ++j) {
      ASSERT_OK(
-          Put("bar" + std::to_string(i * 10 + j), RandomString(&rnd, 1024)));
+          Put("bar" + std::to_string(i * 10 + j), rnd.RandomString(1024)));
    }
    Flush();
  }
@@ -4932,7 +5006,7 @@ TEST_F(DBCompactionTest, CompactionDuringShutdown) {
  for (auto i = 0; i < 2; ++i) {
    for (auto j = 0; j < 10; ++j) {
      ASSERT_OK(
-          Put("foo" + std::to_string(i * 10 + j), RandomString(&rnd, 1024)));
+          Put("foo" + std::to_string(i * 10 + j), rnd.RandomString(1024)));
    }
    Flush();
  }
@@ -4955,7 +5029,7 @@ TEST_P(DBCompactionTestWithParam, FixFileIngestionCompactionDeadlock) {

  // Generate an external SST file containing a single key, i.e. 99
  std::string sst_files_dir = dbname_ + "/sst_files/";
-  test::DestroyDir(env_, sst_files_dir);
+  DestroyDir(env_, sst_files_dir);
  ASSERT_OK(env_->CreateDir(sst_files_dir));
  SstFileWriter sst_writer(EnvOptions(), options);
  const std::string sst_file_path = sst_files_dir + "test.sst";
@@ -4982,7 +5056,7 @@ TEST_P(DBCompactionTestWithParam, FixFileIngestionCompactionDeadlock) {
  // Generate level0_stop_writes_trigger L0 files to trigger write stop
  for (int i = 0; i != options.level0_file_num_compaction_trigger; ++i) {
    for (int j = 0; j != kNumKeysPerFile; ++j) {
-      ASSERT_OK(Put(Key(j), RandomString(&rnd, 990)));
+      ASSERT_OK(Put(Key(j), rnd.RandomString(990)));
    }
    if (0 == i) {
      // When we reach here, the memtables have kNumKeysPerFile keys. Note that
@@ -5074,7 +5148,7 @@ TEST_F(DBCompactionTest, ConsistencyFailTest2) {
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  Random rnd(301);
-  std::string value = RandomString(&rnd, 1000);
+  std::string value = rnd.RandomString(1000);

  ASSERT_OK(Put("foo1", value));
  ASSERT_OK(Put("z", ""));
@@ -5121,7 +5195,7 @@ TEST_P(DBCompactionTestWithParam,
  const size_t kValueSize = 1 << 20;
  Random rnd(301);
  std::atomic<int> pick_intra_l0_count(0);
-  std::string value(RandomString(&rnd, kValueSize));
+  std::string value(rnd.RandomString(kValueSize));

  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"DBCompactionTestWithParam::FlushAfterIntraL0:1",
@@ -5163,7 +5237,6 @@ TEST_P(DBCompactionTestWithParam,
  // Put one key, to make biggest log sequence number in this memtable is bigger
  // than sst which would be ingested in next step.
  ASSERT_OK(Put(Key(2), "b"));
-  ASSERT_EQ(10, NumTableFilesAtLevel(0));
  dbfull()->TEST_WaitForCompact();
  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  std::vector<std::vector<FileMetaData>> level_to_files;
@@ -5189,8 +5262,8 @@ TEST_P(DBCompactionTestWithParam,

  const size_t kValueSize = 1 << 20;
  Random rnd(301);
-  std::string value(RandomString(&rnd, kValueSize));
-  std::string value2(RandomString(&rnd, kValueSize));
+  std::string value(rnd.RandomString(kValueSize));
+  std::string value2(rnd.RandomString(kValueSize));
  std::string bigvalue = value + value;

  // prevents trivial move
@@ -5260,6 +5333,108 @@ TEST_P(DBCompactionTestWithParam,
  }
 }

+TEST_F(DBCompactionTest, UpdateLevelSubCompactionTest) {
+  Options options = CurrentOptions();
+  options.max_subcompactions = 10;
+  options.target_file_size_base = 1 << 10;  // 1KB
+  DestroyAndReopen(options);
+
+  bool has_compaction = false;
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "LevelCompactionPicker::PickCompaction:Return", [&](void* arg) {
+        Compaction* compaction = reinterpret_cast<Compaction*>(arg);
+        ASSERT_TRUE(compaction->max_subcompactions() == 10);
+        has_compaction = true;
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  ASSERT_TRUE(dbfull()->GetDBOptions().max_subcompactions == 10);
+  // Trigger compaction
+  for (int i = 0; i < 32; i++) {
+    for (int j = 0; j < 5000; j++) {
+      Put(std::to_string(j), std::string(1, 'A'));
+    }
+    ASSERT_OK(Flush());
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+  }
+  dbfull()->TEST_WaitForCompact();
+  ASSERT_TRUE(has_compaction);
+
+  has_compaction = false;
+  ASSERT_OK(dbfull()->SetDBOptions({{"max_subcompactions", "2"}}));
+  ASSERT_TRUE(dbfull()->GetDBOptions().max_subcompactions == 2);
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "LevelCompactionPicker::PickCompaction:Return", [&](void* arg) {
+        Compaction* compaction = reinterpret_cast<Compaction*>(arg);
+        ASSERT_TRUE(compaction->max_subcompactions() == 2);
+        has_compaction = true;
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  // Trigger compaction
+  for (int i = 0; i < 32; i++) {
+    for (int j = 0; j < 5000; j++) {
+      Put(std::to_string(j), std::string(1, 'A'));
+    }
+    ASSERT_OK(Flush());
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+  }
+  dbfull()->TEST_WaitForCompact();
+  ASSERT_TRUE(has_compaction);
+}
+
+TEST_F(DBCompactionTest, UpdateUniversalSubCompactionTest) {
+  Options options = CurrentOptions();
+  options.max_subcompactions = 10;
+  options.compaction_style = kCompactionStyleUniversal;
+  options.target_file_size_base = 1 << 10;  // 1KB
+  DestroyAndReopen(options);
+
+  bool has_compaction = false;
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "UniversalCompactionBuilder::PickCompaction:Return", [&](void* arg) {
+        Compaction* compaction = reinterpret_cast<Compaction*>(arg);
+        ASSERT_TRUE(compaction->max_subcompactions() == 10);
+        has_compaction = true;
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  // Trigger compaction
+  for (int i = 0; i < 32; i++) {
+    for (int j = 0; j < 5000; j++) {
+      Put(std::to_string(j), std::string(1, 'A'));
+    }
+    ASSERT_OK(Flush());
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+  }
+  dbfull()->TEST_WaitForCompact();
+  ASSERT_TRUE(has_compaction);
+  has_compaction = false;
+
+  ASSERT_OK(dbfull()->SetDBOptions({{"max_subcompactions", "2"}}));
+  ASSERT_TRUE(dbfull()->GetDBOptions().max_subcompactions == 2);
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "UniversalCompactionBuilder::PickCompaction:Return", [&](void* arg) {
+        Compaction* compaction = reinterpret_cast<Compaction*>(arg);
+        ASSERT_TRUE(compaction->max_subcompactions() == 2);
+        has_compaction = true;
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  // Trigger compaction
+  for (int i = 0; i < 32; i++) {
+    for (int j = 0; j < 5000; j++) {
+      Put(std::to_string(j), std::string(1, 'A'));
+    }
+    ASSERT_OK(Flush());
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+  }
+  dbfull()->TEST_WaitForCompact();
+  ASSERT_TRUE(has_compaction);
+}
+
 #endif // !defined(ROCKSDB_LITE)
 }  // namespace ROCKSDB_NAMESPACE

@@ -15,6 +15,7 @@
 #include "db/db_test_util.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
+#include "util/random.h"

 namespace ROCKSDB_NAMESPACE {
 class DBTestDynamicLevel : public DBTestBase {
@@ -80,9 +81,9 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase) {

      for (int i = 0; i < kNKeys; i++) {
        int key = keys[i];
-        ASSERT_OK(Put(Key(kNKeys + key), RandomString(&rnd, 102)));
-        ASSERT_OK(Put(Key(key), RandomString(&rnd, 102)));
-        ASSERT_OK(Put(Key(kNKeys * 2 + key), RandomString(&rnd, 102)));
+        ASSERT_OK(Put(Key(kNKeys + key), rnd.RandomString(102)));
+        ASSERT_OK(Put(Key(key), rnd.RandomString(102)));
+        ASSERT_OK(Put(Key(kNKeys * 2 + key), rnd.RandomString(102)));
        ASSERT_OK(Delete(Key(kNKeys + keys[i / 10])));
        env_->SleepForMicroseconds(5000);
      }
@@ -158,7 +159,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  // Put about 28K to L0
  for (int i = 0; i < 70; i++) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 380)));
+                  rnd.RandomString(380)));
  }
  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "false"},
@@ -175,7 +176,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  }));
  for (int i = 0; i < 70; i++) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 380)));
+                  rnd.RandomString(380)));
  }

  ASSERT_OK(dbfull()->SetOptions({
@@ -197,7 +198,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  // Write about 40K more
  for (int i = 0; i < 100; i++) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 380)));
+                  rnd.RandomString(380)));
  }
  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "false"},
@@ -216,7 +217,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  // Each file is about 11KB, with 9KB of data.
  for (int i = 0; i < 1300; i++) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 380)));
+                  rnd.RandomString(380)));
  }

  // Make sure that the compaction starts before the last bit of data is
@@ -257,7 +258,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:1");
  for (int i = 0; i < 2; i++) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 380)));
+                  rnd.RandomString(380)));
  }
  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:2");

@@ -310,15 +311,15 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesCompactRange) {

  // Put about 7K to L0
  for (int i = 0; i < 140; i++) {
-    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 80)));
+    ASSERT_OK(
+        Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))), rnd.RandomString(80)));
  }
  Flush();
  dbfull()->TEST_WaitForCompact();
  if (NumTableFilesAtLevel(0) == 0) {
    // Make sure level 0 is not empty
-    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
-                  RandomString(&rnd, 80)));
+    ASSERT_OK(
+        Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))), rnd.RandomString(80)));
    Flush();
  }

@@ -382,7 +383,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBaseInc) {
  const int total_keys = 3000;
  const int random_part_size = 100;
  for (int i = 0; i < total_keys; i++) {
-    std::string value = RandomString(&rnd, random_part_size);
+    std::string value = rnd.RandomString(random_part_size);
    PutFixed32(&value, static_cast<uint32_t>(i));
    ASSERT_OK(Put(Key(i), value));
  }
@@ -441,8 +442,8 @@ TEST_F(DBTestDynamicLevel, DISABLED_MigrateToDynamicLevelMaxBytesBase) {

  int total_keys = 1000;
  for (int i = 0; i < total_keys; i++) {
-    ASSERT_OK(Put(Key(i), RandomString(&rnd, 102)));
-    ASSERT_OK(Put(Key(kMaxKey + i), RandomString(&rnd, 102)));
+    ASSERT_OK(Put(Key(i), rnd.RandomString(102)));
+    ASSERT_OK(Put(Key(kMaxKey + i), rnd.RandomString(102)));
    ASSERT_OK(Delete(Key(i / 10)));
  }
  verify_func(total_keys, false);
@@ -475,8 +476,8 @@ TEST_F(DBTestDynamicLevel, DISABLED_MigrateToDynamicLevelMaxBytesBase) {

  int total_keys2 = 2000;
  for (int i = total_keys; i < total_keys2; i++) {
-    ASSERT_OK(Put(Key(i), RandomString(&rnd, 102)));
-    ASSERT_OK(Put(Key(kMaxKey + i), RandomString(&rnd, 102)));
+    ASSERT_OK(Put(Key(i), rnd.RandomString(102)));
+    ASSERT_OK(Put(Key(kMaxKey + i), rnd.RandomString(102)));
    ASSERT_OK(Delete(Key(i / 10)));
  }

@@ -23,57 +23,6 @@

 namespace ROCKSDB_NAMESPACE {

-Status DBImpl::DisableFileDeletions() {
-  InstrumentedMutexLock l(&mutex_);
-  ++disable_delete_obsolete_files_;
-  if (disable_delete_obsolete_files_ == 1) {
-    ROCKS_LOG_INFO(immutable_db_options_.info_log, "File Deletions Disabled");
-  } else {
-    ROCKS_LOG_WARN(immutable_db_options_.info_log,
-                   "File Deletions Disabled, but already disabled. Counter: %d",
-                   disable_delete_obsolete_files_);
-  }
-  return Status::OK();
-}
-
-Status DBImpl::EnableFileDeletions(bool force) {
-  // Job id == 0 means that this is not our background process, but rather
-  // user thread
-  JobContext job_context(0);
-  bool file_deletion_enabled = false;
-  {
-    InstrumentedMutexLock l(&mutex_);
-    if (force) {
-      // if force, we need to enable file deletions right away
-      disable_delete_obsolete_files_ = 0;
-    } else if (disable_delete_obsolete_files_ > 0) {
-      --disable_delete_obsolete_files_;
-    }
-    if (disable_delete_obsolete_files_ == 0)  {
-      file_deletion_enabled = true;
-      FindObsoleteFiles(&job_context, true);
-      bg_cv_.SignalAll();
-    }
-  }
-  if (file_deletion_enabled) {
-    ROCKS_LOG_INFO(immutable_db_options_.info_log, "File Deletions Enabled");
-    if (job_context.HaveSomethingToDelete()) {
-      PurgeObsoleteFiles(job_context);
-    }
-  } else {
-    ROCKS_LOG_WARN(immutable_db_options_.info_log,
-                   "File Deletions Enable, but not really enabled. Counter: %d",
-                   disable_delete_obsolete_files_);
-  }
-  job_context.Clean();
-  LogFlush(immutable_db_options_.info_log);
-  return Status::OK();
-}
-
-int DBImpl::IsFileDeletionsEnabled() const {
-  return !disable_delete_obsolete_files_;
-}
-
 Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
                            uint64_t* manifest_file_size,
                            bool flush_memtable) {
@@ -13,10 +13,10 @@
 #include "db/db_test_util.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
-#include "test_util/fault_injection_test_env.h"
 #include "test_util/sync_point.h"
 #include "util/cast_util.h"
 #include "util/mutexlock.h"
+#include "utilities/fault_injection_env.h"

 namespace ROCKSDB_NAMESPACE {

@@ -89,7 +89,7 @@ TEST_F(DBFlushTest, SyncFail) {
  CreateAndReopenWithCF({"pikachu"}, options);
  Put("key", "value");
  auto* cfd =
-      reinterpret_cast<ColumnFamilyHandleImpl*>(db_->DefaultColumnFamily())
+      static_cast_with_check<ColumnFamilyHandleImpl>(db_->DefaultColumnFamily())
          ->cfd();
  FlushOptions flush_options;
  flush_options.wait = false;
@@ -379,9 +379,9 @@ TEST_F(DBFlushTest, FireOnFlushCompletedAfterCommittedResult) {
      DBImpl* db_impl = static_cast_with_check<DBImpl>(db);
      InstrumentedMutex* mutex = db_impl->mutex();
      mutex->Lock();
-      auto* cfd =
-          reinterpret_cast<ColumnFamilyHandleImpl*>(db->DefaultColumnFamily())
-              ->cfd();
+      auto* cfd = static_cast_with_check<ColumnFamilyHandleImpl>(
+                      db->DefaultColumnFamily())
+                      ->cfd();
      ASSERT_LT(seq, cfd->imm()->current()->GetEarliestSequenceNumber());
      mutex->Unlock();
    }
@@ -499,13 +499,13 @@ TEST_P(DBAtomicFlushTest, AtomicFlushTriggeredByMemTableFull) {
  TEST_SYNC_POINT(
      "DBAtomicFlushTest::AtomicFlushTriggeredByMemTableFull:BeforeCheck");
  if (options.atomic_flush) {
-    for (size_t i = 0; i != num_cfs - 1; ++i) {
+    for (size_t i = 0; i + 1 != num_cfs; ++i) {
      auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
      ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
      ASSERT_TRUE(cfh->cfd()->mem()->IsEmpty());
    }
  } else {
-    for (size_t i = 0; i != num_cfs - 1; ++i) {
+    for (size_t i = 0; i + 1 != num_cfs; ++i) {
      auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
      ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
      ASSERT_FALSE(cfh->cfd()->mem()->IsEmpty());
@@ -87,10 +87,10 @@
 #include "table/get_context.h"
 #include "table/merging_iterator.h"
 #include "table/multiget_context.h"
+#include "table/sst_file_dumper.h"
 #include "table/table_builder.h"
 #include "table/two_level_iterator.h"
 #include "test_util/sync_point.h"
-#include "tools/sst_dump_tool_imp.h"
 #include "util/autovector.h"
 #include "util/build_version.h"
 #include "util/cast_util.h"
@@ -249,7 +249,8 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
      new ColumnFamilyMemTablesImpl(versions_->GetColumnFamilySet()));

  DumpRocksDBBuildVersion(immutable_db_options_.info_log.get());
-  DumpDBFileSummary(immutable_db_options_, dbname_);
+  SetDbSessionId();
+  DumpDBFileSummary(immutable_db_options_, dbname_, db_session_id_);
  immutable_db_options_.Dump(immutable_db_options_.info_log.get());
  mutable_db_options_.Dump(immutable_db_options_.info_log.get());
  DumpSupportInfo(immutable_db_options_.info_log.get());
@@ -312,8 +313,37 @@ Status DBImpl::ResumeImpl() {
  }

  // Make sure the IO Status stored in version set is set to OK.
+  bool file_deletion_disabled = !IsFileDeletionsEnabled();
  if (s.ok()) {
-    versions_->SetIOStatusOK();
+    IOStatus io_s = versions_->io_status();
+    if (io_s.IsIOError()) {
+      // If resuming from IOError resulted from MANIFEST write, then assert
+      // that we must have already set the MANIFEST writer to nullptr during
+      // clean-up phase MANIFEST writing. We must have also disabled file
+      // deletions.
+      assert(!versions_->descriptor_log_);
+      assert(file_deletion_disabled);
+      // Since we are trying to recover from MANIFEST write error, we need to
+      // switch to a new MANIFEST anyway. The old MANIFEST can be corrupted.
+      // Therefore, force writing a dummy version edit because we do not know
+      // whether there are flush jobs with non-empty data to flush, triggering
+      // appends to MANIFEST.
+      VersionEdit edit;
+      auto cfh =
+          static_cast_with_check<ColumnFamilyHandleImpl>(default_cf_handle_);
+      assert(cfh);
+      ColumnFamilyData* cfd = cfh->cfd();
+      const MutableCFOptions& cf_opts = *cfd->GetLatestMutableCFOptions();
+      s = versions_->LogAndApply(cfd, cf_opts, &edit, &mutex_,
+                                 directories_.GetDbDir());
+      if (!s.ok()) {
+        io_s = versions_->io_status();
+        if (!io_s.ok()) {
+          s = error_handler_.SetBGError(io_s,
+                                        BackgroundErrorReason::kManifestWrite);
+        }
+      }
+    }
  }

  // We cannot guarantee consistency of the WAL. So force flush Memtables of
@@ -364,6 +394,13 @@ Status DBImpl::ResumeImpl() {
  job_context.Clean();

  if (s.ok()) {
+    assert(versions_->io_status().ok());
+    // If we reach here, we should re-enable file deletions if it was disabled
+    // during previous error handling.
+    if (file_deletion_disabled) {
+      // Always return ok
+      EnableFileDeletions(/*force=*/true);
+    }
    ROCKS_LOG_INFO(immutable_db_options_.info_log, "Successfully resumed DB");
  }
  mutex_.Lock();
@@ -648,9 +685,17 @@ void DBImpl::StartTimedTasks() {
    stats_dump_period_sec = mutable_db_options_.stats_dump_period_sec;
    if (stats_dump_period_sec > 0) {
      if (!thread_dump_stats_) {
+        // In case of many `DB::Open()` in rapid succession we can have all
+        // threads dumping at once, which causes severe lock contention in
+        // jemalloc. Ensure successive `DB::Open()`s are staggered by at least
+        // one second in the common case.
+        static std::atomic<uint64_t> stats_dump_threads_started(0);
        thread_dump_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
            [this]() { DBImpl::DumpStats(); }, "dump_st", env_,
-            static_cast<uint64_t>(stats_dump_period_sec) * kMicrosInSecond));
+            static_cast<uint64_t>(stats_dump_period_sec) * kMicrosInSecond,
+            stats_dump_threads_started.fetch_add(1) %
+                static_cast<uint64_t>(stats_dump_period_sec) *
+                kMicrosInSecond));
      }
    }
    stats_persist_period_sec = mutable_db_options_.stats_persist_period_sec;
@@ -909,7 +954,8 @@ Status DBImpl::SetOptions(
  (void)options_map;
  return Status::NotSupported("Not supported in ROCKSDB LITE");
 #else
-  auto* cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+  auto* cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
  if (options_map.empty()) {
    ROCKS_LOG_WARN(immutable_db_options_.info_log,
                   "SetOptions() on column family [%s], empty input",
@@ -1045,9 +1091,18 @@ Status DBImpl::SetDBOptions(
          mutex_.Lock();
        }
        if (new_options.stats_dump_period_sec > 0) {
+          // In case many DBs have `stats_dump_period_sec` enabled in rapid
+          // succession, we can have all threads dumping at once, which causes
+          // severe lock contention in jemalloc. Ensure successive enabling of
+          // `stats_dump_period_sec` are staggered by at least one second in the
+          // common case.
+          static std::atomic<uint64_t> stats_dump_threads_started(0);
          thread_dump_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
              [this]() { DBImpl::DumpStats(); }, "dump_st", env_,
              static_cast<uint64_t>(new_options.stats_dump_period_sec) *
+                  kMicrosInSecond,
+              stats_dump_threads_started.fetch_add(1) %
+                  static_cast<uint64_t>(new_options.stats_dump_period_sec) *
                  kMicrosInSecond));
        } else {
          thread_dump_stats_.reset();
@@ -1316,7 +1371,7 @@ InternalIterator* DBImpl::NewInternalIterator(
  if (column_family == nullptr) {
    cfd = default_cf_handle_->cfd();
  } else {
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+    auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
    cfd = cfh->cfd();
  }

@@ -1545,8 +1600,8 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
  StopWatch sw(env_, stats_, DB_GET);
  PERF_TIMER_GUARD(get_snapshot_time);

-  auto cfh =
-      reinterpret_cast<ColumnFamilyHandleImpl*>(get_impl_options.column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(
+      get_impl_options.column_family);
  auto cfd = cfh->cfd();

  if (tracer_) {
@@ -1747,7 +1802,7 @@ std::vector<Status> DBImpl::MultiGet(
  std::unordered_map<uint32_t, MultiGetColumnFamilyData> multiget_cf_data(
      column_family.size());
  for (auto cf : column_family) {
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(cf);
+    auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(cf);
    auto cfd = cfh->cfd();
    if (multiget_cf_data.find(cfd->GetID()) == multiget_cf_data.end()) {
      multiget_cf_data.emplace(cfd->GetID(),
@@ -1787,6 +1842,7 @@ std::vector<Status> DBImpl::MultiGet(
  // merge_operands will contain the sequence of merges in the latter case.
  size_t num_found = 0;
  size_t keys_read;
+  uint64_t curr_value_size = 0;
  for (keys_read = 0; keys_read < num_keys; ++keys_read) {
    merge_context.Clear();
    Status& s = stat_list[keys_read];
@@ -1830,6 +1886,13 @@ std::vector<Status> DBImpl::MultiGet(
    if (s.ok()) {
      bytes_read += value->size();
      num_found++;
+      curr_value_size += value->size();
+      if (curr_value_size > read_options.value_size_soft_limit) {
+        while (++keys_read < num_keys) {
+          stat_list[keys_read] = Status::Aborted();
+        }
+        break;
+      }
    }

    if (read_options.deadline.count() &&
@@ -2084,11 +2147,11 @@ void DBImpl::MultiGet(const ReadOptions& read_options, const size_t num_keys,
    }
  }
  if (!s.ok()) {
-    assert(s.IsTimedOut());
+    assert(s.IsTimedOut() || s.IsAborted());
    for (++cf_iter; cf_iter != multiget_cf_data.end(); ++cf_iter) {
      for (size_t i = cf_iter->start; i < cf_iter->start + cf_iter->num_keys;
           ++i) {
-        *sorted_keys[i]->s = Status::TimedOut();
+        *sorted_keys[i]->s = s;
      }
    }
  }
@@ -2141,10 +2204,11 @@ void DBImpl::PrepareMultiGetKeys(
        KeyContext* lhs = (*sorted_keys)[index - 1];
        KeyContext* rhs = (*sorted_keys)[index];
        ColumnFamilyHandleImpl* cfh =
-            reinterpret_cast<ColumnFamilyHandleImpl*>(lhs->column_family);
+            static_cast_with_check<ColumnFamilyHandleImpl>(lhs->column_family);
        uint32_t cfd_id1 = cfh->cfd()->GetID();
        const Comparator* comparator = cfh->cfd()->user_comparator();
-        cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(lhs->column_family);
+        cfh =
+            static_cast_with_check<ColumnFamilyHandleImpl>(lhs->column_family);
        uint32_t cfd_id2 = cfh->cfd()->GetID();

        assert(cfd_id1 <= cfd_id2);
@@ -2243,7 +2307,7 @@ void DBImpl::MultiGetWithCallback(
  Status s = MultiGetImpl(read_options, 0, num_keys, sorted_keys,
                          multiget_cf_data[0].super_version, consistent_seqnum,
                          nullptr, nullptr);
-  assert(s.ok() || s.IsTimedOut());
+  assert(s.ok() || s.IsTimedOut() || s.IsAborted());
  ReturnAndCleanupSuperVersion(multiget_cf_data[0].cfd,
                               multiget_cf_data[0].super_version);
 }
@@ -2271,6 +2335,7 @@ Status DBImpl::MultiGetImpl(
  // merge_operands will contain the sequence of merges in the latter case.
  size_t keys_left = num_keys;
  Status s;
+  uint64_t curr_value_size = 0;
  while (keys_left) {
    if (read_options.deadline.count() &&
        env_->NowMicros() >
@@ -2285,6 +2350,7 @@ Status DBImpl::MultiGetImpl(
    MultiGetContext ctx(sorted_keys, start_key + num_keys - keys_left,
                        batch_size, snapshot, read_options);
    MultiGetRange range = ctx.GetMultiGetRange();
+    range.AddValueSize(curr_value_size);
    bool lookup_current = false;

    keys_left -= batch_size;
@@ -2315,6 +2381,11 @@ Status DBImpl::MultiGetImpl(
      super_version->current->MultiGet(read_options, &range, callback,
                                       is_blob_index);
    }
+    curr_value_size = range.GetValueSize();
+    if (curr_value_size > read_options.value_size_soft_limit) {
+      s = Status::Aborted();
+      break;
+    }
  }

  // Post processing (decrement reference counts and record statistics)
@@ -2329,11 +2400,11 @@ Status DBImpl::MultiGetImpl(
    }
  }
  if (keys_left) {
-    assert(s.IsTimedOut());
+    assert(s.IsTimedOut() || s.IsAborted());
    for (size_t i = start_key + num_keys - keys_left; i < start_key + num_keys;
         ++i) {
      KeyContext* key = (*sorted_keys)[i];
-      *key->s = Status::TimedOut();
+      *key->s = s;
    }
  }

@@ -2501,7 +2572,7 @@ Status DBImpl::CreateColumnFamilyImpl(const ColumnFamilyOptions& cf_options,
  // this is outside the mutex
  if (s.ok()) {
    NewThreadStatusCfInfo(
-        reinterpret_cast<ColumnFamilyHandleImpl*>(*handle)->cfd());
+        static_cast_with_check<ColumnFamilyHandleImpl>(*handle)->cfd());
  }
  return s;
 }
@@ -2538,7 +2609,7 @@ Status DBImpl::DropColumnFamilies(
 }

 Status DBImpl::DropColumnFamilyImpl(ColumnFamilyHandle* column_family) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  if (cfd->GetID() == 0) {
    return Status::InvalidArgument("Can't drop default column family");
@@ -2634,8 +2705,6 @@ Iterator* DBImpl::NewIterator(const ReadOptions& read_options,
    return NewErrorIterator(
        Status::NotSupported("Managed iterator is not supported anymore."));
  }
-  // We will eventually support deadline for iterators too, but safeguard
-  // for now
  if (read_options.deadline != std::chrono::microseconds::zero()) {
    return NewErrorIterator(
        Status::NotSupported("ReadOptions deadline is not supported"));
@@ -2654,8 +2723,9 @@ Iterator* DBImpl::NewIterator(const ReadOptions& read_options,
        "Iterator requested internal keys which are too old and are not"
        " guaranteed to be preserved, try larger iter_start_seqnum opt."));
  }
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
-  auto cfd = cfh->cfd();
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
+  ColumnFamilyData* cfd = cfh->cfd();
+  assert(cfd != nullptr);
  ReadCallback* read_callback = nullptr;  // No read callback provided.
  if (read_options.tailing) {
 #ifdef ROCKSDB_LITE
@@ -2676,10 +2746,11 @@ Iterator* DBImpl::NewIterator(const ReadOptions& read_options,
    // Note: no need to consider the special case of
    // last_seq_same_as_publish_seq_==false since NewIterator is overridden in
    // WritePreparedTxnDB
-    auto snapshot = read_options.snapshot != nullptr
-                        ? read_options.snapshot->GetSequenceNumber()
-                        : versions_->LastSequence();
-    result = NewIteratorImpl(read_options, cfd, snapshot, read_callback);
+    result = NewIteratorImpl(read_options, cfd,
+                             (read_options.snapshot != nullptr)
+                                 ? read_options.snapshot->GetSequenceNumber()
+                                 : kMaxSequenceNumber,
+                             read_callback);
  }
  return result;
 }
@@ -2692,6 +2763,24 @@ ArenaWrappedDBIter* DBImpl::NewIteratorImpl(const ReadOptions& read_options,
                                            bool allow_refresh) {
  SuperVersion* sv = cfd->GetReferencedSuperVersion(this);

+  TEST_SYNC_POINT("DBImpl::NewIterator:1");
+  TEST_SYNC_POINT("DBImpl::NewIterator:2");
+
+  if (snapshot == kMaxSequenceNumber) {
+    // Note that the snapshot is assigned AFTER referencing the super
+    // version because otherwise a flush happening in between may compact away
+    // data for the snapshot, so the reader would see neither data that was be
+    // visible to the snapshot before compaction nor the newer data inserted
+    // afterwards.
+    // Note that the super version might not contain all the data available
+    // to this snapshot, but in that case it can see all the data in the
+    // super version, which is a valid consistent state after the user
+    // calls NewIterator().
+    snapshot = versions_->LastSequence();
+    TEST_SYNC_POINT("DBImpl::NewIterator:3");
+    TEST_SYNC_POINT("DBImpl::NewIterator:4");
+  }
+
  // Try to generate a DB iterator tree in continuous memory area to be
  // cache friendly. Here is an example of result:
  // +-------------------------------+
@@ -2769,7 +2858,7 @@ Status DBImpl::NewIterators(
        "Tailing iterator not supported in RocksDB lite");
 #else
    for (auto cfh : column_families) {
-      auto cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(cfh)->cfd();
+      auto cfd = static_cast_with_check<ColumnFamilyHandleImpl>(cfh)->cfd();
      SuperVersion* sv = cfd->GetReferencedSuperVersion(this);
      auto iter = new ForwardIterator(this, read_options, cfd, sv,
                                      /* allow_unprepared_value */ true);
@@ -2789,7 +2878,8 @@ Status DBImpl::NewIterators(
                        : versions_->LastSequence();
    for (size_t i = 0; i < column_families.size(); ++i) {
      auto* cfd =
-          reinterpret_cast<ColumnFamilyHandleImpl*>(column_families[i])->cfd();
+          static_cast_with_check<ColumnFamilyHandleImpl>(column_families[i])
+              ->cfd();
      iterators->push_back(
          NewIteratorImpl(read_options, cfd, snapshot, read_callback));
    }
@@ -2896,7 +2986,7 @@ void DBImpl::ReleaseSnapshot(const Snapshot* s) {
 #ifndef ROCKSDB_LITE
 Status DBImpl::GetPropertiesOfAllTables(ColumnFamilyHandle* column_family,
                                        TablePropertiesCollection* props) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();

  // Increment the ref count
@@ -2918,7 +3008,7 @@ Status DBImpl::GetPropertiesOfAllTables(ColumnFamilyHandle* column_family,
 Status DBImpl::GetPropertiesOfTablesInRange(ColumnFamilyHandle* column_family,
                                            const Range* range, std::size_t n,
                                            TablePropertiesCollection* props) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();

  // Increment the ref count
@@ -2954,7 +3044,7 @@ FileSystem* DBImpl::GetFileSystem() const {

 Options DBImpl::GetOptions(ColumnFamilyHandle* column_family) const {
  InstrumentedMutexLock l(&mutex_);
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  return Options(BuildDBOptions(immutable_db_options_, mutable_db_options_),
                 cfh->cfd()->GetLatestCFOptions());
 }
@@ -2968,7 +3058,8 @@ bool DBImpl::GetProperty(ColumnFamilyHandle* column_family,
                         const Slice& property, std::string* value) {
  const DBPropertyInfo* property_info = GetPropertyInfo(property);
  value->clear();
-  auto cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+  auto cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
  if (property_info == nullptr) {
    return false;
  } else if (property_info->handle_int) {
@@ -3002,7 +3093,8 @@ bool DBImpl::GetMapProperty(ColumnFamilyHandle* column_family,
                            std::map<std::string, std::string>* value) {
  const DBPropertyInfo* property_info = GetPropertyInfo(property);
  value->clear();
-  auto cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+  auto cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
  if (property_info == nullptr) {
    return false;
  } else if (property_info->handle_map) {
@@ -3021,7 +3113,8 @@ bool DBImpl::GetIntProperty(ColumnFamilyHandle* column_family,
  if (property_info == nullptr || property_info->handle_int == nullptr) {
    return false;
  }
-  auto cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+  auto cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
  return GetIntPropertyInternal(cfd, *property_info, false, value);
 }

@@ -3194,7 +3287,7 @@ void DBImpl::GetApproximateMemTableStats(ColumnFamilyHandle* column_family,
                                         uint64_t* const count,
                                         uint64_t* const size) {
  ColumnFamilyHandleImpl* cfh =
-      reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  ColumnFamilyData* cfd = cfh->cfd();
  SuperVersion* sv = GetAndRefSuperVersion(cfd);

@@ -3219,7 +3312,7 @@ Status DBImpl::GetApproximateSizes(const SizeApproximationOptions& options,
  }

  Version* v;
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  SuperVersion* sv = GetAndRefSuperVersion(cfd);
  v = sv->current;
@@ -3377,7 +3470,7 @@ Status DBImpl::DeleteFilesInRanges(ColumnFamilyHandle* column_family,
                                   const RangePtr* ranges, size_t n,
                                   bool include_end) {
  Status status;
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  ColumnFamilyData* cfd = cfh->cfd();
  VersionEdit edit;
  std::set<FileMetaData*> deleted_files;
@@ -3467,10 +3560,16 @@ void DBImpl::GetLiveFilesMetaData(std::vector<LiveFileMetaData>* metadata) {
  versions_->GetLiveFilesMetaData(metadata);
 }

+Status DBImpl::GetLiveFilesChecksumInfo(FileChecksumList* checksum_list) {
+  InstrumentedMutexLock l(&mutex_);
+  return versions_->GetLiveFilesChecksumInfo(checksum_list);
+}
+
 void DBImpl::GetColumnFamilyMetaData(ColumnFamilyHandle* column_family,
                                     ColumnFamilyMetaData* cf_meta) {
  assert(column_family);
-  auto* cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+  auto* cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
  auto* sv = GetAndRefSuperVersion(cfd);
  {
    // Without mutex, Version::GetColumnFamilyMetaData will have data race with
@@ -3586,6 +3685,37 @@ Status DBImpl::GetDbIdentityFromIdentityFile(std::string* identity) const {
  return s;
 }

+Status DBImpl::GetDbSessionId(std::string& session_id) const {
+  session_id.assign(db_session_id_);
+  return Status::OK();
+}
+
+void DBImpl::SetDbSessionId() {
+  // GenerateUniqueId() generates an identifier that has a negligible
+  // probability of being duplicated, ~128 bits of entropy
+  std::string uuid = env_->GenerateUniqueId();
+
+  // Hash and reformat that down to a more compact format, 20 characters
+  // in base-36 ([0-9A-Z]), which is ~103 bits of entropy, which is enough
+  // to expect no collisions across a billion servers each opening DBs
+  // a million times (~2^50). Benefits vs. raw unique id:
+  // * Save ~ dozen bytes per SST file
+  // * Shorter shared backup file names (some platforms have low limits)
+  // * Visually distinct from DB id format
+  uint64_t a = NPHash64(uuid.data(), uuid.size(), 1234U);
+  uint64_t b = NPHash64(uuid.data(), uuid.size(), 5678U);
+  db_session_id_.resize(20);
+  static const char* const base36 = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ";
+  size_t i = 0;
+  for (; i < 10U; ++i, a /= 36U) {
+    db_session_id_[i] = base36[a % 36];
+  }
+  for (; i < 20U; ++i, b /= 36U) {
+    db_session_id_[i] = base36[b % 36];
+  }
+  TEST_SYNC_POINT_CALLBACK("DBImpl::SetDbSessionId", &db_session_id_);
+}
+
 // Default implementation -- returns not supported status
 Status DB::CreateColumnFamily(const ColumnFamilyOptions& /*cf_options*/,
                              const std::string& /*column_family_name*/,
@@ -4165,7 +4295,8 @@ Status DBImpl::IngestExternalFiles(
        static_cast<ColumnFamilyHandleImpl*>(args[i].column_family)->cfd();
    SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
    exec_results[i].second = ingestion_jobs[i].Prepare(
-        args[i].external_files, start_file_number, super_version);
+        args[i].external_files, args[i].files_checksums,
+        args[i].files_checksum_func_names, start_file_number, super_version);
    exec_results[i].first = true;
    CleanupSuperVersion(super_version);
  }
@@ -4176,7 +4307,8 @@ Status DBImpl::IngestExternalFiles(
        static_cast<ColumnFamilyHandleImpl*>(args[0].column_family)->cfd();
    SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
    exec_results[0].second = ingestion_jobs[0].Prepare(
-        args[0].external_files, next_file_number, super_version);
+        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);
  }
@@ -4352,6 +4484,14 @@ Status DBImpl::IngestExternalFiles(
 #endif  // !NDEBUG
        }
      }
+    } else if (versions_->io_status().IsIOError()) {
+      // Error while writing to MANIFEST.
+      // In fact, versions_->io_status() can also be the result of renaming
+      // CURRENT file. With current code, it's just difficult to tell. So just
+      // be pessimistic and try write to a new MANIFEST.
+      // TODO: distinguish between MANIFEST write and CURRENT renaming
+      const IOStatus& io_s = versions_->io_status();
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite);
    }

    // Resume writes to the DB
@@ -4411,7 +4551,7 @@ Status DBImpl::CreateColumnFamilyWithImport(
  }

  // Import sst files from metadata.
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(*handle);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(*handle);
  auto cfd = cfh->cfd();
  ImportColumnFamilyJob import_job(env_, versions_.get(), cfd,
                                   immutable_db_options_, file_options_,
@@ -350,12 +350,20 @@ class DBImpl : public DB {

  virtual Status GetDbIdentityFromIdentityFile(std::string* identity) const;

+  virtual Status GetDbSessionId(std::string& session_id) const override;
+
  ColumnFamilyHandle* DefaultColumnFamily() const override;

  ColumnFamilyHandle* PersistentStatsColumnFamily() const;

  virtual Status Close() override;

+  virtual Status DisableFileDeletions() override;
+
+  virtual Status EnableFileDeletions(bool force) override;
+
+  virtual bool IsFileDeletionsEnabled() const;
+
  Status GetStatsHistory(
      uint64_t start_time, uint64_t end_time,
      std::unique_ptr<StatsHistoryIterator>* stats_iterator) override;
@@ -363,9 +371,6 @@ class DBImpl : public DB {
 #ifndef ROCKSDB_LITE
  using DB::ResetStats;
  virtual Status ResetStats() override;
-  virtual Status DisableFileDeletions() override;
-  virtual Status EnableFileDeletions(bool force) override;
-  virtual int IsFileDeletionsEnabled() const;
  // All the returned filenames start with "/"
  virtual Status GetLiveFiles(std::vector<std::string>&,
                              uint64_t* manifest_file_size,
@@ -388,6 +393,9 @@ class DBImpl : public DB {
  virtual void GetLiveFilesMetaData(
      std::vector<LiveFileMetaData>* metadata) override;

+  virtual Status GetLiveFilesChecksumInfo(
+      FileChecksumList* checksum_list) override;
+
  // Obtains the meta data of the specified column family of the DB.
  // Status::NotFound() will be returned if the current DB does not have
  // any column family match the specified name.
@@ -477,6 +485,7 @@ class DBImpl : public DB {
  Status GetImpl(const ReadOptions& options, const Slice& key,
                 GetImplOptions& get_impl_options);

+  // If `snapshot` == kMaxSequenceNumber, set a recent one inside the file.
  ArenaWrappedDBIter* NewIteratorImpl(const ReadOptions& options,
                                      ColumnFamilyData* cfd,
                                      SequenceNumber snapshot,
@@ -843,8 +852,8 @@ class DBImpl : public DB {
  InstrumentedMutex* mutex() const { return &mutex_; }

  // Initialize a brand new DB. The DB directory is expected to be empty before
-  // calling it.
-  Status NewDB();
+  // calling it. Push new manifest file name into `new_filenames`.
+  Status NewDB(std::vector<std::string>* new_filenames);

  // This is to be used only by internal rocksdb classes.
  static Status Open(const DBOptions& db_options, const std::string& name,
@@ -980,6 +989,9 @@ class DBImpl : public DB {
 protected:
  const std::string dbname_;
  std::string db_id_;
+  // db_session_id_ is an identifier that gets reset
+  // every time the DB is opened
+  std::string db_session_id_;
  std::unique_ptr<VersionSet> versions_;
  // Flag to check whether we allocated and own the info log file
  bool own_info_log_;
@@ -1161,6 +1173,10 @@ class DBImpl : public DB {
  // bump up the version set's next_file_number_ to be 1 + largest_file_number.
  Status FinishBestEffortsRecovery();

+  // SetDbSessionId() should be called in the constuctor DBImpl()
+  // to ensure that db_session_id_ gets updated every time the DB is opened
+  void SetDbSessionId();
+
 private:
  friend class DB;
  friend class ErrorHandler;
@@ -1186,7 +1202,7 @@ class DBImpl : public DB {
  friend class StatsHistoryTest_PersistentStatsCreateColumnFamilies_Test;
 #ifndef NDEBUG
  friend class DBTest2_ReadCallbackTest_Test;
-  friend class WriteCallbackTest_WriteWithCallbackTest_Test;
+  friend class WriteCallbackPTest_WriteWithCallbackTest_Test;
  friend class XFTransactionWriteHandler;
  friend class DBBlobIndexTest;
  friend class WriteUnpreparedTransactionTest_RecoveryTest_Test;
@@ -1694,8 +1710,8 @@ class DBImpl : public DB {
  size_t GetWalPreallocateBlockSize(uint64_t write_buffer_size) const;
  Env::WriteLifeTimeHint CalculateWALWriteHint() { return Env::WLTH_SHORT; }

-  Status CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
-                   size_t preallocate_block_size, log::Writer** new_log);
+  IOStatus CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
+                     size_t preallocate_block_size, log::Writer** new_log);

  // Validate self-consistency of DB options
  static Status ValidateOptions(const DBOptions& db_options);
@@ -1779,6 +1795,8 @@ class DBImpl : public DB {
      SuperVersion* sv, SequenceNumber snap_seqnum, ReadCallback* callback,
      bool* is_blob_index);

+  Status DisableFileDeletionsWithLock();
+
  // table_cache_ provides its own synchronization
  std::shared_ptr<Cache> table_cache_;

@@ -154,7 +154,8 @@ Status DBImpl::FlushMemTableToOutputFile(
      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);
+      true /* sync_output_directory */, true /* write_manifest */, thread_pri,
+      db_id_, db_session_id_);
  FileMetaData file_meta;

  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:BeforePickMemtables");
@@ -193,7 +194,9 @@ Status DBImpl::FlushMemTableToOutputFile(
  } else {
    flush_job.Cancel();
  }
-  io_s = flush_job.io_status();
+  if (io_s.ok()) {
+    io_s = flush_job.io_status();
+  }

  if (s.ok()) {
    InstallSuperVersionAndScheduleWork(cfd, superversion_context,
@@ -210,7 +213,15 @@ Status DBImpl::FlushMemTableToOutputFile(
  if (!s.ok() && !s.IsShutdownInProgress() && !s.IsColumnFamilyDropped()) {
    if (!io_s.ok() && !io_s.IsShutdownInProgress() &&
        !io_s.IsColumnFamilyDropped()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+      // Error while writing to MANIFEST.
+      // In fact, versions_->io_status() can also be the result of renaming
+      // CURRENT file. With current code, it's just difficult to tell. So just
+      // be pessimistic and try write to a new MANIFEST.
+      // TODO: distinguish between MANIFEST write and CURRENT renaming
+      auto err_reason = versions_->io_status().ok()
+                            ? BackgroundErrorReason::kFlush
+                            : BackgroundErrorReason::kManifestWrite;
+      error_handler_.SetBGError(io_s, err_reason);
    } else {
      Status new_bg_error = s;
      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
@@ -345,7 +356,7 @@ 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));
+        thread_pri, db_id_, db_session_id_));
    jobs.back()->PickMemTable();
  }

@@ -573,8 +584,16 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
  // Need to undo atomic flush if something went wrong, i.e. s is not OK and
  // it is not because of CF drop.
  if (!s.ok() && !s.IsColumnFamilyDropped()) {
-    if (!io_s.ok() && io_s.IsColumnFamilyDropped()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+    if (!io_s.ok() && !io_s.IsColumnFamilyDropped()) {
+      // Error while writing to MANIFEST.
+      // In fact, versions_->io_status() can also be the result of renaming
+      // CURRENT file. With current code, it's just difficult to tell. So just
+      // be pessimistic and try write to a new MANIFEST.
+      // TODO: distinguish between MANIFEST write and CURRENT renaming
+      auto err_reason = versions_->io_status().ok()
+                            ? BackgroundErrorReason::kFlush
+                            : BackgroundErrorReason::kManifestWrite;
+      error_handler_.SetBGError(io_s, err_reason);
    } else {
      Status new_bg_error = s;
      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
@@ -678,7 +697,7 @@ void DBImpl::NotifyOnFlushCompleted(
 Status DBImpl::CompactRange(const CompactRangeOptions& options,
                            ColumnFamilyHandle* column_family,
                            const Slice* begin, const Slice* end) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();

  if (options.target_path_id >= cfd->ioptions()->cf_paths.size()) {
@@ -868,7 +887,8 @@ Status DBImpl::CompactFiles(const CompactionOptions& compact_options,
    return Status::InvalidArgument("ColumnFamilyHandle must be non-null.");
  }

-  auto cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+  auto cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
  assert(cfd);

  Status s;
@@ -1001,7 +1021,7 @@ Status DBImpl::CompactFilesImpl(
  assert(cfd->compaction_picker());
  c.reset(cfd->compaction_picker()->CompactFiles(
      compact_options, input_files, output_level, version->storage_info(),
-      *cfd->GetLatestMutableCFOptions(), output_path_id));
+      *cfd->GetLatestMutableCFOptions(), mutable_db_options_, output_path_id));
  // we already sanitized the set of input files and checked for conflicts
  // without releasing the lock, so we're guaranteed a compaction can be formed.
  assert(c != nullptr);
@@ -1031,7 +1051,8 @@ Status DBImpl::CompactFilesImpl(
      snapshot_checker, table_cache_, &event_logger_,
      c->mutable_cf_options()->paranoid_file_checks,
      c->mutable_cf_options()->report_bg_io_stats, dbname_,
-      &compaction_job_stats, Env::Priority::USER, &manual_compaction_paused_);
+      &compaction_job_stats, Env::Priority::USER, &manual_compaction_paused_,
+      db_id_, db_session_id_);

  // Creating a compaction influences the compaction score because the score
  // takes running compactions into account (by skipping files that are already
@@ -1088,7 +1109,12 @@ Status DBImpl::CompactFilesImpl(
                   "[%s] [JOB %d] Compaction error: %s",
                   c->column_family_data()->GetName().c_str(),
                   job_context->job_id, status.ToString().c_str());
-    error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
+    IOStatus io_s = compaction_job.io_status();
+    if (!io_s.ok()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kCompaction);
+    } else {
+      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
+    }
  }

  if (output_file_names != nullptr) {
@@ -1341,7 +1367,7 @@ Status DBImpl::ReFitLevel(ColumnFamilyData* cfd, int level, int target_level) {
 }

 int DBImpl::NumberLevels(ColumnFamilyHandle* column_family) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  return cfh->cfd()->NumberLevels();
 }

@@ -1350,7 +1376,7 @@ int DBImpl::MaxMemCompactionLevel(ColumnFamilyHandle* /*column_family*/) {
 }

 int DBImpl::Level0StopWriteTrigger(ColumnFamilyHandle* column_family) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  InstrumentedMutexLock l(&mutex_);
  return cfh->cfd()
      ->GetSuperVersion()
@@ -1359,7 +1385,7 @@ int DBImpl::Level0StopWriteTrigger(ColumnFamilyHandle* column_family) {

 Status DBImpl::Flush(const FlushOptions& flush_options,
                     ColumnFamilyHandle* column_family) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  ROCKS_LOG_INFO(immutable_db_options_.info_log, "[%s] Manual flush start.",
                 cfh->GetName().c_str());
  Status s;
@@ -1510,9 +1536,9 @@ Status DBImpl::RunManualCompaction(
        scheduled ||
        (((manual.manual_end = &manual.tmp_storage1) != nullptr) &&
         ((compaction = manual.cfd->CompactRange(
-               *manual.cfd->GetLatestMutableCFOptions(), manual.input_level,
-               manual.output_level, compact_range_options, manual.begin,
-               manual.end, &manual.manual_end, &manual_conflict,
+               *manual.cfd->GetLatestMutableCFOptions(), mutable_db_options_,
+               manual.input_level, manual.output_level, compact_range_options,
+               manual.begin, manual.end, &manual.manual_end, &manual_conflict,
               max_file_num_to_ignore)) == nullptr &&
          manual_conflict))) {
      // exclusive manual compactions should not see a conflict during
@@ -2616,7 +2642,8 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
      // compaction is not necessary. Need to make sure mutex is held
      // until we make a copy in the following code
      TEST_SYNC_POINT("DBImpl::BackgroundCompaction():BeforePickCompaction");
-      c.reset(cfd->PickCompaction(*mutable_cf_options, log_buffer));
+      c.reset(cfd->PickCompaction(*mutable_cf_options, mutable_db_options_,
+                                  log_buffer));
      TEST_SYNC_POINT("DBImpl::BackgroundCompaction():AfterPickCompaction");

      if (c != nullptr) {
@@ -2687,7 +2714,6 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
    for (const auto& f : *c->inputs(0)) {
      c->edit()->DeleteFile(c->level(), f->fd.GetNumber());
    }
-    versions_->SetIOStatusOK();
    status = versions_->LogAndApply(c->column_family_data(),
                                    *c->mutable_cf_options(), c->edit(),
                                    &mutex_, directories_.GetDbDir());
@@ -2745,7 +2771,6 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
      }
    }

-    versions_->SetIOStatusOK();
    status = versions_->LogAndApply(c->column_family_data(),
                                    *c->mutable_cf_options(), c->edit(),
                                    &mutex_, directories_.GetDbDir());
@@ -2822,7 +2847,8 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
        &event_logger_, c->mutable_cf_options()->paranoid_file_checks,
        c->mutable_cf_options()->report_bg_io_stats, dbname_,
        &compaction_job_stats, thread_pri,
-        is_manual ? &manual_compaction_paused_ : nullptr);
+        is_manual ? &manual_compaction_paused_ : nullptr, db_id_,
+        db_session_id_);
    compaction_job.Prepare();

    NotifyOnCompactionBegin(c->column_family_data(), c.get(), status,
@@ -2877,7 +2903,15 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
    ROCKS_LOG_WARN(immutable_db_options_.info_log, "Compaction error: %s",
                   status.ToString().c_str());
    if (!io_s.ok()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kCompaction);
+      // Error while writing to MANIFEST.
+      // In fact, versions_->io_status() can also be the result of renaming
+      // CURRENT file. With current code, it's just difficult to tell. So just
+      // be pessimistic and try write to a new MANIFEST.
+      // TODO: distinguish between MANIFEST write and CURRENT renaming
+      auto err_reason = versions_->io_status().ok()
+                            ? BackgroundErrorReason::kCompaction
+                            : BackgroundErrorReason::kManifestWrite;
+      error_handler_.SetBGError(io_s, err_reason);
    } else {
      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
    }
@@ -47,7 +47,7 @@ int64_t DBImpl::TEST_MaxNextLevelOverlappingBytes(
  if (column_family == nullptr) {
    cfd = default_cf_handle_->cfd();
  } else {
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+    auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
    cfd = cfh->cfd();
  }
  InstrumentedMutexLock l(&mutex_);
@@ -57,7 +57,7 @@ int64_t DBImpl::TEST_MaxNextLevelOverlappingBytes(
 void DBImpl::TEST_GetFilesMetaData(
    ColumnFamilyHandle* column_family,
    std::vector<std::vector<FileMetaData>>* metadata) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  InstrumentedMutexLock l(&mutex_);
  metadata->resize(NumberLevels());
@@ -88,7 +88,7 @@ Status DBImpl::TEST_CompactRange(int level, const Slice* begin,
  if (column_family == nullptr) {
    cfd = default_cf_handle_->cfd();
  } else {
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+    auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
    cfd = cfh->cfd();
  }
  int output_level =
@@ -131,7 +131,7 @@ Status DBImpl::TEST_FlushMemTable(bool wait, bool allow_write_stall,
  if (cfh == nullptr) {
    cfd = default_cf_handle_->cfd();
  } else {
-    auto cfhi = reinterpret_cast<ColumnFamilyHandleImpl*>(cfh);
+    auto cfhi = static_cast_with_check<ColumnFamilyHandleImpl>(cfh);
    cfd = cfhi->cfd();
  }
  return FlushMemTable(cfd, fo, FlushReason::kTest);
@@ -152,7 +152,7 @@ Status DBImpl::TEST_WaitForFlushMemTable(ColumnFamilyHandle* column_family) {
  if (column_family == nullptr) {
    cfd = default_cf_handle_->cfd();
  } else {
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+    auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
    cfd = cfh->cfd();
  }
  return WaitForFlushMemTable(cfd, nullptr, false);
@@ -242,7 +242,7 @@ Status DBImpl::TEST_GetLatestMutableCFOptions(
    ColumnFamilyHandle* column_family, MutableCFOptions* mutable_cf_options) {
  InstrumentedMutexLock l(&mutex_);

-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  *mutable_cf_options = *cfh->cfd()->GetLatestMutableCFOptions();
  return Status::OK();
 }
@@ -7,22 +7,22 @@
 // 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 <vector>

 #include "db/column_family.h"
+#include "db/db_impl/db_impl.h"
 #include "db/job_context.h"
 #include "db/version_set.h"
 #include "rocksdb/status.h"
+#include "util/cast_util.h"

 namespace ROCKSDB_NAMESPACE {

 #ifndef ROCKSDB_LITE
 Status DBImpl::SuggestCompactRange(ColumnFamilyHandle* column_family,
                                   const Slice* begin, const Slice* end) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  InternalKey start_key, end_key;
  if (begin != nullptr) {
@@ -36,6 +36,62 @@ uint64_t DBImpl::MinObsoleteSstNumberToKeep() {
  return std::numeric_limits<uint64_t>::max();
 }

+Status DBImpl::DisableFileDeletions() {
+  InstrumentedMutexLock l(&mutex_);
+  return DisableFileDeletionsWithLock();
+}
+
+Status DBImpl::DisableFileDeletionsWithLock() {
+  mutex_.AssertHeld();
+  ++disable_delete_obsolete_files_;
+  if (disable_delete_obsolete_files_ == 1) {
+    ROCKS_LOG_INFO(immutable_db_options_.info_log, "File Deletions Disabled");
+  } else {
+    ROCKS_LOG_WARN(immutable_db_options_.info_log,
+                   "File Deletions Disabled, but already disabled. Counter: %d",
+                   disable_delete_obsolete_files_);
+  }
+  return Status::OK();
+}
+
+Status DBImpl::EnableFileDeletions(bool force) {
+  // Job id == 0 means that this is not our background process, but rather
+  // user thread
+  JobContext job_context(0);
+  bool file_deletion_enabled = false;
+  {
+    InstrumentedMutexLock l(&mutex_);
+    if (force) {
+      // if force, we need to enable file deletions right away
+      disable_delete_obsolete_files_ = 0;
+    } else if (disable_delete_obsolete_files_ > 0) {
+      --disable_delete_obsolete_files_;
+    }
+    if (disable_delete_obsolete_files_ == 0) {
+      file_deletion_enabled = true;
+      FindObsoleteFiles(&job_context, true);
+      bg_cv_.SignalAll();
+    }
+  }
+  if (file_deletion_enabled) {
+    ROCKS_LOG_INFO(immutable_db_options_.info_log, "File Deletions Enabled");
+    if (job_context.HaveSomethingToDelete()) {
+      PurgeObsoleteFiles(job_context);
+    }
+  } else {
+    ROCKS_LOG_WARN(immutable_db_options_.info_log,
+                   "File Deletions Enable, but not really enabled. Counter: %d",
+                   disable_delete_obsolete_files_);
+  }
+  job_context.Clean();
+  LogFlush(immutable_db_options_.info_log);
+  return Status::OK();
+}
+
+bool DBImpl::IsFileDeletionsEnabled() const {
+  return 0 == disable_delete_obsolete_files_;
+}
+
 // * Returns the list of live files in 'sst_live' and 'blob_live'.
 // If it's doing full scan:
 // * Returns the list of all files in the filesystem in
@@ -134,7 +190,7 @@ 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);  // Ignore errors
+      env_->GetChildren(path, &files).PermitUncheckedError();  // Ignore errors
      for (const std::string& file : files) {
        uint64_t number;
        FileType type;
@@ -686,13 +742,15 @@ uint64_t PrecomputeMinLogNumberToKeep(
 Status DBImpl::FinishBestEffortsRecovery() {
  mutex_.AssertHeld();
  std::vector<std::string> paths;
-  paths.push_back(dbname_);
+  paths.push_back(NormalizePath(dbname_ + std::string(1, kFilePathSeparator)));
  for (const auto& db_path : immutable_db_options_.db_paths) {
-    paths.push_back(db_path.path);
+    paths.push_back(
+        NormalizePath(db_path.path + std::string(1, kFilePathSeparator)));
  }
  for (const auto* cfd : *versions_->GetColumnFamilySet()) {
    for (const auto& cf_path : cfd->ioptions()->cf_paths) {
-      paths.push_back(cf_path.path);
+      paths.push_back(
+          NormalizePath(cf_path.path + std::string(1, kFilePathSeparator)));
    }
  }
  // Dedup paths
@@ -711,7 +769,8 @@ Status DBImpl::FinishBestEffortsRecovery() {
      if (!ParseFileName(fname, &number, &type)) {
        continue;
      }
-      const std::string normalized_fpath = NormalizePath(path + fname);
+      // path ends with '/' or '\\'
+      const std::string normalized_fpath = path + fname;
      largest_file_number = std::max(largest_file_number, number);
      if (type == kTableFile && number >= next_file_number &&
          files_to_delete.find(normalized_fpath) == files_to_delete.end()) {
@@ -253,7 +253,7 @@ Status DBImpl::ValidateOptions(const DBOptions& db_options) {
  return Status::OK();
 }

-Status DBImpl::NewDB() {
+Status DBImpl::NewDB(std::vector<std::string>* new_filenames) {
  VersionEdit new_db;
  Status s = SetIdentityFile(env_, dbname_);
  if (!s.ok()) {
@@ -293,6 +293,10 @@ Status DBImpl::NewDB() {
  if (s.ok()) {
    // Make "CURRENT" file that points to the new manifest file.
    s = SetCurrentFile(fs_.get(), dbname_, 1, directories_.GetDbDir());
+    if (new_filenames) {
+      new_filenames->emplace_back(
+          manifest.substr(manifest.find_last_of("/\\") + 1));
+    }
  } else {
    fs_->DeleteFile(manifest, IOOptions(), nullptr);
  }
@@ -356,6 +360,7 @@ Status DBImpl::Recover(

  bool is_new_db = false;
  assert(db_lock_ == nullptr);
+  std::vector<std::string> files_in_dbname;
  if (!read_only) {
    Status s = directories_.SetDirectories(fs_.get(), dbname_,
                                           immutable_db_options_.wal_dir,
@@ -370,10 +375,35 @@ Status DBImpl::Recover(
    }

    std::string current_fname = CurrentFileName(dbname_);
-    s = env_->FileExists(current_fname);
+    // Path to any MANIFEST file in the db dir. It does not matter which one.
+    // Since best-efforts recovery ignores CURRENT file, existence of a
+    // MANIFEST indicates the recovery to recover existing db. If no MANIFEST
+    // can be found, a new db will be created.
+    std::string manifest_path;
+    if (!immutable_db_options_.best_efforts_recovery) {
+      s = env_->FileExists(current_fname);
+    } else {
+      s = Status::NotFound();
+      Status io_s = env_->GetChildren(dbname_, &files_in_dbname);
+      if (!io_s.ok()) {
+        s = io_s;
+        files_in_dbname.clear();
+      }
+      for (const std::string& file : files_in_dbname) {
+        uint64_t number = 0;
+        FileType type = kLogFile;  // initialize
+        if (ParseFileName(file, &number, &type) && type == kDescriptorFile) {
+          // Found MANIFEST (descriptor log), thus best-efforts recovery does
+          // not have to treat the db as empty.
+          s = Status::OK();
+          manifest_path = dbname_ + "/" + file;
+          break;
+        }
+      }
+    }
    if (s.IsNotFound()) {
      if (immutable_db_options_.create_if_missing) {
-        s = NewDB();
+        s = NewDB(&files_in_dbname);
        is_new_db = true;
        if (!s.ok()) {
          return s;
@@ -398,14 +428,14 @@ Status DBImpl::Recover(
      FileOptions customized_fs(file_options_);
      customized_fs.use_direct_reads |=
          immutable_db_options_.use_direct_io_for_flush_and_compaction;
-      s = fs_->NewRandomAccessFile(current_fname, customized_fs, &idfile,
-                                   nullptr);
+      const std::string& fname =
+          manifest_path.empty() ? current_fname : manifest_path;
+      s = fs_->NewRandomAccessFile(fname, customized_fs, &idfile, nullptr);
      if (!s.ok()) {
        std::string error_str = s.ToString();
        // Check if unsupported Direct I/O is the root cause
        customized_fs.use_direct_reads = false;
-        s = fs_->NewRandomAccessFile(current_fname, customized_fs, &idfile,
-                                     nullptr);
+        s = fs_->NewRandomAccessFile(fname, customized_fs, &idfile, nullptr);
        if (s.ok()) {
          return Status::InvalidArgument(
              "Direct I/O is not supported by the specified DB.");
@@ -415,6 +445,16 @@ Status DBImpl::Recover(
        }
      }
    }
+  } else if (immutable_db_options_.best_efforts_recovery) {
+    assert(files_in_dbname.empty());
+    Status s = env_->GetChildren(dbname_, &files_in_dbname);
+    if (s.IsNotFound()) {
+      return Status::InvalidArgument(dbname_,
+                                     "does not exist (open for read only)");
+    } else if (s.IsIOError()) {
+      return s;
+    }
+    assert(s.ok());
  }
  assert(db_id_.empty());
  Status s;
@@ -422,9 +462,13 @@ Status DBImpl::Recover(
  if (!immutable_db_options_.best_efforts_recovery) {
    s = versions_->Recover(column_families, read_only, &db_id_);
  } else {
-    s = versions_->TryRecover(column_families, read_only, &db_id_,
-                              &missing_table_file);
+    assert(!files_in_dbname.empty());
+    s = versions_->TryRecover(column_families, read_only, files_in_dbname,
+                              &db_id_, &missing_table_file);
    if (s.ok()) {
+      // TryRecover may delete previous column_family_set_.
+      column_family_memtables_.reset(
+          new ColumnFamilyMemTablesImpl(versions_->GetColumnFamilySet()));
      s = FinishBestEffortsRecovery();
    }
  }
@@ -480,6 +524,7 @@ Status DBImpl::Recover(
    s = InitPersistStatsColumnFamily();
  }

+  std::vector<std::string> files_in_wal_dir;
  if (s.ok()) {
    // Initial max_total_in_memory_state_ before recovery logs. Log recovery
    // may check this value to decide whether to flush.
@@ -506,9 +551,8 @@ Status DBImpl::Recover(
    // Note that prev_log_number() is no longer used, but we pay
    // attention to it in case we are recovering a database
    // produced by an older version of rocksdb.
-    std::vector<std::string> filenames;
    if (!immutable_db_options_.best_efforts_recovery) {
-      s = env_->GetChildren(immutable_db_options_.wal_dir, &filenames);
+      s = env_->GetChildren(immutable_db_options_.wal_dir, &files_in_wal_dir);
    }
    if (s.IsNotFound()) {
      return Status::InvalidArgument("wal_dir not found",
@@ -518,15 +562,15 @@ Status DBImpl::Recover(
    }

    std::vector<uint64_t> logs;
-    for (size_t i = 0; i < filenames.size(); i++) {
+    for (const auto& file : files_in_wal_dir) {
      uint64_t number;
      FileType type;
-      if (ParseFileName(filenames[i], &number, &type) && type == kLogFile) {
+      if (ParseFileName(file, &number, &type) && type == kLogFile) {
        if (is_new_db) {
          return Status::Corruption(
              "While creating a new Db, wal_dir contains "
              "existing log file: ",
-              filenames[i]);
+              file);
        } else {
          logs.push_back(number);
        }
@@ -578,15 +622,24 @@ Status DBImpl::Recover(
    // to reflect the most recent OPTIONS file. It does not matter for regular
    // read-write db instance because options_file_number_ will later be
    // updated to versions_->NewFileNumber() in RenameTempFileToOptionsFile.
-    std::vector<std::string> file_names;
+    std::vector<std::string> filenames;
    if (s.ok()) {
-      s = env_->GetChildren(GetName(), &file_names);
+      const std::string normalized_dbname = NormalizePath(dbname_);
+      const std::string normalized_wal_dir =
+          NormalizePath(immutable_db_options_.wal_dir);
+      if (immutable_db_options_.best_efforts_recovery) {
+        filenames = std::move(files_in_dbname);
+      } else if (normalized_dbname == normalized_wal_dir) {
+        filenames = std::move(files_in_wal_dir);
+      } else {
+        s = env_->GetChildren(GetName(), &filenames);
+      }
    }
    if (s.ok()) {
      uint64_t number = 0;
      uint64_t options_file_number = 0;
      FileType type;
-      for (const auto& fname : file_names) {
+      for (const auto& fname : filenames) {
        if (ParseFileName(fname, &number, &type) && type == kOptionsFile) {
          options_file_number = std::max(number, options_file_number);
        }
@@ -594,7 +647,6 @@ Status DBImpl::Recover(
      versions_->options_file_number_ = options_file_number;
    }
  }
-
  return s;
 }

@@ -701,10 +753,10 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
    Status* status;  // nullptr if immutable_db_options_.paranoid_checks==false
    void Corruption(size_t bytes, const Status& s) override {
      ROCKS_LOG_WARN(info_log, "%s%s: dropping %d bytes; %s",
-                     (this->status == nullptr ? "(ignoring error) " : ""),
-                     fname, static_cast<int>(bytes), s.ToString().c_str());
-      if (this->status != nullptr && this->status->ok()) {
-        *this->status = s;
+                     (status == nullptr ? "(ignoring error) " : ""), fname,
+                     static_cast<int>(bytes), s.ToString().c_str());
+      if (status != nullptr && status->ok()) {
+        *status = s;
      }
    }
  };
@@ -827,6 +879,8 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
    Slice record;
    WriteBatch batch;

+    TEST_SYNC_POINT_CALLBACK("DBImpl::RecoverLogFiles:BeforeReadWal",
+                             /*arg=*/nullptr);
    while (!stop_replay_by_wal_filter &&
           reader.ReadRecord(&record, &scratch,
                             immutable_db_options_.wal_recovery_mode) &&
@@ -991,6 +1045,16 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
        status = Status::OK();
      } else if (immutable_db_options_.wal_recovery_mode ==
                 WALRecoveryMode::kPointInTimeRecovery) {
+        if (status.IsIOError()) {
+          ROCKS_LOG_ERROR(immutable_db_options_.info_log,
+                          "IOError during point-in-time reading log #%" PRIu64
+                          " seq #%" PRIu64
+                          ". %s. This likely mean loss of synced WAL, "
+                          "thus recovery fails.",
+                          log_number, *next_sequence,
+                          status.ToString().c_str());
+          return status;
+        }
        // We should ignore the error but not continue replaying
        status = Status::OK();
        stop_replay_for_corruption = true;
@@ -1251,7 +1315,8 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
          mutable_cf_options.compression_opts, paranoid_file_checks,
          cfd->internal_stats(), TableFileCreationReason::kRecovery, &io_s,
          &event_logger_, job_id, Env::IO_HIGH, nullptr /* table_properties */,
-          -1 /* level */, current_time, write_hint);
+          -1 /* level */, current_time, 0 /* oldest_key_time */, write_hint,
+          0 /* file_creation_time */, db_id_, db_session_id_);
      LogFlush(immutable_db_options_.info_log);
      ROCKS_LOG_DEBUG(immutable_db_options_.info_log,
                      "[%s] [WriteLevel0TableForRecovery]"
@@ -1323,9 +1388,10 @@ Status DB::Open(const DBOptions& db_options, const std::string& dbname,
                      !kSeqPerBatch, kBatchPerTxn);
 }

-Status DBImpl::CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
-                         size_t preallocate_block_size, log::Writer** new_log) {
-  Status s;
+IOStatus DBImpl::CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
+                           size_t preallocate_block_size,
+                           log::Writer** new_log) {
+  IOStatus io_s;
  std::unique_ptr<FSWritableFile> lfile;

  DBOptions db_options =
@@ -1343,13 +1409,13 @@ Status DBImpl::CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
        LogFileName(immutable_db_options_.wal_dir, recycle_log_number);
    TEST_SYNC_POINT("DBImpl::CreateWAL:BeforeReuseWritableFile1");
    TEST_SYNC_POINT("DBImpl::CreateWAL:BeforeReuseWritableFile2");
-    s = fs_->ReuseWritableFile(log_fname, old_log_fname, opt_file_options,
-                               &lfile, /*dbg=*/nullptr);
+    io_s = fs_->ReuseWritableFile(log_fname, old_log_fname, opt_file_options,
+                                  &lfile, /*dbg=*/nullptr);
  } else {
-    s = NewWritableFile(fs_.get(), log_fname, &lfile, opt_file_options);
+    io_s = NewWritableFile(fs_.get(), log_fname, &lfile, opt_file_options);
  }

-  if (s.ok()) {
+  if (io_s.ok()) {
    lfile->SetWriteLifeTimeHint(CalculateWALWriteHint());
    lfile->SetPreallocationBlockSize(preallocate_block_size);

@@ -1361,7 +1427,7 @@ Status DBImpl::CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
                               immutable_db_options_.recycle_log_file_num > 0,
                               immutable_db_options_.manual_wal_flush);
  }
-  return s;
+  return io_s;
 }

 Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
@@ -1566,6 +1632,12 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
  auto sfm = static_cast<SstFileManagerImpl*>(
      impl->immutable_db_options_.sst_file_manager.get());
  if (s.ok() && sfm) {
+    // Set Statistics ptr for SstFileManager to dump the stats of
+    // DeleteScheduler.
+    sfm->SetStatisticsPtr(impl->immutable_db_options_.statistics);
+    ROCKS_LOG_INFO(impl->immutable_db_options_.info_log,
+                   "SstFileManager instance %p", sfm);
+
    // Notify SstFileManager about all sst files that already exist in
    // db_paths[0] and cf_paths[0] when the DB is opened.

@@ -4,13 +4,14 @@
 //  (found in the LICENSE.Apache file in the root directory).

 #include "db/db_impl/db_impl_readonly.h"
-#include "db/arena_wrapped_db_iter.h"

+#include "db/arena_wrapped_db_iter.h"
 #include "db/compacted_db_impl.h"
 #include "db/db_impl/db_impl.h"
 #include "db/db_iter.h"
 #include "db/merge_context.h"
 #include "monitoring/perf_context_imp.h"
+#include "util/cast_util.h"

 namespace ROCKSDB_NAMESPACE {

@@ -35,7 +36,7 @@ Status DBImplReadOnly::Get(const ReadOptions& read_options,
  PERF_TIMER_GUARD(get_snapshot_time);
  Status s;
  SequenceNumber snapshot = versions_->LastSequence();
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  if (tracer_) {
    InstrumentedMutexLock lock(&trace_mutex_);
@@ -70,7 +71,7 @@ Status DBImplReadOnly::Get(const ReadOptions& read_options,

 Iterator* DBImplReadOnly::NewIterator(const ReadOptions& read_options,
                                      ColumnFamilyHandle* column_family) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  SuperVersion* super_version = cfd->GetSuperVersion()->Ref();
  SequenceNumber latest_snapshot = versions_->LastSequence();
@@ -111,7 +112,7 @@ Status DBImplReadOnly::NewIterators(
          : latest_snapshot;

  for (auto cfh : column_families) {
-    auto* cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(cfh)->cfd();
+    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,
@@ -128,12 +129,38 @@ Status DBImplReadOnly::NewIterators(
  return Status::OK();
 }

+namespace {
+// Return OK if dbname exists in the file system
+// or create_if_missing is false
+Status OpenForReadOnlyCheckExistence(const DBOptions& db_options,
+                                     const std::string& dbname) {
+  Status s;
+  if (!db_options.create_if_missing) {
+    // Attempt to read "CURRENT" file
+    const std::shared_ptr<FileSystem>& fs = db_options.env->GetFileSystem();
+    std::string manifest_path;
+    uint64_t manifest_file_number;
+    s = VersionSet::GetCurrentManifestPath(dbname, fs.get(), &manifest_path,
+                                           &manifest_file_number);
+    if (!s.ok()) {
+      return Status::NotFound(CurrentFileName(dbname), "does not exist");
+    }
+  }
+  return s;
+}
+}  // namespace
+
 Status DB::OpenForReadOnly(const Options& options, const std::string& dbname,
                           DB** dbptr, bool /*error_if_log_file_exist*/) {
+  // If dbname does not exist in the file system, should not do anything
+  Status s = OpenForReadOnlyCheckExistence(options, dbname);
+  if (!s.ok()) {
+    return s;
+  }
+
  *dbptr = nullptr;

  // Try to first open DB as fully compacted DB
-  Status s;
  s = CompactedDBImpl::Open(options, dbname, dbptr);
  if (s.ok()) {
    return s;
@@ -146,7 +173,8 @@ Status DB::OpenForReadOnly(const Options& options, const std::string& dbname,
      ColumnFamilyDescriptor(kDefaultColumnFamilyName, cf_options));
  std::vector<ColumnFamilyHandle*> handles;

-  s = DB::OpenForReadOnly(db_options, dbname, column_families, &handles, dbptr);
+  s = DBImplReadOnly::OpenForReadOnlyWithoutCheck(
+      db_options, dbname, column_families, &handles, dbptr);
  if (s.ok()) {
    assert(handles.size() == 1);
    // i can delete the handle since DBImpl is always holding a
@@ -161,6 +189,22 @@ Status DB::OpenForReadOnly(
    const std::vector<ColumnFamilyDescriptor>& column_families,
    std::vector<ColumnFamilyHandle*>* handles, DB** dbptr,
    bool error_if_log_file_exist) {
+  // If dbname does not exist in the file system, should not do anything
+  Status s = OpenForReadOnlyCheckExistence(db_options, dbname);
+  if (!s.ok()) {
+    return s;
+  }
+
+  return DBImplReadOnly::OpenForReadOnlyWithoutCheck(
+      db_options, dbname, column_families, handles, dbptr,
+      error_if_log_file_exist);
+}
+
+Status DBImplReadOnly::OpenForReadOnlyWithoutCheck(
+    const DBOptions& db_options, const std::string& dbname,
+    const std::vector<ColumnFamilyDescriptor>& column_families,
+    std::vector<ColumnFamilyHandle*>* handles, DB** dbptr,
+    bool error_if_log_file_exist) {
  *dbptr = nullptr;
  handles->clear();

@@ -193,7 +237,7 @@ Status DB::OpenForReadOnly(
    *dbptr = impl;
    for (auto* h : *handles) {
      impl->NewThreadStatusCfInfo(
-          reinterpret_cast<ColumnFamilyHandleImpl*>(h)->cfd());
+          static_cast_with_check<ColumnFamilyHandleImpl>(h)->cfd());
    }
  } else {
    for (auto h : *handles) {
@@ -130,6 +130,15 @@ class DBImplReadOnly : public DBImpl {
  }

 private:
+  // A "helper" function for DB::OpenForReadOnly without column families
+  // to reduce unnecessary I/O
+  // It has the same functionality as DB::OpenForReadOnly with column families
+  // but does not check the existence of dbname in the file system
+  static Status OpenForReadOnlyWithoutCheck(
+      const DBOptions& db_options, const std::string& dbname,
+      const std::vector<ColumnFamilyDescriptor>& column_families,
+      std::vector<ColumnFamilyHandle*>* handles, DB** dbptr,
+      bool error_if_log_file_exist = false);
  friend class DB;
 };
 }  // namespace ROCKSDB_NAMESPACE
@@ -388,7 +388,7 @@ Iterator* DBImplSecondary::NewIterator(const ReadOptions& read_options,
        "ReadTier::kPersistedData is not yet supported in iterators."));
  }
  Iterator* result = nullptr;
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  auto cfd = cfh->cfd();
  ReadCallback* read_callback = nullptr;  // No read callback provided.
  if (read_options.tailing) {
@@ -642,7 +642,7 @@ Status DB::OpenAsSecondary(
    *dbptr = impl;
    for (auto h : *handles) {
      impl->NewThreadStatusCfInfo(
-          reinterpret_cast<ColumnFamilyHandleImpl*>(h)->cfd());
+          static_cast_with_check<ColumnFamilyHandleImpl>(h)->cfd());
    }
  } else {
    for (auto h : *handles) {
@@ -6,14 +6,15 @@
 // 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 "db/db_impl/db_impl.h"
 #include "db/error_handler.h"
 #include "db/event_helpers.h"
 #include "monitoring/perf_context_imp.h"
 #include "options/options_helper.h"
 #include "test_util/sync_point.h"
+#include "util/cast_util.h"

 namespace ROCKSDB_NAMESPACE {
 // Convenience methods
@@ -24,7 +25,7 @@ Status DBImpl::Put(const WriteOptions& o, ColumnFamilyHandle* column_family,

 Status DBImpl::Merge(const WriteOptions& o, ColumnFamilyHandle* column_family,
                     const Slice& key, const Slice& val) {
-  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
+  auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  if (!cfh->cfd()->ioptions()->merge_operator) {
    return Status::NotSupported("Provide a merge_operator when opening DB");
  } else {
@@ -1637,6 +1638,7 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
  std::unique_ptr<WritableFile> lfile;
  log::Writer* new_log = nullptr;
  MemTable* new_mem = nullptr;
+  IOStatus io_s;

  // Recoverable state is persisted in WAL. After memtable switch, WAL might
  // be deleted, so we write the state to memtable to be persisted as well.
@@ -1682,8 +1684,11 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
  if (creating_new_log) {
    // TODO: Write buffer size passed in should be max of all CF's instead
    // of mutable_cf_options.write_buffer_size.
-    s = CreateWAL(new_log_number, recycle_log_number, preallocate_block_size,
-                  &new_log);
+    io_s = CreateWAL(new_log_number, recycle_log_number, preallocate_block_size,
+                     &new_log);
+    if (s.ok()) {
+      s = io_s;
+    }
  }
  if (s.ok()) {
    SequenceNumber seq = versions_->LastSequence();
@@ -1709,7 +1714,10 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
    if (!logs_.empty()) {
      // Alway flush the buffer of the last log before switching to a new one
      log::Writer* cur_log_writer = logs_.back().writer;
-      s = cur_log_writer->WriteBuffer();
+      io_s = cur_log_writer->WriteBuffer();
+      if (s.ok()) {
+        s = io_s;
+      }
      if (!s.ok()) {
        ROCKS_LOG_WARN(immutable_db_options_.info_log,
                       "[%s] Failed to switch from #%" PRIu64 " to #%" PRIu64
@@ -1744,7 +1752,11 @@ 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
-    error_handler_.SetBGError(s, BackgroundErrorReason::kMemTable);
+    if (!io_s.ok()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kMemTable);
+    } else {
+      error_handler_.SetBGError(s, BackgroundErrorReason::kMemTable);
+    }
    // Read back bg_error in order to get the right severity
    s = error_handler_.GetBGError();
    return s;
@@ -1837,8 +1849,30 @@ Status DB::Put(const WriteOptions& opt, ColumnFamilyHandle* column_family,

 Status DB::Delete(const WriteOptions& opt, ColumnFamilyHandle* column_family,
                  const Slice& key) {
-  WriteBatch batch;
-  batch.Delete(column_family, key);
+  if (nullptr == opt.timestamp) {
+    WriteBatch batch;
+    Status s = batch.Delete(column_family, key);
+    if (!s.ok()) {
+      return s;
+    }
+    return Write(opt, &batch);
+  }
+  const Slice* ts = opt.timestamp;
+  assert(ts != nullptr);
+  const size_t ts_sz = ts->size();
+  constexpr size_t kKeyAndValueLenSize = 11;
+  constexpr size_t kWriteBatchOverhead =
+      WriteBatchInternal::kHeader + sizeof(ValueType) + kKeyAndValueLenSize;
+  WriteBatch batch(key.size() + ts_sz + kWriteBatchOverhead, /*max_bytes=*/0,
+                   ts_sz);
+  Status s = batch.Delete(column_family, key);
+  if (!s.ok()) {
+    return s;
+  }
+  s = batch.AssignTimestamp(*ts);
+  if (!s.ok()) {
+    return s;
+  }
  return Write(opt, &batch);
 }

@@ -10,8 +10,8 @@
 #include "db/db_impl/db_impl_secondary.h"
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
-#include "test_util/fault_injection_test_env.h"
 #include "test_util/sync_point.h"
+#include "utilities/fault_injection_env.h"

 namespace ROCKSDB_NAMESPACE {

@@ -17,7 +17,8 @@
 namespace ROCKSDB_NAMESPACE {

 void DumpDBFileSummary(const ImmutableDBOptions& options,
-                       const std::string& dbname) {
+                       const std::string& dbname,
+                       const std::string& session_id) {
  if (options.info_log == nullptr) {
    return;
  }
@@ -32,6 +33,8 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
  std::string file_info, wal_info;

  Header(options.info_log, "DB SUMMARY\n");
+  Header(options.info_log, "DB Session ID:  %s\n", session_id.c_str());
+
  // Get files in dbname dir
  if (!env->GetChildren(dbname, &files).ok()) {
    Error(options.info_log,
@@ -10,5 +10,6 @@

 namespace ROCKSDB_NAMESPACE {
 void DumpDBFileSummary(const ImmutableDBOptions& options,
-                       const std::string& dbname);
+                       const std::string& dbname,
+                       const std::string& session_id = "");
 }  // namespace ROCKSDB_NAMESPACE
@@ -9,6 +9,7 @@

 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/random.h"

 namespace ROCKSDB_NAMESPACE {

@@ -281,8 +282,8 @@ TEST_F(DBIOFailureTest, FlushSstRangeSyncError) {

  Random rnd(301);
  std::string rnd_str =
-      RandomString(&rnd, static_cast<int>(options.bytes_per_sync / 2));
-  std::string rnd_str_512kb = RandomString(&rnd, 512 * 1024);
+      rnd.RandomString(static_cast<int>(options.bytes_per_sync / 2));
+  std::string rnd_str_512kb = rnd.RandomString(512 * 1024);

  ASSERT_OK(Put(1, "foo", "bar"));
  // First 1MB doesn't get range synced
@@ -330,8 +331,8 @@ TEST_F(DBIOFailureTest, CompactSstRangeSyncError) {

  Random rnd(301);
  std::string rnd_str =
-      RandomString(&rnd, static_cast<int>(options.bytes_per_sync / 2));
-  std::string rnd_str_512kb = RandomString(&rnd, 512 * 1024);
+      rnd.RandomString(static_cast<int>(options.bytes_per_sync / 2));
+  std::string rnd_str_512kb = rnd.RandomString(512 * 1024);

  ASSERT_OK(Put(1, "foo", "bar"));
  // First 1MB doesn't get range synced
@@ -258,7 +258,7 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
      // prone to bugs causing the same user key with the same sequence number.
      // Note that with current timestamp implementation, the same user key can
      // have different timestamps and zero sequence number on the bottommost
-      // level. This will change in the future.
+      // level. This may change in the future.
      if ((!is_prev_key_seqnum_zero || timestamp_size_ > 0) &&
          skipping_saved_key &&
          CompareKeyForSkip(ikey_.user_key, saved_key_.GetUserKey()) <= 0) {
@@ -276,6 +276,7 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
        reseek_done = false;
        switch (ikey_.type) {
          case kTypeDeletion:
+          case kTypeDeletionWithTimestamp:
          case kTypeSingleDeletion:
            // Arrange to skip all upcoming entries for this key since
            // they are hidden by this deletion.
@@ -97,7 +97,8 @@ struct StressTestIterator : public InternalIterator {

  bool MaybeFail() {
    if (rnd->Next() >=
-        std::numeric_limits<uint64_t>::max() * error_probability) {
+        static_cast<double>(std::numeric_limits<uint64_t>::max()) *
+            error_probability) {
      return false;
    }
    if (rnd->Next() % 2) {
@@ -114,7 +115,8 @@ struct StressTestIterator : public InternalIterator {

  void MaybeMutate() {
    if (rnd->Next() >=
-        std::numeric_limits<uint64_t>::max() * mutation_probability) {
+        static_cast<double>(std::numeric_limits<uint64_t>::max()) *
+            mutation_probability) {
      return;
    }
    do {
@@ -126,8 +128,9 @@ struct StressTestIterator : public InternalIterator {
      if (data->hidden.empty()) {
        hide_probability = 1;
      }
-      bool do_hide =
-          rnd->Next() < std::numeric_limits<uint64_t>::max() * hide_probability;
+      bool do_hide = rnd->Next() <
+                     static_cast<double>(std::numeric_limits<uint64_t>::max()) *
+                         hide_probability;
      if (do_hide) {
        // Hide a random entry.
        size_t idx = rnd->Next() % data->entries.size();
@@ -17,6 +17,7 @@
 #include "rocksdb/iostats_context.h"
 #include "rocksdb/perf_context.h"
 #include "table/block_based/flush_block_policy.h"
+#include "util/random.h"

 namespace ROCKSDB_NAMESPACE {

@@ -40,7 +41,8 @@ class DBIteratorTest : public DBTestBase,
    if (column_family == nullptr) {
      column_family = db_->DefaultColumnFamily();
    }
-    auto* cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family)->cfd();
+    auto* cfd =
+        static_cast_with_check<ColumnFamilyHandleImpl>(column_family)->cfd();
    SequenceNumber seq = read_options.snapshot != nullptr
                             ? read_options.snapshot->GetSequenceNumber()
                             : db_->GetLatestSequenceNumber();
@@ -193,10 +195,10 @@ TEST_P(DBIteratorTest, IterReseekNewUpperBound) {
  options.compression = kNoCompression;
  Reopen(options);

-  ASSERT_OK(Put("a", RandomString(&rnd, 400)));
-  ASSERT_OK(Put("aabb", RandomString(&rnd, 400)));
-  ASSERT_OK(Put("aaef", RandomString(&rnd, 400)));
-  ASSERT_OK(Put("b", RandomString(&rnd, 400)));
+  ASSERT_OK(Put("a", rnd.RandomString(400)));
+  ASSERT_OK(Put("aabb", rnd.RandomString(400)));
+  ASSERT_OK(Put("aaef", rnd.RandomString(400)));
+  ASSERT_OK(Put("b", rnd.RandomString(400)));
  dbfull()->Flush(FlushOptions());
  ReadOptions opts;
  Slice ub = Slice("aa");
@@ -1359,7 +1361,7 @@ class DBIteratorTestForPinnedData : public DBIteratorTest {

    std::vector<std::string> generated_keys(key_pool);
    for (int i = 0; i < key_pool; i++) {
-      generated_keys[i] = RandomString(&rnd, key_size);
+      generated_keys[i] = rnd.RandomString(key_size);
    }

    std::map<std::string, std::string> true_data;
@@ -1367,7 +1369,7 @@ class DBIteratorTestForPinnedData : public DBIteratorTest {
    std::vector<std::string> deleted_keys;
    for (int i = 0; i < puts; i++) {
      auto& k = generated_keys[rnd.Next() % key_pool];
-      auto v = RandomString(&rnd, val_size);
+      auto v = rnd.RandomString(val_size);

      // Insert data to true_data map and to DB
      true_data[k] = v;
@@ -1530,7 +1532,7 @@ TEST_P(DBIteratorTest, PinnedDataIteratorMultipleFiles) {
  Random rnd(301);
  for (int i = 1; i <= 1000; i++) {
    std::string k = Key(i * 3);
-    std::string v = RandomString(&rnd, 100);
+    std::string v = rnd.RandomString(100);
    ASSERT_OK(Put(k, v));
    true_data[k] = v;
    if (i % 250 == 0) {
@@ -1544,7 +1546,7 @@ TEST_P(DBIteratorTest, PinnedDataIteratorMultipleFiles) {
  // Generate 4 sst files in L0
  for (int i = 1; i <= 1000; i++) {
    std::string k = Key(i * 2);
-    std::string v = RandomString(&rnd, 100);
+    std::string v = rnd.RandomString(100);
    ASSERT_OK(Put(k, v));
    true_data[k] = v;
    if (i % 250 == 0) {
@@ -1556,7 +1558,7 @@ TEST_P(DBIteratorTest, PinnedDataIteratorMultipleFiles) {
  // Add some keys/values in memtables
  for (int i = 1; i <= 1000; i++) {
    std::string k = Key(i);
-    std::string v = RandomString(&rnd, 100);
+    std::string v = rnd.RandomString(100);
    ASSERT_OK(Put(k, v));
    true_data[k] = v;
  }
@@ -1658,8 +1660,8 @@ TEST_P(DBIteratorTest, PinnedDataIteratorReadAfterUpdate) {

  std::map<std::string, std::string> true_data;
  for (int i = 0; i < 1000; i++) {
-    std::string k = RandomString(&rnd, 10);
-    std::string v = RandomString(&rnd, 1000);
+    std::string k = rnd.RandomString(10);
+    std::string v = rnd.RandomString(1000);
    ASSERT_OK(Put(k, v));
    true_data[k] = v;
  }
@@ -1673,7 +1675,7 @@ TEST_P(DBIteratorTest, PinnedDataIteratorReadAfterUpdate) {
    if (rnd.OneIn(2)) {
      ASSERT_OK(Delete(kv.first));
    } else {
-      std::string new_val = RandomString(&rnd, 1000);
+      std::string new_val = rnd.RandomString(1000);
      ASSERT_OK(Put(kv.first, new_val));
    }
  }
@@ -1930,7 +1932,7 @@ TEST_P(DBIteratorTest, IterPrevKeyCrossingBlocksRandomized) {

  for (int i = 0; i < kNumKeys; i++) {
    gen_key = Key(i);
-    gen_val = RandomString(&rnd, kValSize);
+    gen_val = rnd.RandomString(kValSize);

    ASSERT_OK(Put(gen_key, gen_val));
    true_data[gen_key] = gen_val;
@@ -1948,7 +1950,7 @@ TEST_P(DBIteratorTest, IterPrevKeyCrossingBlocksRandomized) {

    for (int j = 0; j < kNumMergeOperands; j++) {
      gen_key = Key(i);
-      gen_val = RandomString(&rnd, kValSize);
+      gen_val = rnd.RandomString(kValSize);

      ASSERT_OK(db_->Merge(WriteOptions(), gen_key, gen_val));
      true_data[gen_key] += "," + gen_val;
@@ -2048,7 +2050,7 @@ TEST_P(DBIteratorTest, IteratorWithLocalStatistics) {
  Random rnd(301);
  for (int i = 0; i < 1000; i++) {
    // Key 10 bytes / Value 10 bytes
-    ASSERT_OK(Put(RandomString(&rnd, 10), RandomString(&rnd, 10)));
+    ASSERT_OK(Put(rnd.RandomString(10), rnd.RandomString(10)));
  }

  std::atomic<uint64_t> total_next(0);
@@ -2704,7 +2706,7 @@ TEST_P(DBIteratorTest, AvoidReseekLevelIterator) {
  Reopen(options);

  Random rnd(301);
-  std::string random_str = RandomString(&rnd, 180);
+  std::string random_str = rnd.RandomString(180);

  ASSERT_OK(Put("1", random_str));
  ASSERT_OK(Put("2", random_str));
@@ -2911,7 +2913,7 @@ TEST_F(DBIteratorWithReadCallbackTest, ReadCallback) {

  SequenceNumber seq2 = db_->GetLatestSequenceNumber();
  auto* cfd =
-      reinterpret_cast<ColumnFamilyHandleImpl*>(db_->DefaultColumnFamily())
+      static_cast_with_check<ColumnFamilyHandleImpl>(db_->DefaultColumnFamily())
          ->cfd();
  // The iterator are suppose to see data before seq1.
  Iterator* iter =
@@ -303,14 +303,14 @@ TEST_F(DBMemTableTest, InsertWithHint) {
  ASSERT_EQ(hint_bar, rep->last_hint_in());
  ASSERT_EQ(hint_bar, rep->last_hint_out());
  ASSERT_EQ(5, rep->num_insert_with_hint());
-  ASSERT_OK(Put("whitelisted", "vvv"));
+  ASSERT_OK(Put("NotInPrefixDomain", "vvv"));
  ASSERT_EQ(5, rep->num_insert_with_hint());
  ASSERT_EQ("foo_v1", Get("foo_k1"));
  ASSERT_EQ("foo_v2", Get("foo_k2"));
  ASSERT_EQ("foo_v3", Get("foo_k3"));
  ASSERT_EQ("bar_v1", Get("bar_k1"));
  ASSERT_EQ("bar_v2", Get("bar_k2"));
-  ASSERT_EQ("vvv", Get("whitelisted"));
+  ASSERT_EQ("vvv", Get("NotInPrefixDomain"));
 }

 TEST_F(DBMemTableTest, ColumnFamilyId) {
@@ -8,11 +8,11 @@
 #include "rocksdb/perf_context.h"
 #include "rocksdb/utilities/debug.h"
 #include "table/block_based/block_builder.h"
-#include "test_util/fault_injection_test_env.h"
 #if !defined(ROCKSDB_LITE)
 #include "test_util/sync_point.h"
 #endif
 #include "rocksdb/merge_operator.h"
+#include "utilities/fault_injection_env.h"
 #include "utilities/merge_operators.h"
 #include "utilities/merge_operators/sortlist.h"
 #include "utilities/merge_operators/string_append/stringappend2.h"