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

.circleci/cat_ignore_eagain

@@ -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

.circleci/config.yml

@@ -11,20 +11,201 @@ executors:
       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 TEST_UINT128_COMPAT=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: echo "APT::Acquire::Retries \"10\";" | sudo tee -a /etc/apt/apt.conf.d/80-retries # llvm.org unreliable
+      - 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: echo "APT::Acquire::Retries \"10\";" | sudo tee -a /etc/apt/apt.conf.d/80-retries # llvm.org unreliable
+      - 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: echo "APT::Acquire::Retries \"10\";" | sudo tee -a /etc/apt/apt.conf.d/80-retries # llvm.org unreliable
+      - 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: echo "APT::Acquire::Retries \"10\";" | sudo tee -a /etc/apt/apt.conf.d/80-retries # llvm.org unreliable
+      - 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: echo "APT::Acquire::Retries \"10\";" | sudo tee -a /etc/apt/apt.conf.d/80-retries # llvm.org unreliable
+      - 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-linux-gcc-4-8:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: large
+    steps:
+      - checkout
+      - run: pyenv global 3.5.2
+      - run: sudo apt-get update -y && sudo apt-get install gcc-4.8 g++-4.8 libgflags-dev
+      - run: CC=gcc-4.8 CXX=g++-4.8 V=1 SKIP_LINK=1 make -j4 all | .circleci/cat_ignore_eagain # Linking broken because libgflags compiled with newer ABI
+
+  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 +227,7 @@ 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 +235,141 @@ 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
+
+  build-linux-non-shm:
+    machine:
+      image: ubuntu-1604:201903-01
+    resource_class: 2xlarge
+    parameters:
+      start_test:
+        default: ""
+        type: string
+      end_test:
+        default: ""
+        type: string
+    steps:
+      - checkout # checkout 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 and unit tests on non-shm"
+          command: |
+            TMPD=/tmp/rocksdb_test_tmp ROCKSDBTESTS_START=<<parameters.start_test>> ROCKSDBTESTS_END=<<parameters.end_test>> make V=1 J=32 -j32 check_some | .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
+  build-linux-non-shm-1:
+    jobs:
+      - build-linux-non-shm:
+          start_test: ""
+          end_test: "db_tailing_iter_test" # make sure unique in src.mk
+  build-linux-non-shm-2:
+    jobs:
+      - build-linux-non-shm:
+          start_test: "db_tailing_iter_test" # make sure unique in src.mk
+          end_test: "db_test2" # make sure unique in src.mk
+  build-linux-non-shm-3:
+    jobs:
+      - build-linux-non-shm:
+          start_test: "db_test2" # make sure unique in src.mk
+          end_test: "arena_test" # make sure unique in src.mk
+  build-linux-non-shm-4:
+    jobs:
+      - build-linux-non-shm:
+          start_test: "compact_on_deletion_collector_test" # make sure unique in src.mk
+          end_test: ""
+  build-linux-gcc-4-8:
+    jobs:
+      - build-linux-gcc-4-8

.circleci/ubsan_suppression_list.txt

@@ -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

.circleci/vs2015_install.ps1

@@ -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
+}

.circleci/vs2017_install.ps1

@@ -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
+}

.gitignore

@@ -1,4 +1,5 @@
 make_config.mk
+rocksdb.pc
 
 *.a
 *.arc
@@ -52,6 +53,7 @@ db_test2
 trace_analyzer
 trace_analyzer_test
 block_cache_trace_analyzer
+io_tracer_parser
 .DS_Store
 .vs
 .vscode

.travis.yml

@@ -87,134 +87,134 @@ matrix:
   - os: linux
     compiler: clang
   # Exclude all but most unique cmake variants for pull requests, but build all in branches
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: amd64
     env: JOB_NAME=cmake
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: amd64
     env: JOB_NAME=cmake-gcc8
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: amd64
     env: JOB_NAME=cmake-gcc9
   # Exclude most osx, arm64 and ppc64le tests for pull requests, but build in branches
   # Temporarily disable ppc64le unit tests in PRs until Travis gets its act together (#6653)
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: TEST_GROUP=platform_dependent
   # NB: the cmake build is a partial java test
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: osx
     env: TEST_GROUP=1
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: TEST_GROUP=1
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: TEST_GROUP=1
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: osx
     env: TEST_GROUP=2
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: TEST_GROUP=2
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: TEST_GROUP=2
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: osx
     env: TEST_GROUP=3
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: TEST_GROUP=3
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: TEST_GROUP=3
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: osx
     env: TEST_GROUP=4
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: TEST_GROUP=4
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: TEST_GROUP=4
-  - if: type = pull_request AND commit_message !~ /java/
+  - if: type = pull_request AND commit_message !~ /FULL_CI/ AND commit_message !~ /java/
     os : osx
     env: JOB_NAME=java_test
-  - if: type = pull_request AND commit_message !~ /java/
+  - if: type = pull_request AND commit_message !~ /FULL_CI/ AND commit_message !~ /java/
     os : linux
     arch: arm64
     env: JOB_NAME=java_test
-  - if: type = pull_request AND commit_message !~ /java/
+  - if: type = pull_request AND commit_message !~ /FULL_CI/ AND commit_message !~ /java/
     os: linux
     arch: ppc64le
     env: JOB_NAME=java_test
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : osx
     env: JOB_NAME=lite_build
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: JOB_NAME=lite_build
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: JOB_NAME=lite_build
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : osx
     env: JOB_NAME=examples
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: JOB_NAME=examples
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: JOB_NAME=examples
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: JOB_NAME=cmake-gcc8
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: JOB_NAME=cmake-gcc8
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: JOB_NAME=cmake-gcc9
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: JOB_NAME=cmake-gcc9
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: JOB_NAME=cmake-gcc9-c++20
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: JOB_NAME=cmake-gcc9-c++20
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : osx
     env: JOB_NAME=status_checked
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os : linux
     arch: arm64
     env: JOB_NAME=status_checked
-  - if: type = pull_request
+  - if: type = pull_request AND commit_message !~ /FULL_CI/
     os: linux
     arch: ppc64le
     env: JOB_NAME=status_checked
@@ -224,38 +224,33 @@ install:
       PATH=$PATH:/usr/local/opt/ccache/libexec;
     fi
   - if [ "${JOB_NAME}" == cmake-gcc8 ]; then
-      sudo apt-get install -y g++-8;
+      sudo apt-get install -y g++-8 || exit $?;
       CC=gcc-8 && CXX=g++-8;
     fi
   - if [ "${JOB_NAME}" == cmake-gcc9 ] || [ "${JOB_NAME}" == cmake-gcc9-c++20 ]; then
-      sudo apt-get install -y g++-9;
+      sudo apt-get install -y g++-9 || exit $?;
       CC=gcc-9 && CXX=g++-9;
     fi
   - if [ "${JOB_NAME}" == cmake-mingw ]; then
-      sudo apt-get install -y mingw-w64 ;
+      sudo apt-get install -y mingw-w64 || exit $?;
     fi
   - if [ "${JOB_NAME}" == make-gcc4.8 ]; then
-      sudo apt-get install -y g++-4.8 ;
+      sudo apt-get install -y g++-4.8 || exit $?;
       CC=gcc-4.8 && CXX=g++-4.8;
     fi
   - if [[ "${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 apt-get install snapd && sudo snap install cmake --beta --classic || exit $?;
+      export PATH=/snap/bin:$PATH;
     fi
   - |
     if [[ "${JOB_NAME}" == java_test || "${JOB_NAME}" == cmake* ]]; then
       # Ensure JDK 8
       if [ "${TRAVIS_OS_NAME}" == osx ]; then
-        brew tap AdoptOpenJDK/openjdk
-        brew cask install adoptopenjdk8
+        brew tap AdoptOpenJDK/openjdk || exit $?
+        brew cask install adoptopenjdk8 || exit $?
         export JAVA_HOME=$(/usr/libexec/java_home)
       else
-        sudo apt-get install -y openjdk-8-jdk
+        sudo apt-get install -y openjdk-8-jdk || exit $?
         export PATH=/usr/lib/jvm/java-8-openjdk-$(dpkg --print-architecture)/bin:$PATH
         export JAVA_HOME=/usr/lib/jvm/java-8-openjdk-$(dpkg --print-architecture)
       fi
@@ -274,19 +269,19 @@ script:
   - if [ `command -v ccache` ]; then ccache -C; fi
   - case $TEST_GROUP in
     platform_dependent)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_END=db_block_cache_test make -j4 all_but_some_tests check_some
+      OPT=-DTRAVIS V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=only make -j4 all_but_some_tests check_some
       ;;
     1)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=db_block_cache_test ROCKSDBTESTS_END=db_iter_test make -j4 check_some
+      OPT=-DTRAVIS V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_END=backupable_db_test make -j4 check_some
       ;;
     2)
-      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
+      OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" V=1 make -j4 tools && OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=backupable_db_test ROCKSDBTESTS_END=db_universal_compaction_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_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=db_universal_compaction_test ROCKSDBTESTS_END=table_properties_collector_test make -j4 check_some
       ;;
     4)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=write_prepared_transaction_test make -j4 check_some
+      OPT=-DTRAVIS V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=table_properties_collector_test make -j4 check_some
       ;;
     esac
   - case $JOB_NAME in

CMakeLists.txt

@@ -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)
@@ -107,6 +107,7 @@ else()
     endif()
   endif()
 
+  set(GFLAGS_LIB)
   if(WITH_GFLAGS)
     # Config with namespace available since gflags 2.2.2
     option(GFLAGS_USE_TARGET_NAMESPACE "Use gflags import target with namespace." ON)
@@ -114,15 +115,17 @@ else()
     if(gflags_FOUND)
       if(TARGET ${GFLAGS_TARGET})
         # Config with GFLAGS_TARGET available since gflags 2.2.0
-        list(APPEND THIRDPARTY_LIBS ${GFLAGS_TARGET})
+        set(GFLAGS_LIB ${GFLAGS_TARGET})
       else()
         # Config with GFLAGS_LIBRARIES available since gflags 2.1.0
-        list(APPEND THIRDPARTY_LIBS ${GFLAGS_LIBRARIES})
+        set(GFLAGS_LIB ${GFLAGS_LIBRARIES})
       endif()
     else()
       find_package(gflags REQUIRED)
-      list(APPEND THIRDPARTY_LIBS gflags::gflags)
+        set(GFLAGS_LIB gflags::gflags)
     endif()
+    include_directories(${GFLAGS_INCLUDE_DIR})
+    list(APPEND THIRDPARTY_LIBS ${GFLAGS_LIB})
     add_definitions(-DGFLAGS=1)
   endif()
 
@@ -251,12 +254,30 @@ endif(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|AARCH64")
 
 option(PORTABLE "build a portable binary" OFF)
 option(FORCE_SSE42 "force building with SSE4.2, even when PORTABLE=ON" OFF)
+option(FORCE_AVX "force building with AVX, even when PORTABLE=ON" OFF)
+option(FORCE_AVX2 "force building with AVX2, even when PORTABLE=ON" OFF)
 if(PORTABLE)
   # MSVC does not need a separate compiler flag to enable SSE4.2; if nmmintrin.h
   # is available, it is available by default.
   if(FORCE_SSE42 AND NOT MSVC)
     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -msse4.2 -mpclmul")
   endif()
+  if(MSVC)
+    if(FORCE_AVX)
+      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /arch:AVX")
+    endif()
+    # MSVC automatically enables BMI / lzcnt with AVX2.
+    if(FORCE_AVX2)
+      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /arch:AVX2")
+    endif()
+  else()
+    if(FORCE_AVX)
+      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mavx")
+    endif()
+    if(FORCE_AVX2)
+      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mavx2 -mbmi -mlzcnt")
+    endif()
+  endif()
 else()
   if(MSVC)
     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /arch:AVX2")
@@ -448,6 +469,8 @@ 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")
@@ -506,7 +529,11 @@ 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()
@@ -537,8 +564,12 @@ set(SOURCES
         cache/sharded_cache.cc
         db/arena_wrapped_db_iter.cc
         db/blob/blob_file_addition.cc
+        db/blob/blob_file_builder.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
@@ -550,6 +581,7 @@ 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
@@ -594,6 +626,7 @@ set(SOURCES
         db/version_edit.cc
         db/version_edit_handler.cc
         db/version_set.cc
+        db/wal_edit.cc
         db/wal_manager.cc
         db/write_batch.cc
         db/write_batch_base.cc
@@ -604,6 +637,7 @@ set(SOURCES
         env/env_encryption.cc
         env/env_hdfs.cc
         env/file_system.cc
+        env/file_system_tracer.cc
         env/mock_env.cc
         file/delete_scheduler.cc
         file/file_prefetch_buffer.cc
@@ -637,6 +671,7 @@ set(SOURCES
         monitoring/perf_level.cc
         monitoring/persistent_stats_history.cc
         monitoring/statistics.cc
+        monitoring/stats_dump_scheduler.cc
         monitoring/thread_status_impl.cc
         monitoring/thread_status_updater.cc
         monitoring/thread_status_util.cc
@@ -689,6 +724,7 @@ 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
@@ -705,6 +741,7 @@ set(SOURCES
         tools/trace_analyzer_tool.cc
         trace_replay/trace_replay.cc
         trace_replay/block_cache_tracer.cc
+        trace_replay/io_tracer.cc
         util/coding.cc
         util/compaction_job_stats_impl.cc
         util/comparator.cc
@@ -730,9 +767,6 @@ 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
@@ -741,6 +775,8 @@ 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
@@ -762,6 +798,8 @@ set(SOURCES
         utilities/simulator_cache/sim_cache.cc
         utilities/table_properties_collectors/compact_on_deletion_collector.cc
         utilities/trace/file_trace_reader_writer.cc
+        utilities/transactions/lock/lock_tracker.cc
+        utilities/transactions/lock/point_lock_tracker.cc
         utilities/transactions/optimistic_transaction_db_impl.cc
         utilities/transactions/optimistic_transaction.cc
         utilities/transactions/pessimistic_transaction.cc
@@ -974,15 +1012,27 @@ if(NOT WIN32 OR ROCKSDB_INSTALL_ON_WINDOWS)
   )
 endif()
 
-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/mock_time_env.cc
   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
+        db/blob/blob_file_builder_test.cc
         db/blob/blob_file_garbage_test.cc
         db/blob/db_blob_index_test.cc
         db/column_family_test.cc
@@ -994,7 +1044,6 @@ 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
@@ -1054,14 +1103,15 @@ if(WITH_TESTS)
         db/version_edit_test.cc
         db/version_set_test.cc
         db/wal_manager_test.cc
+        db/wal_edit_test.cc
         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
         file/delete_scheduler_test.cc
+        file/prefetch_test.cc
         file/random_access_file_reader_test.cc
         logging/auto_roll_logger_test.cc
         logging/env_logger_test.cc
@@ -1074,6 +1124,7 @@ if(WITH_TESTS)
         monitoring/histogram_test.cc
         monitoring/iostats_context_test.cc
         monitoring/statistics_test.cc
+        monitoring/stats_dump_scheduler_test.cc
         monitoring/stats_history_test.cc
         options/options_settable_test.cc
         options/options_test.cc
@@ -1090,6 +1141,8 @@ if(WITH_TESTS)
         table/sst_file_reader_test.cc
         table/table_test.cc
         table/block_fetcher_test.cc
+        test_util/testutil_test.cc
+        trace_replay/io_tracer_test.cc
         tools/block_cache_analyzer/block_cache_trace_analyzer_test.cc
         tools/ldb_cmd_test.cc
         tools/reduce_levels_test.cc
@@ -1139,7 +1192,8 @@ 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()
@@ -1152,8 +1206,6 @@ 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()
@@ -1179,7 +1231,7 @@ 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 ${THIRDPARTY_LIBS} ${ROCKSDB_LIB})
       if(NOT "${exename}" MATCHES "db_sanity_test")
         add_test(NAME ${exename} COMMAND ${exename}${ARTIFACT_SUFFIX})
         add_dependencies(check ${CMAKE_PROJECT_NAME}_${exename}${ARTIFACT_SUFFIX})
@@ -1203,9 +1255,8 @@ if(WITH_TESTS)
 
   if(ROCKSDB_LIB_FOR_C)
     set(C_TESTS db/c_test.c)
-    # C executables must link to a shared object
     add_executable(c_test db/c_test.c)
-    target_link_libraries(c_test ${ROCKSDB_SHARED_LIB} testharness)
+    target_link_libraries(c_test ${ROCKSDB_LIB_FOR_C} testharness)
     add_test(NAME c_test COMMAND c_test${ARTIFACT_SUFFIX})
     add_dependencies(check c_test)
   endif()
@@ -1216,37 +1267,37 @@ if(WITH_BENCHMARK_TOOLS)
     tools/db_bench.cc
     tools/db_bench_tool.cc)
   target_link_libraries(db_bench
-    ${ROCKSDB_LIB})
+    ${ROCKSDB_LIB} ${THIRDPARTY_LIBS})
 
   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()
 
 if(WITH_CORE_TOOLS OR WITH_TOOLS)
@@ -1260,3 +1311,8 @@ 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()

HISTORY.md

@@ -1,8 +1,105 @@
 # Rocksdb Change Log
-## Unreleased
-### Behavior Changes
-* Disable delete triggered compaction (NewCompactOnDeletionCollectorFactory) in universal compaction mode and num_levels = 1 in order to avoid a corruption bug.
+## 6.13.4 (11/15/2020)
+### Bug Fixes
+* Fixed 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.
 
+## 6.13.3 (10/14/2020)
+### Bug Fixes
+* Fix a bug that could cause a stalled write to crash with mixed of slowdown and no_slowdown writes (`WriteOptions.no_slowdown=true`).
+
+## 6.13.2 (10/13/2020)
+### 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.
+
+## 6.13.1 (10/12/2020)
+### 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.
+
+## 6.13 (09/24/2020)
+### Bug fixes
+* Fix a performance regression introduced in 6.4 that makes a upper bound check for every Next() even if keys are within a data block that is within the upper bound.
+* Fix a possible corruption to the LSM state (overlapping files within a level) when a `CompactRange()` for refitting levels (`CompactRangeOptions::change_level == true`) and another manual compaction are executed in parallel.
+* Sanitize `recycle_log_file_num` to zero when the user attempts to enable it in combination with `WALRecoveryMode::kTolerateCorruptedTailRecords`. Previously the two features were allowed together, which compromised the user's configured crash-recovery guarantees.
+* Fix a bug where a level refitting in CompactRange() might race with an automatic compaction that puts the data to the target level of the refitting. The bug has been there for years.
+* Fixed a bug in version 6.12 in which BackupEngine::CreateNewBackup could fail intermittently with non-OK status when backing up a read-write DB configured with a DBOptions::file_checksum_gen_factory.
+* Fix useless no-op compactions scheduled upon snapshot release when options.disable-auto-compactions = true.
+* Fix a bug when max_write_buffer_size_to_maintain is set, immutable flushed memtable destruction is delayed until the next super version is installed. A memtable is not added to delete list because of its reference hold by super version and super version doesn't switch because of empt delete list. So memory usage keeps on increasing beyond write_buffer_size + max_write_buffer_size_to_maintain.
+* Avoid converting MERGES to PUTS when allow_ingest_behind is true.
+* Fix compression dictionary sampling together with `SstFileWriter`. Previously, the dictionary would be trained/finalized immediately with zero samples. Now, the whole `SstFileWriter` file is buffered in memory and then sampled.
+* Fix a bug with `avoid_unnecessary_blocking_io=1` and creating backups (BackupEngine::CreateNewBackup) or checkpoints (Checkpoint::Create). With this setting and WAL enabled, these operations could randomly fail with non-OK status.
+* Fix a bug in which bottommost compaction continues to advance the underlying InternalIterator to skip tombstones even after shutdown.
+
+### New Features
+* A new field `std::string requested_checksum_func_name` is added to `FileChecksumGenContext`, which enables the checksum factory to create generators for a suite of different functions.
+* Added a new subcommand, `ldb unsafe_remove_sst_file`, which removes a lost or corrupt SST file from a DB's metadata. This command involves data loss and must not be used on a live DB.
+
+### Performance Improvements
+* Reduce thread number for multiple DB instances by re-using one global thread for statistics dumping and persisting.
+* Reduce write-amp in heavy write bursts in `kCompactionStyleLevel` compaction style with `level_compaction_dynamic_level_bytes` set.
+* BackupEngine incremental backups no longer read DB table files that are already saved to a shared part of the backup directory, unless `share_files_with_checksum` is used with `kLegacyCrc32cAndFileSize` naming (discouraged).
+  * For `share_files_with_checksum`, we are confident there is no regression (vs. pre-6.12) in detecting DB or backup corruption at backup creation time, mostly because the old design did not leverage this extra checksum computation for detecting inconsistencies at backup creation time.
+  * For `share_table_files` without "checksum" (not recommended), there is a regression in detecting fundamentally unsafe use of the option, greatly mitigated by file size checking (under "Behavior Changes"). Almost no reason to use `share_files_with_checksum=false` should remain.
+  * `DB::VerifyChecksum` and `BackupEngine::VerifyBackup` with checksum checking are still able to catch corruptions that `CreateNewBackup` does not.
+
+### Public API Change
+* Expose kTypeDeleteWithTimestamp in EntryType and update GetEntryType() accordingly.
+* Added file_checksum and file_checksum_func_name to TableFileCreationInfo, which can pass the table file checksum information through the OnTableFileCreated callback during flush and compaction.
+* A warning is added to `DB::DeleteFile()` API describing its known problems and deprecation plan.
+* Add a new stats level, i.e. StatsLevel::kExceptTickers (PR7329) to exclude tickers even if application passes a non-null Statistics object.
+* Added a new status code IOStatus::IOFenced() for the Env/FileSystem to indicate that writes from this instance are fenced off. Like any other background error, this error is returned to the user in Put/Merge/Delete/Flush calls and can be checked using Status::IsIOFenced().
+
+### Behavior Changes
+* File abstraction `FSRandomAccessFile.Prefetch()` default return status is changed from `OK` to `NotSupported`. If the user inherited file doesn't implement prefetch, RocksDB will create internal prefetch buffer to improve read performance.
+* When retryable IO error happens during Flush (manifest write error is excluded) and WAL is disabled, originally it is mapped to kHardError. Now,it is mapped to soft error. So DB will not stall the writes unless the memtable is full. At the same time, when auto resume is triggered to recover the retryable IO error during Flush, SwitchMemtable is not called to avoid generating to many small immutable memtables. If WAL is enabled, no behavior changes.
+* When considering whether a table file is already backed up in a shared part of backup directory, BackupEngine would already query the sizes of source (DB) and pre-existing destination (backup) files. BackupEngine now uses these file sizes to detect corruption, as at least one of (a) old backup, (b) backup in progress, or (c) current DB is corrupt if there's a size mismatch.
+
+### Others
+* Error in prefetching partitioned index blocks will not be swallowed. It will fail the query and return the IOError users.
+
+## 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.
+* 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.
+
+### 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.
@@ -12,6 +109,9 @@
 * 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.
@@ -20,11 +120,19 @@
 * 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 Feature
+### 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

LANGUAGE-BINDINGS.md

@@ -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

README.md

@@ -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.

TARGETS

@@ -78,7 +78,6 @@ ROCKSDB_PREPROCESSOR_FLAGS = [
     # Directories with files for #include
     "-I" + REPO_PATH + "include/",
     "-I" + REPO_PATH,
-    "-I" + REPO_PATH + "third-party/gtest-1.8.1/fused-src/",
 ]
 
 ROCKSDB_ARCH_PREPROCESSOR_FLAGS = {
@@ -123,8 +122,12 @@ cpp_library(
         "cache/sharded_cache.cc",
         "db/arena_wrapped_db_iter.cc",
         "db/blob/blob_file_addition.cc",
+        "db/blob/blob_file_builder.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",
@@ -136,6 +139,7 @@ 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",
@@ -180,6 +184,7 @@ cpp_library(
         "db/version_edit.cc",
         "db/version_edit_handler.cc",
         "db/version_set.cc",
+        "db/wal_edit.cc",
         "db/wal_manager.cc",
         "db/write_batch.cc",
         "db/write_batch_base.cc",
@@ -191,6 +196,7 @@ 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",
@@ -226,6 +232,7 @@ cpp_library(
         "monitoring/perf_level.cc",
         "monitoring/persistent_stats_history.cc",
         "monitoring/statistics.cc",
+        "monitoring/stats_dump_scheduler.cc",
         "monitoring/thread_status_impl.cc",
         "monitoring/thread_status_updater.cc",
         "monitoring/thread_status_updater_debug.cc",
@@ -280,6 +287,7 @@ 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",
@@ -292,6 +300,7 @@ 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",
@@ -319,9 +328,6 @@ 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",
@@ -331,6 +337,8 @@ 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",
@@ -352,6 +360,8 @@ cpp_library(
         "utilities/simulator_cache/sim_cache.cc",
         "utilities/table_properties_collectors/compact_on_deletion_collector.cc",
         "utilities/trace/file_trace_reader_writer.cc",
+        "utilities/transactions/lock/lock_tracker.cc",
+        "utilities/transactions/lock/point_lock_tracker.cc",
         "utilities/transactions/optimistic_transaction.cc",
         "utilities/transactions/optimistic_transaction_db_impl.cc",
         "utilities/transactions/pessimistic_transaction.cc",
@@ -377,6 +387,284 @@ cpp_library(
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
     deps = [],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
+    link_whole = False,
+)
+
+cpp_library(
+    name = "rocksdb_whole_archive_lib",
+    srcs = [
+        "cache/cache.cc",
+        "cache/clock_cache.cc",
+        "cache/lru_cache.cc",
+        "cache/sharded_cache.cc",
+        "db/arena_wrapped_db_iter.cc",
+        "db/blob/blob_file_addition.cc",
+        "db/blob/blob_file_builder.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",
+        "db/compacted_db_impl.cc",
+        "db/compaction/compaction.cc",
+        "db/compaction/compaction_iterator.cc",
+        "db/compaction/compaction_job.cc",
+        "db/compaction/compaction_picker.cc",
+        "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",
+        "db/db_impl/db_impl_compaction_flush.cc",
+        "db/db_impl/db_impl_debug.cc",
+        "db/db_impl/db_impl_experimental.cc",
+        "db/db_impl/db_impl_files.cc",
+        "db/db_impl/db_impl_open.cc",
+        "db/db_impl/db_impl_readonly.cc",
+        "db/db_impl/db_impl_secondary.cc",
+        "db/db_impl/db_impl_write.cc",
+        "db/db_info_dumper.cc",
+        "db/db_iter.cc",
+        "db/dbformat.cc",
+        "db/error_handler.cc",
+        "db/event_helpers.cc",
+        "db/experimental.cc",
+        "db/external_sst_file_ingestion_job.cc",
+        "db/file_indexer.cc",
+        "db/flush_job.cc",
+        "db/flush_scheduler.cc",
+        "db/forward_iterator.cc",
+        "db/import_column_family_job.cc",
+        "db/internal_stats.cc",
+        "db/log_reader.cc",
+        "db/log_writer.cc",
+        "db/logs_with_prep_tracker.cc",
+        "db/malloc_stats.cc",
+        "db/memtable.cc",
+        "db/memtable_list.cc",
+        "db/merge_helper.cc",
+        "db/merge_operator.cc",
+        "db/range_del_aggregator.cc",
+        "db/range_tombstone_fragmenter.cc",
+        "db/repair.cc",
+        "db/snapshot_impl.cc",
+        "db/table_cache.cc",
+        "db/table_properties_collector.cc",
+        "db/transaction_log_impl.cc",
+        "db/trim_history_scheduler.cc",
+        "db/version_builder.cc",
+        "db/version_edit.cc",
+        "db/version_edit_handler.cc",
+        "db/version_set.cc",
+        "db/wal_edit.cc",
+        "db/wal_manager.cc",
+        "db/write_batch.cc",
+        "db/write_batch_base.cc",
+        "db/write_controller.cc",
+        "db/write_thread.cc",
+        "env/env.cc",
+        "env/env_chroot.cc",
+        "env/env_encryption.cc",
+        "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",
+        "file/delete_scheduler.cc",
+        "file/file_prefetch_buffer.cc",
+        "file/file_util.cc",
+        "file/filename.cc",
+        "file/random_access_file_reader.cc",
+        "file/read_write_util.cc",
+        "file/readahead_raf.cc",
+        "file/sequence_file_reader.cc",
+        "file/sst_file_manager_impl.cc",
+        "file/writable_file_writer.cc",
+        "logging/auto_roll_logger.cc",
+        "logging/event_logger.cc",
+        "logging/log_buffer.cc",
+        "memory/arena.cc",
+        "memory/concurrent_arena.cc",
+        "memory/jemalloc_nodump_allocator.cc",
+        "memory/memkind_kmem_allocator.cc",
+        "memtable/alloc_tracker.cc",
+        "memtable/hash_linklist_rep.cc",
+        "memtable/hash_skiplist_rep.cc",
+        "memtable/skiplistrep.cc",
+        "memtable/vectorrep.cc",
+        "memtable/write_buffer_manager.cc",
+        "monitoring/histogram.cc",
+        "monitoring/histogram_windowing.cc",
+        "monitoring/in_memory_stats_history.cc",
+        "monitoring/instrumented_mutex.cc",
+        "monitoring/iostats_context.cc",
+        "monitoring/perf_context.cc",
+        "monitoring/perf_level.cc",
+        "monitoring/persistent_stats_history.cc",
+        "monitoring/statistics.cc",
+        "monitoring/stats_dump_scheduler.cc",
+        "monitoring/thread_status_impl.cc",
+        "monitoring/thread_status_updater.cc",
+        "monitoring/thread_status_updater_debug.cc",
+        "monitoring/thread_status_util.cc",
+        "monitoring/thread_status_util_debug.cc",
+        "options/cf_options.cc",
+        "options/db_options.cc",
+        "options/options.cc",
+        "options/options_helper.cc",
+        "options/options_parser.cc",
+        "port/port_posix.cc",
+        "port/stack_trace.cc",
+        "table/adaptive/adaptive_table_factory.cc",
+        "table/block_based/binary_search_index_reader.cc",
+        "table/block_based/block.cc",
+        "table/block_based/block_based_filter_block.cc",
+        "table/block_based/block_based_table_builder.cc",
+        "table/block_based/block_based_table_factory.cc",
+        "table/block_based/block_based_table_iterator.cc",
+        "table/block_based/block_based_table_reader.cc",
+        "table/block_based/block_builder.cc",
+        "table/block_based/block_prefetcher.cc",
+        "table/block_based/block_prefix_index.cc",
+        "table/block_based/data_block_footer.cc",
+        "table/block_based/data_block_hash_index.cc",
+        "table/block_based/filter_block_reader_common.cc",
+        "table/block_based/filter_policy.cc",
+        "table/block_based/flush_block_policy.cc",
+        "table/block_based/full_filter_block.cc",
+        "table/block_based/hash_index_reader.cc",
+        "table/block_based/index_builder.cc",
+        "table/block_based/index_reader_common.cc",
+        "table/block_based/parsed_full_filter_block.cc",
+        "table/block_based/partitioned_filter_block.cc",
+        "table/block_based/partitioned_index_iterator.cc",
+        "table/block_based/partitioned_index_reader.cc",
+        "table/block_based/reader_common.cc",
+        "table/block_based/uncompression_dict_reader.cc",
+        "table/block_fetcher.cc",
+        "table/cuckoo/cuckoo_table_builder.cc",
+        "table/cuckoo/cuckoo_table_factory.cc",
+        "table/cuckoo/cuckoo_table_reader.cc",
+        "table/format.cc",
+        "table/get_context.cc",
+        "table/iterator.cc",
+        "table/merging_iterator.cc",
+        "table/meta_blocks.cc",
+        "table/persistent_cache_helper.cc",
+        "table/plain/plain_table_bloom.cc",
+        "table/plain/plain_table_builder.cc",
+        "table/plain/plain_table_factory.cc",
+        "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",
+        "table/two_level_iterator.cc",
+        "test_util/sync_point.cc",
+        "test_util/sync_point_impl.cc",
+        "test_util/transaction_test_util.cc",
+        "tools/dump/db_dump_tool.cc",
+        "tools/ldb_cmd.cc",
+        "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",
+        "util/compaction_job_stats_impl.cc",
+        "util/comparator.cc",
+        "util/compression_context_cache.cc",
+        "util/concurrent_task_limiter_impl.cc",
+        "util/crc32c.cc",
+        "util/dynamic_bloom.cc",
+        "util/file_checksum_helper.cc",
+        "util/hash.cc",
+        "util/murmurhash.cc",
+        "util/random.cc",
+        "util/rate_limiter.cc",
+        "util/slice.cc",
+        "util/status.cc",
+        "util/string_util.cc",
+        "util/thread_local.cc",
+        "util/threadpool_imp.cc",
+        "util/xxhash.cc",
+        "utilities/backupable/backupable_db.cc",
+        "utilities/blob_db/blob_compaction_filter.cc",
+        "utilities/blob_db/blob_db.cc",
+        "utilities/blob_db/blob_db_impl.cc",
+        "utilities/blob_db/blob_db_impl_filesnapshot.cc",
+        "utilities/blob_db/blob_dump_tool.cc",
+        "utilities/blob_db/blob_file.cc",
+        "utilities/cassandra/cassandra_compaction_filter.cc",
+        "utilities/cassandra/format.cc",
+        "utilities/cassandra/merge_operator.cc",
+        "utilities/checkpoint/checkpoint_impl.cc",
+        "utilities/compaction_filters/remove_emptyvalue_compactionfilter.cc",
+        "utilities/convenience/info_log_finder.cc",
+        "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",
+        "utilities/merge_operators/max.cc",
+        "utilities/merge_operators/put.cc",
+        "utilities/merge_operators/sortlist.cc",
+        "utilities/merge_operators/string_append/stringappend.cc",
+        "utilities/merge_operators/string_append/stringappend2.cc",
+        "utilities/merge_operators/uint64add.cc",
+        "utilities/object_registry.cc",
+        "utilities/option_change_migration/option_change_migration.cc",
+        "utilities/options/options_util.cc",
+        "utilities/persistent_cache/block_cache_tier.cc",
+        "utilities/persistent_cache/block_cache_tier_file.cc",
+        "utilities/persistent_cache/block_cache_tier_metadata.cc",
+        "utilities/persistent_cache/persistent_cache_tier.cc",
+        "utilities/persistent_cache/volatile_tier_impl.cc",
+        "utilities/simulator_cache/cache_simulator.cc",
+        "utilities/simulator_cache/sim_cache.cc",
+        "utilities/table_properties_collectors/compact_on_deletion_collector.cc",
+        "utilities/trace/file_trace_reader_writer.cc",
+        "utilities/transactions/lock/lock_tracker.cc",
+        "utilities/transactions/lock/point_lock_tracker.cc",
+        "utilities/transactions/optimistic_transaction.cc",
+        "utilities/transactions/optimistic_transaction_db_impl.cc",
+        "utilities/transactions/pessimistic_transaction.cc",
+        "utilities/transactions/pessimistic_transaction_db.cc",
+        "utilities/transactions/snapshot_checker.cc",
+        "utilities/transactions/transaction_base.cc",
+        "utilities/transactions/transaction_db_mutex_impl.cc",
+        "utilities/transactions/transaction_lock_mgr.cc",
+        "utilities/transactions/transaction_util.cc",
+        "utilities/transactions/write_prepared_txn.cc",
+        "utilities/transactions/write_prepared_txn_db.cc",
+        "utilities/transactions/write_unprepared_txn.cc",
+        "utilities/transactions/write_unprepared_txn_db.cc",
+        "utilities/ttl/db_ttl_impl.cc",
+        "utilities/write_batch_with_index/write_batch_with_index.cc",
+        "utilities/write_batch_with_index/write_batch_with_index_internal.cc",
+    ],
+    auto_headers = AutoHeaders.RECURSIVE_GLOB,
+    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 = [],
+    external_deps = ROCKSDB_EXTERNAL_DEPS,
+    link_whole = True,
 )
 
 cpp_library(
@@ -384,8 +672,7 @@ 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/mock_time_env.cc",
         "test_util/testharness.cc",
         "test_util/testutil.cc",
         "tools/block_cache_analyzer/block_cache_trace_analyzer.cc",
@@ -398,11 +685,11 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-    deps = [
-        ":rocksdb_lib",
-        ":rocksdb_third_party_gtest",
+    deps = [":rocksdb_lib"],
+    external_deps = ROCKSDB_EXTERNAL_DEPS + [
+        ("googletest", None, "gtest"),
     ],
-    external_deps = ROCKSDB_EXTERNAL_DEPS,
+    link_whole = False,
 )
 
 cpp_library(
@@ -421,6 +708,7 @@ cpp_library(
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
     deps = [":rocksdb_lib"],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
+    link_whole = False,
 )
 
 cpp_library(
@@ -449,18 +737,26 @@ cpp_library(
     external_deps = ROCKSDB_EXTERNAL_DEPS,
 )
 
-cpp_library(
-    name = "rocksdb_third_party_gtest",
-    srcs = ["third-party/gtest-1.8.1/fused-src/gtest/gtest-all.cc"],
-    headers = ["third-party/gtest-1.8.1/fused-src/gtest/gtest.h"],
-    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 = [],
-    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",
@@ -473,6 +769,7 @@ cpp_library(
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
     deps = [":rocksdb_test_lib"],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
+    link_whole = False,
 )
 
 # [test_name, test_src, test_type, extra_deps, extra_compiler_flags]
@@ -519,6 +816,13 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "blob_file_builder_test",
+        "db/blob/blob_file_builder_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "blob_file_garbage_test",
         "db/blob/blob_file_garbage_test.cc",
@@ -575,13 +879,6 @@ ROCKS_TESTS = [
         [],
         [],
     ],
-    [
-        "c_test",
-        "db/c_test.c",
-        "serial",
-        [],
-        [],
-    ],
     [
         "cache_simulator_test",
         "utilities/simulator_cache/cache_simulator_test.cc",
@@ -648,7 +945,7 @@ ROCKS_TESTS = [
     [
         "column_family_test",
         "db/column_family_test.cc",
-        "serial",
+        "parallel",
         [],
         [],
     ],
@@ -942,7 +1239,7 @@ ROCKS_TESTS = [
     [
         "db_test2",
         "db/db_test2.cc",
-        "serial",
+        "parallel",
         [],
         [],
     ],
@@ -1170,6 +1467,13 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "io_tracer_test",
+        "trace_replay/io_tracer_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "iostats_context_test",
         "monitoring/iostats_context_test.cc",
@@ -1345,6 +1649,13 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "prefetch_test",
+        "file/prefetch_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "prefix_test",
         "db/prefix_test.cc",
@@ -1457,6 +1768,13 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "stats_dump_scheduler_test",
+        "monitoring/stats_dump_scheduler_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "stats_history_test",
         "monitoring/stats_history_test.cc",
@@ -1485,6 +1803,13 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "testutil_test",
+        "test_util/testutil_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "thread_list_test",
         "util/thread_list_test.cc",
@@ -1638,18 +1963,17 @@ 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 + extra_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

appveyor.yml

@@ -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

buckifier/buckify_rocksdb.py

@@ -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):
@@ -101,22 +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
-
-
-# Get gtest dir from Makefile
-def get_gtest_dir(repo_path):
-    for line in open(repo_path + "/Makefile"):
-        if line.strip().startswith("GTEST_DIR ="):
-            return line.split("=")[1].strip()
-    # if not found
-    exit_with_error("Unable to find GTEST_DIR in Makefile")
-
+    return s
 
 # Parse extra dependencies passed by user from command line
 def get_dependencies():
@@ -149,21 +123,28 @@ 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 gtest dir
-    gtest_dir = get_gtest_dir(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, gtest_dir)
+    TARGETS = TARGETSBuilder("%s/TARGETS" % repo_path)
 
     # rocksdb_lib
     TARGETS.add_library(
         "rocksdb_lib",
         src_mk["LIB_SOURCES"] +
         src_mk["TOOL_LIB_SOURCES"])
+    # rocksdb_whole_archive_lib
+    TARGETS.add_library(
+        "rocksdb_whole_archive_lib",
+        src_mk["LIB_SOURCES"] +
+        src_mk["TOOL_LIB_SOURCES"],
+        deps=None,
+        headers=None,
+        extra_external_deps="",
+        link_whole=True)
     # rocksdb_test_lib
     TARGETS.add_library(
         "rocksdb_test_lib",
@@ -171,7 +152,10 @@ 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_third_party_gtest"])
+        [":rocksdb_lib"],
+        extra_external_deps=""" + [
+        ("googletest", None, "gtest"),
+    ]""")
     # rocksdb_tools_lib
     TARGETS.add_library(
         "rocksdb_tools_lib",
@@ -185,40 +169,45 @@ def generate_targets(repo_path, deps_map):
         src_mk.get("ANALYZER_LIB_SOURCES", [])
         + src_mk.get('STRESS_LIB_SOURCES', [])
         + ["test_util/testutil.cc"])
-    # rocksdb_third_party_gtest
-    TARGETS.add_library(
-        "rocksdb_third_party_gtest",
-        [gtest_dir + "gtest/gtest-all.cc"],
-        [],
-        [gtest_dir + "gtest/gtest.h"])
 
     print("Extra dependencies:\n{0}".format(json.dumps(deps_map)))
-    # test for every test we found in the Makefile
+
+    # 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)
+
     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,
+                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:"))

buckifier/targets_builder.py

@@ -25,11 +25,10 @@ def pretty_list(lst, indent=8):
 
 
 class TARGETSBuilder(object):
-    def __init__(self, path, gtest_dir):
+    def __init__(self, path):
         self.path = path
         self.targets_file = open(path, 'w')
-        header = targets_cfg.rocksdb_target_header_template.format(
-            gtest_dir=gtest_dir)
+        header = targets_cfg.rocksdb_target_header_template
         self.targets_file.write(header)
         self.total_lib = 0
         self.total_bin = 0
@@ -39,7 +38,8 @@ 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="", link_whole=False):
         headers_attr_prefix = ""
         if headers is None:
             headers_attr_prefix = "auto_"
@@ -51,7 +51,9 @@ class TARGETSBuilder(object):
             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,
+            link_whole=link_whole))
         self.total_lib = self.total_lib + 1
 
     def add_rocksdb_library(self, name, srcs, headers=None):
@@ -75,6 +77,30 @@ 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,

buckifier/targets_cfg.py

@@ -85,14 +85,13 @@ ROCKSDB_PREPROCESSOR_FLAGS = [
     # Directories with files for #include
     "-I" + REPO_PATH + "include/",
     "-I" + REPO_PATH,
-    "-I" + REPO_PATH + "{gtest_dir}",
 ]
 
-ROCKSDB_ARCH_PREPROCESSOR_FLAGS = {{
+ROCKSDB_ARCH_PREPROCESSOR_FLAGS = {
     "x86_64": [
         "-DHAVE_PCLMUL",
     ],
-}}
+}
 
 build_mode = read_config("fbcode", "build_mode")
 
@@ -134,7 +133,8 @@ cpp_library(
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
     deps = [{deps}],
-    external_deps = ROCKSDB_EXTERNAL_DEPS,
+    external_deps = ROCKSDB_EXTERNAL_DEPS{extra_external_deps},
+    link_whole = {link_whole},
 )
 """
 
@@ -183,18 +183,17 @@ 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 + extra_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

build_tools/build_detect_platform

@@ -60,8 +60,16 @@ if [ -z "$ROCKSDB_NO_FBCODE" -a -d /mnt/gvfs/third-party ]; then
       source "$PWD/build_tools/fbcode_config4.8.1.sh"
     elif [ -n "$ROCKSDB_FBCODE_BUILD_WITH_5xx" ]; then
       source "$PWD/build_tools/fbcode_config.sh"
-    else
+    elif [ -n "$ROCKSDB_FBCODE_BUILD_WITH_PLATFORM007" ]; then
       source "$PWD/build_tools/fbcode_config_platform007.sh"
+    elif [ -n "$ROCKSDB_FBCODE_BUILD_WITH_PLATFORM009" ]; then
+      source "$PWD/build_tools/fbcode_config_platform009.sh"
+    elif [ -z "$USE_CLANG" ]; then
+        # Still use platform007 for gcc by default for build break on
+        # some hosts.
+      source "$PWD/build_tools/fbcode_config_platform007.sh"
+    else
+      source "$PWD/build_tools/fbcode_config_platform009.sh"
     fi
 fi
 
@@ -89,6 +97,16 @@ 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 +169,7 @@ 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 $CFLAGS -x c++ - -luring -o /dev/null 2>/dev/null  <<EOF
@@ -192,6 +210,17 @@ EOF
         PLATFORM_LDFLAGS="$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 +314,32 @@ 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 +467,7 @@ 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 $COMMON_FLAGS -lmemkind -x c++ - -o /dev/null 2>/dev/null  <<EOF
@@ -446,7 +483,7 @@ EOF
             JAVA_LDFLAGS="$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 $CFLAGS -x c++ - -o /dev/null 2>/dev/null  <<EOF
@@ -612,6 +649,13 @@ else
   if test "$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 "$TRY_SSE_ETC"; then
@@ -621,11 +665,18 @@ if test "$TRY_SSE_ETC"; then
   # It doesn't even really check that your current CPU is compatible.
   #
   # SSE4.2 available since nehalem, ca. 2008-2010
+  # Includes POPCNT for BitsSetToOne, BitParity
   TRY_SSE42="-msse4.2"
   # PCLMUL available since westmere, ca. 2010-2011
   TRY_PCLMUL="-mpclmul"
   # AVX2 available since haswell, ca. 2013-2015
   TRY_AVX2="-mavx2"
+  # BMI available since haswell, ca. 2013-2015
+  # Primarily for TZCNT for CountTrailingZeroBits
+  TRY_BMI="-mbmi"
+  # LZCNT available since haswell, ca. 2013-2015
+  # For FloorLog2
+  TRY_LZCNT="-mlzcnt"
 fi
 
 $CXX $PLATFORM_CXXFLAGS $COMMON_FLAGS $TRY_SSE42 -x c++ - -o /dev/null 2>/dev/null <<EOF
@@ -674,6 +725,34 @@ elif test "$USE_SSE"; then
   echo "warning: USE_SSE specified but compiler could not use AVX2 intrinsics, disabling" >&2
 fi
 
+$CXX $PLATFORM_CXXFLAGS $COMMON_FLAGS $TRY_BMI -x c++ - -o /dev/null 2>/dev/null <<EOF
+  #include <cstdint>
+  #include <immintrin.h>
+  int main(int argc, char *argv[]) {
+    (void)argv;
+    return (int)_tzcnt_u64((uint64_t)argc);
+  }
+EOF
+if [ "$?" = 0 ]; then
+  COMMON_FLAGS="$COMMON_FLAGS $TRY_BMI -DHAVE_BMI"
+elif test "$USE_SSE"; then
+  echo "warning: USE_SSE specified but compiler could not use BMI intrinsics, disabling" >&2
+fi
+
+$CXX $PLATFORM_CXXFLAGS $COMMON_FLAGS $TRY_LZCNT -x c++ - -o /dev/null 2>/dev/null <<EOF
+  #include <cstdint>
+  #include <immintrin.h>
+  int main(int argc, char *argv[]) {
+    (void)argv;
+    return (int)_lzcnt_u64((uint64_t)argc);
+  }
+EOF
+if [ "$?" = 0 ]; then
+  COMMON_FLAGS="$COMMON_FLAGS $TRY_LZCNT -DHAVE_LZCNT"
+elif test "$USE_SSE"; then
+  echo "warning: USE_SSE specified but compiler could not use LZCNT intrinsics, disabling" >&2
+fi
+
 $CXX $PLATFORM_CXXFLAGS $COMMON_FLAGS -x c++ - -o /dev/null 2>/dev/null <<EOF
   #include <cstdint>
   int main() {
@@ -730,6 +809,7 @@ 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"

build_tools/dependencies_platform009.sh

@@ -0,0 +1,20 @@
+# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+GCC_BASE=/mnt/gvfs/third-party2/gcc/1795efe5f06778c15a92c8f9a2aba5dc496d9d4d/9.x/centos7-native/3bed279
+CLANG_BASE=/mnt/gvfs/third-party2/llvm-fb/7318eaac22659b6ff2fe43918e4b69fd0772a8a7/9.0.0/platform009/651ee30
+LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/4959b39cfbe5965a37c861c4c327fa7c5c759b87/9.x/platform009/9202ce7
+GLIBC_BASE=/mnt/gvfs/third-party2/glibc/45ce3375cdc77ecb2520bbf8f0ecddd3f98efd7a/2.30/platform009/f259413
+SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/be4de3205e029101b18aa8103daa696c2bef3b19/1.1.3/platform009/7f3b187
+ZLIB_BASE=/mnt/gvfs/third-party2/zlib/3c160ac5c67e257501e24c6c1d00ad5e01d73db6/1.2.8/platform009/7f3b187
+BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/73a237ac5bc0a5f5d67b39b8d253cfebaab88684/1.0.6/platform009/7f3b187
+LZ4_BASE=/mnt/gvfs/third-party2/lz4/ec6573523b0ce55ef6373a4801189027cf07bb2c/1.9.1/platform009/7f3b187
+ZSTD_BASE=/mnt/gvfs/third-party2/zstd/64c58a207d2495e83abc57a500a956df09b79a7c/1.4.x/platform009/ba86d1f
+GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/824d0a8a5abb5b121afd1b35fc3896407ea50092/2.2.0/platform009/7f3b187
+JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/d9aef9feb850b168a68736420f217b01cce11a89/master/platform009/c305944
+NUMA_BASE=/mnt/gvfs/third-party2/numa/0af65f71e23a67bf65dc91b11f95caa39325c432/2.0.11/platform009/7f3b187
+LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/02486dac347645d31dce116f44e1de3177315be2/1.4/platform009/5191652
+TBB_BASE=/mnt/gvfs/third-party2/tbb/2e0ec671e550bfca347300bf3f789d9c0fff24ad/2018_U5/platform009/7f3b187
+LIBURING_BASE=/mnt/gvfs/third-party2/liburing/70dbd9cfee63a25611417d09433a86d7711b3990/20200729/platform009/7f3b187
+KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/32b8a2407b634df3f8f948ba373fc4acc6a18296/fb/platform009/da39a3e
+BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/08634589372fa5f237bfd374e8c644a8364e78c1/2.32/platform009/ba86d1f/
+VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/6ae525939ad02e5e676855082fbbc7828dbafeac/3.15.0/platform009/7f3b187
+LUA_BASE=/mnt/gvfs/third-party2/lua/162efd9561a3d21f6869f4814011e9cf1b3ff4dc/5.3.4/platform009/a6271c4

build_tools/fbcode_config.sh

@@ -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"
 

build_tools/fbcode_config4.8.1.sh

@@ -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/"
 

build_tools/fbcode_config_platform007.sh

@@ -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"
 

build_tools/fbcode_config_platform009.sh

@@ -0,0 +1,172 @@
+#!/bin/sh
+# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+#
+# Set environment variables so that we can compile rocksdb using
+# fbcode settings.  It uses the latest g++ and clang compilers and also
+# uses jemalloc
+# Environment variables that change the behavior of this script:
+# PIC_BUILD -- if true, it will only take pic versions of libraries from fbcode. libraries that don't have pic variant will not be included
+
+
+BASEDIR=`dirname $BASH_SOURCE`
+source "$BASEDIR/dependencies_platform009.sh"
+
+CFLAGS=""
+
+# libgcc
+LIBGCC_INCLUDE="$LIBGCC_BASE/include/c++/9.3.0"
+LIBGCC_LIBS=" -L $LIBGCC_BASE/lib"
+
+# glibc
+GLIBC_INCLUDE="$GLIBC_BASE/include"
+GLIBC_LIBS=" -L $GLIBC_BASE/lib"
+
+# snappy
+SNAPPY_INCLUDE=" -I $SNAPPY_BASE/include/"
+if test -z $PIC_BUILD; then
+  SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy.a"
+else
+  SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy_pic.a"
+fi
+CFLAGS+=" -DSNAPPY"
+
+if test -z $PIC_BUILD; then
+  # location of zlib headers and libraries
+  ZLIB_INCLUDE=" -I $ZLIB_BASE/include/"
+  ZLIB_LIBS=" $ZLIB_BASE/lib/libz.a"
+  CFLAGS+=" -DZLIB"
+
+  # location of bzip headers and libraries
+  BZIP_INCLUDE=" -I $BZIP2_BASE/include/"
+  BZIP_LIBS=" $BZIP2_BASE/lib/libbz2.a"
+  CFLAGS+=" -DBZIP2"
+
+  LZ4_INCLUDE=" -I $LZ4_BASE/include/"
+  LZ4_LIBS=" $LZ4_BASE/lib/liblz4.a"
+  CFLAGS+=" -DLZ4"
+fi
+
+ZSTD_INCLUDE=" -I $ZSTD_BASE/include/"
+if test -z $PIC_BUILD; then
+  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd.a"
+else
+  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd_pic.a"
+fi
+CFLAGS+=" -DZSTD"
+
+# location of gflags headers and libraries
+GFLAGS_INCLUDE=" -I $GFLAGS_BASE/include/"
+if test -z $PIC_BUILD; then
+  GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags.a"
+else
+  GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags_pic.a"
+fi
+CFLAGS+=" -DGFLAGS=gflags"
+
+# location of jemalloc
+JEMALLOC_INCLUDE=" -I $JEMALLOC_BASE/include/"
+JEMALLOC_LIB=" $JEMALLOC_BASE/lib/libjemalloc.a"
+
+if test -z $PIC_BUILD; then
+  # location of numa
+  NUMA_INCLUDE=" -I $NUMA_BASE/include/"
+  NUMA_LIB=" $NUMA_BASE/lib/libnuma.a"
+  CFLAGS+=" -DNUMA"
+
+  # location of libunwind
+  LIBUNWIND="$LIBUNWIND_BASE/lib/libunwind.a"
+fi
+
+# location of TBB
+TBB_INCLUDE=" -isystem $TBB_BASE/include/"
+if test -z $PIC_BUILD; then
+  TBB_LIBS="$TBB_BASE/lib/libtbb.a"
+else
+  TBB_LIBS="$TBB_BASE/lib/libtbb_pic.a"
+fi
+CFLAGS+=" -DTBB"
+
+# location of LIBURING
+LIBURING_INCLUDE=" -isystem $LIBURING_BASE/include/"
+if test -z $PIC_BUILD; then
+  LIBURING_LIBS="$LIBURING_BASE/lib/liburing.a"
+else
+  LIBURING_LIBS="$LIBURING_BASE/lib/liburing_pic.a"
+fi
+CFLAGS+=" -DLIBURING"
+
+test "$USE_SSE" || USE_SSE=1
+export USE_SSE
+test "$PORTABLE" || PORTABLE=1
+export PORTABLE
+
+BINUTILS="$BINUTILS_BASE/bin"
+AR="$BINUTILS/ar"
+
+DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE $TBB_INCLUDE $LIBURING_INCLUDE"
+
+STDLIBS="-L $GCC_BASE/lib64"
+
+CLANG_BIN="$CLANG_BASE/bin"
+CLANG_LIB="$CLANG_BASE/lib"
+CLANG_SRC="$CLANG_BASE/../../src"
+
+CLANG_ANALYZER="$CLANG_BIN/clang++"
+CLANG_SCAN_BUILD="$CLANG_SRC/llvm/clang/tools/scan-build/bin/scan-build"
+
+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"
+  CFLAGS+=" -isystem $GLIBC_INCLUDE"
+  JEMALLOC=1
+else
+  # clang
+  CLANG_INCLUDE="$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"
+
+  CFLAGS+=" -B$BINUTILS/gold -nostdinc -nostdlib"
+  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/9.x "
+  CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/9.x/x86_64-facebook-linux "
+  CFLAGS+=" -isystem $GLIBC_INCLUDE"
+  CFLAGS+=" -isystem $LIBGCC_INCLUDE"
+  CFLAGS+=" -isystem $CLANG_INCLUDE"
+  CFLAGS+=" -isystem $KERNEL_HEADERS_INCLUDE/linux "
+  CFLAGS+=" -isystem $KERNEL_HEADERS_INCLUDE "
+  CFLAGS+=" -Wno-expansion-to-defined "
+  CXXFLAGS="-nostdinc++"
+fi
+
+CFLAGS+=" $DEPS_INCLUDE"
+CFLAGS+=" -DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX -DROCKSDB_FALLOCATE_PRESENT -DROCKSDB_MALLOC_USABLE_SIZE -DROCKSDB_RANGESYNC_PRESENT -DROCKSDB_SCHED_GETCPU_PRESENT -DROCKSDB_SUPPORT_THREAD_LOCAL -DHAVE_SSE42 -DROCKSDB_IOURING_PRESENT"
+CXXFLAGS+=" $CFLAGS"
+
+EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB $TBB_LIBS $LIBURING_LIBS"
+EXEC_LDFLAGS+=" -B$BINUTILS/gold"
+EXEC_LDFLAGS+=" -Wl,--dynamic-linker,/usr/local/fbcode/platform009/lib/ld.so"
+EXEC_LDFLAGS+=" $LIBUNWIND"
+EXEC_LDFLAGS+=" -Wl,-rpath=/usr/local/fbcode/platform009/lib"
+# required by libtbb
+EXEC_LDFLAGS+=" -ldl"
+
+PLATFORM_LDFLAGS="$LIBGCC_LIBS $GLIBC_LIBS $STDLIBS -lgcc -lstdc++"
+
+EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $TBB_LIBS $LIBURING_LIBS"
+
+VALGRIND_VER="$VALGRIND_BASE/bin/"
+
+# lua not supported because it's on track for deprecation, I think
+LUA_PATH=
+LUA_LIB=
+
+USE_FOLLY_DISTRIBUTED_MUTEX=1
+
+export CC CXX AR CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE CLANG_ANALYZER CLANG_SCAN_BUILD LUA_PATH LUA_LIB

build_tools/make_package.sh

@@ -103,31 +103,26 @@ function main() {
   gem_install fpm
 
   make static_lib
-  make install INSTALL_PATH=package
-
-  cd package
-
-  LIB_DIR=lib
-  if [[ -z "$ARCH" ]]; then
-      ARCH=$(getconf LONG_BIT)
-  fi
-  if [[ ("$FPM_OUTPUT" = "rpm") && ($ARCH -eq 64) ]]; then
-      mv lib lib64
-      LIB_DIR=lib64
+  LIBDIR=/usr/lib
+  if [[ $FPM_OUTPUT = "rpm" ]]; then
+      LIBDIR=$(rpm --eval '%_libdir')
   fi
 
+  rm -rf package
+  make install DESTDIR=package PREFIX=/usr LIBDIR=$LIBDIR
+
   fpm \
     -s dir \
     -t $FPM_OUTPUT \
+    -C package \
     -n rocksdb \
     -v $1 \
-    --prefix /usr \
     --url http://rocksdb.org/ \
     -m rocksdb@fb.com \
     --license BSD \
     --vendor Facebook \
     --description "RocksDB is an embeddable persistent key-value store for fast storage." \
-    include $LIB_DIR
+    usr
 }
 
 # shellcheck disable=SC2068

build_tools/rocksdb-lego-determinator

@@ -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
 #
@@ -897,6 +1101,7 @@ run_regression()
   strip librocksdb.a
   send_size_to_ods static_lib_stripped $(stat --printf="%s" librocksdb.a)
 
+  make clean
   make -j$(nproc) shared_lib
   send_size_to_ods shared_lib $(stat --printf="%s" `readlink -f librocksdb.so`)
   strip `readlink -f librocksdb.so`
@@ -909,6 +1114,7 @@ run_regression()
   strip librocksdb.a
   send_size_to_ods static_lib_lite_stripped $(stat --printf="%s" librocksdb.a)
 
+  make clean
   make LITE=1 -j$(nproc) shared_lib
   send_size_to_ods shared_lib_lite $(stat --printf="%s" `readlink -f librocksdb.so`)
   strip `readlink -f librocksdb.so`
@@ -994,6 +1200,12 @@ 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 +1221,12 @@ 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 +1239,12 @@ 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 +1260,12 @@ 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
     ;;

cache/cache_test.cc

@@ -117,8 +117,8 @@ class CacheTest : public testing::TestWithParam<std::string> {
 
   void Insert(std::shared_ptr<Cache> cache, int key, int value,
               int charge = 1) {
-    cache->Insert(EncodeKey(key), EncodeValue(value), charge,
-                  &CacheTest::Deleter);
+    EXPECT_OK(cache->Insert(EncodeKey(key), EncodeValue(value), charge,
+                            &CacheTest::Deleter));
   }
 
   void Erase(std::shared_ptr<Cache> cache, int key) {
@@ -167,9 +167,10 @@ TEST_P(CacheTest, UsageTest) {
   for (int i = 1; i < 100; ++i) {
     std::string key(i, 'a');
     auto kv_size = key.size() + 5;
-    cache->Insert(key, reinterpret_cast<void*>(value), kv_size, dumbDeleter);
-    precise_cache->Insert(key, reinterpret_cast<void*>(value), kv_size,
-                          dumbDeleter);
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), kv_size,
+                            dumbDeleter));
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    kv_size, dumbDeleter));
     usage += kv_size;
     ASSERT_EQ(usage, cache->GetUsage());
     ASSERT_LT(usage, precise_cache->GetUsage());
@@ -183,10 +184,10 @@ TEST_P(CacheTest, UsageTest) {
   // make sure the cache will be overloaded
   for (uint64_t i = 1; i < kCapacity; ++i) {
     auto key = ToString(i);
-    cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
-                  dumbDeleter);
-    precise_cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
-                          dumbDeleter);
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
+                            dumbDeleter));
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    key.size() + 5, dumbDeleter));
   }
 
   // the usage should be close to the capacity
@@ -215,11 +216,12 @@ TEST_P(CacheTest, PinnedUsageTest) {
     auto kv_size = key.size() + 5;
     Cache::Handle* handle;
     Cache::Handle* handle_in_precise_cache;
-    cache->Insert(key, reinterpret_cast<void*>(value), kv_size, dumbDeleter,
-                  &handle);
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), kv_size,
+                            dumbDeleter, &handle));
     assert(handle);
-    precise_cache->Insert(key, reinterpret_cast<void*>(value), kv_size,
-                          dumbDeleter, &handle_in_precise_cache);
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    kv_size, dumbDeleter,
+                                    &handle_in_precise_cache));
     assert(handle_in_precise_cache);
     pinned_usage += kv_size;
     ASSERT_EQ(pinned_usage, cache->GetPinnedUsage());
@@ -254,10 +256,10 @@ TEST_P(CacheTest, PinnedUsageTest) {
   // check that overloading the cache does not change the pinned usage
   for (uint64_t i = 1; i < 2 * kCapacity; ++i) {
     auto key = ToString(i);
-    cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
-                  dumbDeleter);
-    precise_cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
-                          dumbDeleter);
+    ASSERT_OK(cache->Insert(key, reinterpret_cast<void*>(value), key.size() + 5,
+                            dumbDeleter));
+    ASSERT_OK(precise_cache->Insert(key, reinterpret_cast<void*>(value),
+                                    key.size() + 5, dumbDeleter));
   }
   ASSERT_EQ(pinned_usage, cache->GetPinnedUsage());
   ASSERT_EQ(precise_cache_pinned_usage, precise_cache->GetPinnedUsage());

cache/lru_cache_test.cc

@@ -37,8 +37,9 @@ class LRUCacheTest : public testing::Test {
 
   void Insert(const std::string& key,
               Cache::Priority priority = Cache::Priority::LOW) {
-    cache_->Insert(key, 0 /*hash*/, nullptr /*value*/, 1 /*charge*/,
-                   nullptr /*deleter*/, nullptr /*handle*/, priority);
+    EXPECT_OK(cache_->Insert(key, 0 /*hash*/, nullptr /*value*/, 1 /*charge*/,
+                             nullptr /*deleter*/, nullptr /*handle*/,
+                             priority));
   }
 
   void Insert(char key, Cache::Priority priority = Cache::Priority::LOW) {

db/arena_wrapped_db_iter.cc

@@ -45,6 +45,7 @@ void ArenaWrappedDBIter::Init(Env* env, const ReadOptions& read_options,
                               true, max_sequential_skip_in_iteration,
                               read_callback, db_impl, cfd, allow_blob);
   sv_number_ = version_number;
+  read_options_ = read_options;
   allow_refresh_ = allow_refresh;
 }
 
@@ -56,8 +57,9 @@ 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 +67,7 @@ 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 +81,7 @@ 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();
@@ -96,8 +99,7 @@ ArenaWrappedDBIter* NewArenaWrappedDbIterator(
              max_sequential_skip_in_iterations, version_number, read_callback,
              db_impl, cfd, allow_blob, allow_refresh);
   if (db_impl != nullptr && cfd != nullptr && allow_refresh) {
-    iter->StoreRefreshInfo(read_options, db_impl, cfd, read_callback,
-                           allow_blob);
+    iter->StoreRefreshInfo(db_impl, cfd, read_callback, allow_blob);
   }
 
   return iter;

db/arena_wrapped_db_iter.h

@@ -41,6 +41,7 @@ class ArenaWrappedDBIter : public Iterator {
   virtual ReadRangeDelAggregator* GetRangeDelAggregator() {
     return db_iter_->GetRangeDelAggregator();
   }
+  const ReadOptions& GetReadOptions() { return read_options_; }
 
   // Set the internal iterator wrapped inside the DB Iterator. Usually it is
   // a merging iterator.
@@ -79,10 +80,8 @@ class ArenaWrappedDBIter : public Iterator {
 
   // Store some parameters so we can refresh the iterator at a later point
   // with these same params
-  void StoreRefreshInfo(const ReadOptions& read_options, DBImpl* db_impl,
-                        ColumnFamilyData* cfd, ReadCallback* read_callback,
-                        bool allow_blob) {
-    read_options_ = read_options;
+  void StoreRefreshInfo(DBImpl* db_impl, ColumnFamilyData* cfd,
+                        ReadCallback* read_callback, bool allow_blob) {
     db_impl_ = db_impl;
     cfd_ = cfd;
     read_callback_ = read_callback;

db/blob/blob_file_builder.cc

@@ -0,0 +1,304 @@
+//  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 "db/blob/blob_file_builder.h"
+
+#include <cassert>
+
+#include "db/blob/blob_file_addition.h"
+#include "db/blob/blob_index.h"
+#include "db/blob/blob_log_format.h"
+#include "db/blob/blob_log_writer.h"
+#include "db/version_set.h"
+#include "file/filename.h"
+#include "file/read_write_util.h"
+#include "file/writable_file_writer.h"
+#include "logging/logging.h"
+#include "options/cf_options.h"
+#include "rocksdb/slice.h"
+#include "rocksdb/status.h"
+#include "test_util/sync_point.h"
+#include "util/compression.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+BlobFileBuilder::BlobFileBuilder(
+    VersionSet* versions, Env* env, FileSystem* fs,
+    const ImmutableCFOptions* immutable_cf_options,
+    const MutableCFOptions* mutable_cf_options, const FileOptions* file_options,
+    int job_id, uint32_t column_family_id,
+    const std::string& column_family_name, Env::IOPriority io_priority,
+    Env::WriteLifeTimeHint write_hint,
+    std::vector<BlobFileAddition>* blob_file_additions)
+    : BlobFileBuilder([versions]() { return versions->NewFileNumber(); }, env,
+                      fs, immutable_cf_options, mutable_cf_options,
+                      file_options, job_id, column_family_id,
+                      column_family_name, io_priority, write_hint,
+                      blob_file_additions) {}
+
+BlobFileBuilder::BlobFileBuilder(
+    std::function<uint64_t()> file_number_generator, Env* env, FileSystem* fs,
+    const ImmutableCFOptions* immutable_cf_options,
+    const MutableCFOptions* mutable_cf_options, const FileOptions* file_options,
+    int job_id, uint32_t column_family_id,
+    const std::string& column_family_name, Env::IOPriority io_priority,
+    Env::WriteLifeTimeHint write_hint,
+    std::vector<BlobFileAddition>* blob_file_additions)
+    : file_number_generator_(std::move(file_number_generator)),
+      env_(env),
+      fs_(fs),
+      immutable_cf_options_(immutable_cf_options),
+      min_blob_size_(mutable_cf_options->min_blob_size),
+      blob_file_size_(mutable_cf_options->blob_file_size),
+      blob_compression_type_(mutable_cf_options->blob_compression_type),
+      file_options_(file_options),
+      job_id_(job_id),
+      column_family_id_(column_family_id),
+      column_family_name_(column_family_name),
+      io_priority_(io_priority),
+      write_hint_(write_hint),
+      blob_file_additions_(blob_file_additions),
+      blob_count_(0),
+      blob_bytes_(0) {
+  assert(file_number_generator_);
+  assert(env_);
+  assert(fs_);
+  assert(immutable_cf_options_);
+  assert(file_options_);
+  assert(blob_file_additions_);
+}
+
+BlobFileBuilder::~BlobFileBuilder() = default;
+
+Status BlobFileBuilder::Add(const Slice& key, const Slice& value,
+                            std::string* blob_index) {
+  assert(blob_index);
+  assert(blob_index->empty());
+
+  if (value.size() < min_blob_size_) {
+    return Status::OK();
+  }
+
+  {
+    const Status s = OpenBlobFileIfNeeded();
+    if (!s.ok()) {
+      return s;
+    }
+  }
+
+  Slice blob = value;
+  std::string compressed_blob;
+
+  {
+    const Status s = CompressBlobIfNeeded(&blob, &compressed_blob);
+    if (!s.ok()) {
+      return s;
+    }
+  }
+
+  uint64_t blob_file_number = 0;
+  uint64_t blob_offset = 0;
+
+  {
+    const Status s =
+        WriteBlobToFile(key, blob, &blob_file_number, &blob_offset);
+    if (!s.ok()) {
+      return s;
+    }
+  }
+
+  {
+    const Status s = CloseBlobFileIfNeeded();
+    if (!s.ok()) {
+      return s;
+    }
+  }
+
+  BlobIndex::EncodeBlob(blob_index, blob_file_number, blob_offset, blob.size(),
+                        blob_compression_type_);
+
+  return Status::OK();
+}
+
+Status BlobFileBuilder::Finish() {
+  if (!IsBlobFileOpen()) {
+    return Status::OK();
+  }
+
+  return CloseBlobFile();
+}
+
+bool BlobFileBuilder::IsBlobFileOpen() const { return !!writer_; }
+
+Status BlobFileBuilder::OpenBlobFileIfNeeded() {
+  if (IsBlobFileOpen()) {
+    return Status::OK();
+  }
+
+  assert(!blob_count_);
+  assert(!blob_bytes_);
+
+  assert(file_number_generator_);
+  const uint64_t blob_file_number = file_number_generator_();
+
+  assert(immutable_cf_options_);
+  assert(!immutable_cf_options_->cf_paths.empty());
+  const std::string blob_file_path = BlobFileName(
+      immutable_cf_options_->cf_paths.front().path, blob_file_number);
+
+  std::unique_ptr<FSWritableFile> file;
+
+  {
+    TEST_SYNC_POINT("BlobFileBuilder::OpenBlobFileIfNeeded:NewWritableFile");
+
+    assert(file_options_);
+    const Status s =
+        NewWritableFile(fs_, blob_file_path, &file, *file_options_);
+    if (!s.ok()) {
+      return s;
+    }
+  }
+
+  assert(file);
+  file->SetIOPriority(io_priority_);
+  file->SetWriteLifeTimeHint(write_hint_);
+
+  Statistics* const statistics = immutable_cf_options_->statistics;
+
+  std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
+      std::move(file), blob_file_path, *file_options_, env_,
+      nullptr /*IOTracer*/, statistics, immutable_cf_options_->listeners,
+      immutable_cf_options_->file_checksum_gen_factory));
+
+  std::unique_ptr<BlobLogWriter> blob_log_writer(
+      new BlobLogWriter(std::move(file_writer), env_, statistics,
+                        blob_file_number, immutable_cf_options_->use_fsync));
+
+  constexpr bool has_ttl = false;
+  constexpr ExpirationRange expiration_range;
+
+  BlobLogHeader header(column_family_id_, blob_compression_type_, has_ttl,
+                       expiration_range);
+
+  {
+    TEST_SYNC_POINT("BlobFileBuilder::OpenBlobFileIfNeeded:WriteHeader");
+
+    const Status s = blob_log_writer->WriteHeader(header);
+    if (!s.ok()) {
+      return s;
+    }
+  }
+
+  writer_ = std::move(blob_log_writer);
+
+  assert(IsBlobFileOpen());
+
+  return Status::OK();
+}
+
+Status BlobFileBuilder::CompressBlobIfNeeded(
+    Slice* blob, std::string* compressed_blob) const {
+  assert(blob);
+  assert(compressed_blob);
+  assert(compressed_blob->empty());
+
+  if (blob_compression_type_ == kNoCompression) {
+    return Status::OK();
+  }
+
+  CompressionOptions opts;
+  CompressionContext context(blob_compression_type_);
+  constexpr uint64_t sample_for_compression = 0;
+
+  CompressionInfo info(opts, context, CompressionDict::GetEmptyDict(),
+                       blob_compression_type_, sample_for_compression);
+
+  constexpr uint32_t compression_format_version = 2;
+
+  if (!CompressData(*blob, info, compression_format_version, compressed_blob)) {
+    return Status::Corruption("Error compressing blob");
+  }
+
+  *blob = Slice(*compressed_blob);
+
+  return Status::OK();
+}
+
+Status BlobFileBuilder::WriteBlobToFile(const Slice& key, const Slice& blob,
+                                        uint64_t* blob_file_number,
+                                        uint64_t* blob_offset) {
+  assert(IsBlobFileOpen());
+  assert(blob_file_number);
+  assert(blob_offset);
+
+  uint64_t key_offset = 0;
+
+  TEST_SYNC_POINT("BlobFileBuilder::WriteBlobToFile:AddRecord");
+
+  const Status s = writer_->AddRecord(key, blob, &key_offset, blob_offset);
+  if (!s.ok()) {
+    return s;
+  }
+
+  *blob_file_number = writer_->get_log_number();
+
+  ++blob_count_;
+  blob_bytes_ += BlobLogRecord::kHeaderSize + key.size() + blob.size();
+
+  return Status::OK();
+}
+
+Status BlobFileBuilder::CloseBlobFile() {
+  assert(IsBlobFileOpen());
+
+  BlobLogFooter footer;
+  footer.blob_count = blob_count_;
+
+  std::string checksum_method;
+  std::string checksum_value;
+
+  TEST_SYNC_POINT("BlobFileBuilder::WriteBlobToFile:AppendFooter");
+
+  const Status s =
+      writer_->AppendFooter(footer, &checksum_method, &checksum_value);
+  if (!s.ok()) {
+    return s;
+  }
+
+  const uint64_t blob_file_number = writer_->get_log_number();
+
+  assert(blob_file_additions_);
+  blob_file_additions_->emplace_back(blob_file_number, blob_count_, blob_bytes_,
+                                     std::move(checksum_method),
+                                     std::move(checksum_value));
+
+  assert(immutable_cf_options_);
+  ROCKS_LOG_INFO(immutable_cf_options_->info_log,
+                 "[%s] [JOB %d] Generated blob file #%" PRIu64 ": %" PRIu64
+                 " total blobs, %" PRIu64 " total bytes",
+                 column_family_name_.c_str(), job_id_, blob_file_number,
+                 blob_count_, blob_bytes_);
+
+  writer_.reset();
+  blob_count_ = 0;
+  blob_bytes_ = 0;
+
+  return Status::OK();
+}
+
+Status BlobFileBuilder::CloseBlobFileIfNeeded() {
+  assert(IsBlobFileOpen());
+
+  const WritableFileWriter* const file_writer = writer_->file();
+  assert(file_writer);
+
+  if (file_writer->GetFileSize() < blob_file_size_) {
+    return Status::OK();
+  }
+
+  return CloseBlobFile();
+}
+
+}  // namespace ROCKSDB_NAMESPACE

db/blob/blob_file_builder.h

@@ -0,0 +1,88 @@
+//  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).
+#pragma once
+
+#include <cinttypes>
+#include <functional>
+#include <memory>
+#include <string>
+#include <vector>
+
+#include "rocksdb/compression_type.h"
+#include "rocksdb/env.h"
+#include "rocksdb/rocksdb_namespace.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class VersionSet;
+class FileSystem;
+struct ImmutableCFOptions;
+struct MutableCFOptions;
+struct FileOptions;
+class BlobFileAddition;
+class Status;
+class Slice;
+class BlobLogWriter;
+
+class BlobFileBuilder {
+ public:
+  BlobFileBuilder(VersionSet* versions, Env* env, FileSystem* fs,
+                  const ImmutableCFOptions* immutable_cf_options,
+                  const MutableCFOptions* mutable_cf_options,
+                  const FileOptions* file_options, int job_id,
+                  uint32_t column_family_id,
+                  const std::string& column_family_name,
+                  Env::IOPriority io_priority,
+                  Env::WriteLifeTimeHint write_hint,
+                  std::vector<BlobFileAddition>* blob_file_additions);
+
+  BlobFileBuilder(std::function<uint64_t()> file_number_generator, Env* env,
+                  FileSystem* fs,
+                  const ImmutableCFOptions* immutable_cf_options,
+                  const MutableCFOptions* mutable_cf_options,
+                  const FileOptions* file_options, int job_id,
+                  uint32_t column_family_id,
+                  const std::string& column_family_name,
+                  Env::IOPriority io_priority,
+                  Env::WriteLifeTimeHint write_hint,
+                  std::vector<BlobFileAddition>* blob_file_additions);
+
+  BlobFileBuilder(const BlobFileBuilder&) = delete;
+  BlobFileBuilder& operator=(const BlobFileBuilder&) = delete;
+
+  ~BlobFileBuilder();
+
+  Status Add(const Slice& key, const Slice& value, std::string* blob_index);
+  Status Finish();
+
+ private:
+  bool IsBlobFileOpen() const;
+  Status OpenBlobFileIfNeeded();
+  Status CompressBlobIfNeeded(Slice* blob, std::string* compressed_blob) const;
+  Status WriteBlobToFile(const Slice& key, const Slice& blob,
+                         uint64_t* blob_file_number, uint64_t* blob_offset);
+  Status CloseBlobFile();
+  Status CloseBlobFileIfNeeded();
+
+  std::function<uint64_t()> file_number_generator_;
+  Env* env_;
+  FileSystem* fs_;
+  const ImmutableCFOptions* immutable_cf_options_;
+  uint64_t min_blob_size_;
+  uint64_t blob_file_size_;
+  CompressionType blob_compression_type_;
+  const FileOptions* file_options_;
+  int job_id_;
+  uint32_t column_family_id_;
+  std::string column_family_name_;
+  Env::IOPriority io_priority_;
+  Env::WriteLifeTimeHint write_hint_;
+  std::vector<BlobFileAddition>* blob_file_additions_;
+  std::unique_ptr<BlobLogWriter> writer_;
+  uint64_t blob_count_;
+  uint64_t blob_bytes_;
+};
+
+}  // namespace ROCKSDB_NAMESPACE

db/blob/blob_file_builder_test.cc

@@ -0,0 +1,594 @@
+//  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 "db/blob/blob_file_builder.h"
+
+#include <cassert>
+#include <cinttypes>
+#include <string>
+#include <utility>
+#include <vector>
+
+#include "db/blob/blob_file_addition.h"
+#include "db/blob/blob_index.h"
+#include "db/blob/blob_log_format.h"
+#include "db/blob/blob_log_reader.h"
+#include "env/composite_env_wrapper.h"
+#include "env/mock_env.h"
+#include "file/filename.h"
+#include "file/random_access_file_reader.h"
+#include "options/cf_options.h"
+#include "rocksdb/env.h"
+#include "rocksdb/file_checksum.h"
+#include "rocksdb/options.h"
+#include "test_util/sync_point.h"
+#include "test_util/testharness.h"
+#include "util/compression.h"
+#include "utilities/fault_injection_env.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class TestFileNumberGenerator {
+ public:
+  uint64_t operator()() { return ++next_file_number_; }
+
+ private:
+  uint64_t next_file_number_ = 1;
+};
+
+class BlobFileBuilderTest : public testing::Test {
+ protected:
+  BlobFileBuilderTest() : mock_env_(Env::Default()), fs_(&mock_env_) {}
+
+  void VerifyBlobFile(const ImmutableCFOptions& immutable_cf_options,
+                      uint64_t blob_file_number, uint32_t column_family_id,
+                      CompressionType blob_compression_type,
+                      const std::vector<std::pair<std::string, std::string>>&
+                          expected_key_value_pairs,
+                      const std::vector<std::string>& blob_indexes) {
+    assert(expected_key_value_pairs.size() == blob_indexes.size());
+
+    const std::string blob_file_path = BlobFileName(
+        immutable_cf_options.cf_paths.front().path, blob_file_number);
+
+    std::unique_ptr<FSRandomAccessFile> file;
+    constexpr IODebugContext* dbg = nullptr;
+    ASSERT_OK(
+        fs_.NewRandomAccessFile(blob_file_path, file_options_, &file, dbg));
+
+    std::unique_ptr<RandomAccessFileReader> file_reader(
+        new RandomAccessFileReader(std::move(file), blob_file_path,
+                                   &mock_env_));
+
+    constexpr Statistics* statistics = nullptr;
+    BlobLogReader blob_log_reader(std::move(file_reader), &mock_env_,
+                                  statistics);
+
+    BlobLogHeader header;
+    ASSERT_OK(blob_log_reader.ReadHeader(&header));
+    ASSERT_EQ(header.version, kVersion1);
+    ASSERT_EQ(header.column_family_id, column_family_id);
+    ASSERT_EQ(header.compression, blob_compression_type);
+    ASSERT_FALSE(header.has_ttl);
+    ASSERT_EQ(header.expiration_range, ExpirationRange());
+
+    for (size_t i = 0; i < expected_key_value_pairs.size(); ++i) {
+      BlobLogRecord record;
+      uint64_t blob_offset = 0;
+
+      ASSERT_OK(blob_log_reader.ReadRecord(
+          &record, BlobLogReader::kReadHeaderKeyBlob, &blob_offset));
+
+      // Check the contents of the blob file
+      const auto& expected_key_value = expected_key_value_pairs[i];
+      const auto& key = expected_key_value.first;
+      const auto& value = expected_key_value.second;
+
+      ASSERT_EQ(record.key_size, key.size());
+      ASSERT_EQ(record.value_size, value.size());
+      ASSERT_EQ(record.expiration, 0);
+      ASSERT_EQ(record.key, key);
+      ASSERT_EQ(record.value, value);
+
+      // Make sure the blob reference returned by the builder points to the
+      // right place
+      BlobIndex blob_index;
+      ASSERT_OK(blob_index.DecodeFrom(blob_indexes[i]));
+      ASSERT_FALSE(blob_index.IsInlined());
+      ASSERT_FALSE(blob_index.HasTTL());
+      ASSERT_EQ(blob_index.file_number(), blob_file_number);
+      ASSERT_EQ(blob_index.offset(), blob_offset);
+      ASSERT_EQ(blob_index.size(), value.size());
+    }
+
+    BlobLogFooter footer;
+    ASSERT_OK(blob_log_reader.ReadFooter(&footer));
+    ASSERT_EQ(footer.blob_count, expected_key_value_pairs.size());
+    ASSERT_EQ(footer.expiration_range, ExpirationRange());
+  }
+
+  MockEnv mock_env_;
+  LegacyFileSystemWrapper fs_;
+  FileOptions file_options_;
+};
+
+TEST_F(BlobFileBuilderTest, BuildAndCheckOneFile) {
+  // Build a single blob file
+  constexpr size_t number_of_blobs = 10;
+  constexpr size_t key_size = 1;
+  constexpr size_t value_size = 4;
+  constexpr size_t value_offset = 1234;
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&mock_env_,
+                            "BlobFileBuilderTest_BuildAndCheckOneFile"),
+      0);
+  options.enable_blob_files = true;
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(
+      TestFileNumberGenerator(), &mock_env_, &fs_, &immutable_cf_options,
+      &mutable_cf_options, &file_options_, job_id, column_family_id,
+      column_family_name, io_priority, write_hint, &blob_file_additions);
+
+  std::vector<std::pair<std::string, std::string>> expected_key_value_pairs(
+      number_of_blobs);
+  std::vector<std::string> blob_indexes(number_of_blobs);
+
+  for (size_t i = 0; i < number_of_blobs; ++i) {
+    auto& expected_key_value = expected_key_value_pairs[i];
+
+    auto& key = expected_key_value.first;
+    key = std::to_string(i);
+    assert(key.size() == key_size);
+
+    auto& value = expected_key_value.second;
+    value = std::to_string(i + value_offset);
+    assert(value.size() == value_size);
+
+    auto& blob_index = blob_indexes[i];
+
+    ASSERT_OK(builder.Add(key, value, &blob_index));
+    ASSERT_FALSE(blob_index.empty());
+  }
+
+  ASSERT_OK(builder.Finish());
+
+  // Check the metadata generated
+  ASSERT_EQ(blob_file_additions.size(), 1);
+
+  const auto& blob_file_addition = blob_file_additions[0];
+
+  constexpr uint64_t blob_file_number = 2;
+
+  ASSERT_EQ(blob_file_addition.GetBlobFileNumber(), blob_file_number);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobCount(), number_of_blobs);
+  ASSERT_EQ(
+      blob_file_addition.GetTotalBlobBytes(),
+      number_of_blobs * (BlobLogRecord::kHeaderSize + key_size + value_size));
+
+  // Verify the contents of the new blob file as well as the blob references
+  VerifyBlobFile(immutable_cf_options, blob_file_number, column_family_id,
+                 kNoCompression, expected_key_value_pairs, blob_indexes);
+}
+
+TEST_F(BlobFileBuilderTest, BuildAndCheckMultipleFiles) {
+  // Build multiple blob files: file size limit is set to the size of a single
+  // value, so each blob ends up in a file of its own
+  constexpr size_t number_of_blobs = 10;
+  constexpr size_t key_size = 1;
+  constexpr size_t value_size = 10;
+  constexpr size_t value_offset = 1234567890;
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&mock_env_,
+                            "BlobFileBuilderTest_BuildAndCheckMultipleFiles"),
+      0);
+  options.enable_blob_files = true;
+  options.blob_file_size = value_size;
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(
+      TestFileNumberGenerator(), &mock_env_, &fs_, &immutable_cf_options,
+      &mutable_cf_options, &file_options_, job_id, column_family_id,
+      column_family_name, io_priority, write_hint, &blob_file_additions);
+
+  std::vector<std::pair<std::string, std::string>> expected_key_value_pairs(
+      number_of_blobs);
+  std::vector<std::string> blob_indexes(number_of_blobs);
+
+  for (size_t i = 0; i < number_of_blobs; ++i) {
+    auto& expected_key_value = expected_key_value_pairs[i];
+
+    auto& key = expected_key_value.first;
+    key = std::to_string(i);
+    assert(key.size() == key_size);
+
+    auto& value = expected_key_value.second;
+    value = std::to_string(i + value_offset);
+    assert(value.size() == value_size);
+
+    auto& blob_index = blob_indexes[i];
+
+    ASSERT_OK(builder.Add(key, value, &blob_index));
+    ASSERT_FALSE(blob_index.empty());
+  }
+
+  ASSERT_OK(builder.Finish());
+
+  // Check the metadata generated
+  ASSERT_EQ(blob_file_additions.size(), number_of_blobs);
+
+  for (size_t i = 0; i < number_of_blobs; ++i) {
+    const auto& blob_file_addition = blob_file_additions[i];
+
+    ASSERT_EQ(blob_file_addition.GetBlobFileNumber(), i + 2);
+    ASSERT_EQ(blob_file_addition.GetTotalBlobCount(), 1);
+    ASSERT_EQ(blob_file_addition.GetTotalBlobBytes(),
+              BlobLogRecord::kHeaderSize + key_size + value_size);
+  }
+
+  // Verify the contents of the new blob files as well as the blob references
+  for (size_t i = 0; i < number_of_blobs; ++i) {
+    std::vector<std::pair<std::string, std::string>> expected_key_value_pair{
+        expected_key_value_pairs[i]};
+    std::vector<std::string> blob_index{blob_indexes[i]};
+
+    VerifyBlobFile(immutable_cf_options, i + 2, column_family_id,
+                   kNoCompression, expected_key_value_pair, blob_index);
+  }
+}
+
+TEST_F(BlobFileBuilderTest, InlinedValues) {
+  // All values are below the min_blob_size threshold; no blob files get written
+  constexpr size_t number_of_blobs = 10;
+  constexpr size_t key_size = 1;
+  constexpr size_t value_size = 10;
+  constexpr size_t value_offset = 1234567890;
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&mock_env_, "BlobFileBuilderTest_InlinedValues"),
+      0);
+  options.enable_blob_files = true;
+  options.min_blob_size = 1024;
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(
+      TestFileNumberGenerator(), &mock_env_, &fs_, &immutable_cf_options,
+      &mutable_cf_options, &file_options_, job_id, column_family_id,
+      column_family_name, io_priority, write_hint, &blob_file_additions);
+
+  for (size_t i = 0; i < number_of_blobs; ++i) {
+    const std::string key = std::to_string(i);
+    assert(key.size() == key_size);
+
+    const std::string value = std::to_string(i + value_offset);
+    assert(value.size() == value_size);
+
+    std::string blob_index;
+    ASSERT_OK(builder.Add(key, value, &blob_index));
+    ASSERT_TRUE(blob_index.empty());
+  }
+
+  ASSERT_OK(builder.Finish());
+
+  // Check the metadata generated
+  ASSERT_TRUE(blob_file_additions.empty());
+}
+
+TEST_F(BlobFileBuilderTest, Compression) {
+  // Build a blob file with a compressed blob
+  if (!Snappy_Supported()) {
+    return;
+  }
+
+  constexpr size_t key_size = 1;
+  constexpr size_t value_size = 100;
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&mock_env_, "BlobFileBuilderTest_Compression"), 0);
+  options.enable_blob_files = true;
+  options.blob_compression_type = kSnappyCompression;
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(
+      TestFileNumberGenerator(), &mock_env_, &fs_, &immutable_cf_options,
+      &mutable_cf_options, &file_options_, job_id, column_family_id,
+      column_family_name, io_priority, write_hint, &blob_file_additions);
+
+  const std::string key("1");
+  const std::string uncompressed_value(value_size, 'x');
+
+  std::string blob_index;
+
+  ASSERT_OK(builder.Add(key, uncompressed_value, &blob_index));
+  ASSERT_FALSE(blob_index.empty());
+
+  ASSERT_OK(builder.Finish());
+
+  // Check the metadata generated
+  ASSERT_EQ(blob_file_additions.size(), 1);
+
+  const auto& blob_file_addition = blob_file_additions[0];
+
+  constexpr uint64_t blob_file_number = 2;
+
+  ASSERT_EQ(blob_file_addition.GetBlobFileNumber(), blob_file_number);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobCount(), 1);
+
+  CompressionOptions opts;
+  CompressionContext context(kSnappyCompression);
+  constexpr uint64_t sample_for_compression = 0;
+
+  CompressionInfo info(opts, context, CompressionDict::GetEmptyDict(),
+                       kSnappyCompression, sample_for_compression);
+
+  std::string compressed_value;
+  ASSERT_TRUE(Snappy_Compress(info, uncompressed_value.data(),
+                              uncompressed_value.size(), &compressed_value));
+
+  ASSERT_EQ(blob_file_addition.GetTotalBlobBytes(),
+            BlobLogRecord::kHeaderSize + key_size + compressed_value.size());
+
+  // Verify the contents of the new blob file as well as the blob reference
+  std::vector<std::pair<std::string, std::string>> expected_key_value_pairs{
+      {key, compressed_value}};
+  std::vector<std::string> blob_indexes{blob_index};
+
+  VerifyBlobFile(immutable_cf_options, blob_file_number, column_family_id,
+                 kSnappyCompression, expected_key_value_pairs, blob_indexes);
+}
+
+TEST_F(BlobFileBuilderTest, CompressionError) {
+  // Simulate an error during compression
+  if (!Snappy_Supported()) {
+    return;
+  }
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&mock_env_, "BlobFileBuilderTest_CompressionError"),
+      0);
+  options.enable_blob_files = true;
+  options.blob_compression_type = kSnappyCompression;
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(
+      TestFileNumberGenerator(), &mock_env_, &fs_, &immutable_cf_options,
+      &mutable_cf_options, &file_options_, job_id, column_family_id,
+      column_family_name, io_priority, write_hint, &blob_file_additions);
+
+  SyncPoint::GetInstance()->SetCallBack("CompressData:TamperWithReturnValue",
+                                        [](void* arg) {
+                                          bool* ret = static_cast<bool*>(arg);
+                                          *ret = false;
+                                        });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  constexpr char key[] = "1";
+  constexpr char value[] = "deadbeef";
+
+  std::string blob_index;
+
+  ASSERT_TRUE(builder.Add(key, value, &blob_index).IsCorruption());
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+TEST_F(BlobFileBuilderTest, Checksum) {
+  // Build a blob file with checksum
+
+  class DummyFileChecksumGenerator : public FileChecksumGenerator {
+   public:
+    void Update(const char* /* data */, size_t /* n */) override {}
+
+    void Finalize() override {}
+
+    std::string GetChecksum() const override { return std::string("dummy"); }
+
+    const char* Name() const override { return "DummyFileChecksum"; }
+  };
+
+  class DummyFileChecksumGenFactory : public FileChecksumGenFactory {
+   public:
+    std::unique_ptr<FileChecksumGenerator> CreateFileChecksumGenerator(
+        const FileChecksumGenContext& /* context */) override {
+      return std::unique_ptr<FileChecksumGenerator>(
+          new DummyFileChecksumGenerator);
+    }
+
+    const char* Name() const override { return "DummyFileChecksumGenFactory"; }
+  };
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&mock_env_, "BlobFileBuilderTest_Checksum"), 0);
+  options.enable_blob_files = true;
+  options.file_checksum_gen_factory =
+      std::make_shared<DummyFileChecksumGenFactory>();
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(
+      TestFileNumberGenerator(), &mock_env_, &fs_, &immutable_cf_options,
+      &mutable_cf_options, &file_options_, job_id, column_family_id,
+      column_family_name, io_priority, write_hint, &blob_file_additions);
+
+  const std::string key("1");
+  const std::string value("deadbeef");
+
+  std::string blob_index;
+
+  ASSERT_OK(builder.Add(key, value, &blob_index));
+  ASSERT_FALSE(blob_index.empty());
+
+  ASSERT_OK(builder.Finish());
+
+  // Check the metadata generated
+  ASSERT_EQ(blob_file_additions.size(), 1);
+
+  const auto& blob_file_addition = blob_file_additions[0];
+
+  constexpr uint64_t blob_file_number = 2;
+
+  ASSERT_EQ(blob_file_addition.GetBlobFileNumber(), blob_file_number);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobCount(), 1);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobBytes(),
+            BlobLogRecord::kHeaderSize + key.size() + value.size());
+  ASSERT_EQ(blob_file_addition.GetChecksumMethod(), "DummyFileChecksum");
+  ASSERT_EQ(blob_file_addition.GetChecksumValue(), "dummy");
+
+  // Verify the contents of the new blob file as well as the blob reference
+  std::vector<std::pair<std::string, std::string>> expected_key_value_pairs{
+      {key, value}};
+  std::vector<std::string> blob_indexes{blob_index};
+
+  VerifyBlobFile(immutable_cf_options, blob_file_number, column_family_id,
+                 kNoCompression, expected_key_value_pairs, blob_indexes);
+}
+
+class BlobFileBuilderIOErrorTest
+    : public testing::Test,
+      public testing::WithParamInterface<std::string> {
+ protected:
+  BlobFileBuilderIOErrorTest()
+      : mock_env_(Env::Default()),
+        fault_injection_env_(&mock_env_),
+        fs_(&fault_injection_env_),
+        sync_point_(GetParam()) {}
+
+  MockEnv mock_env_;
+  FaultInjectionTestEnv fault_injection_env_;
+  LegacyFileSystemWrapper fs_;
+  FileOptions file_options_;
+  std::string sync_point_;
+};
+
+INSTANTIATE_TEST_CASE_P(
+    BlobFileBuilderTest, BlobFileBuilderIOErrorTest,
+    ::testing::ValuesIn(std::vector<std::string>{
+        "BlobFileBuilder::OpenBlobFileIfNeeded:NewWritableFile",
+        "BlobFileBuilder::OpenBlobFileIfNeeded:WriteHeader",
+        "BlobFileBuilder::WriteBlobToFile:AddRecord",
+        "BlobFileBuilder::WriteBlobToFile:AppendFooter"}));
+
+TEST_P(BlobFileBuilderIOErrorTest, IOError) {
+  // Simulate an I/O error during the specified step of Add()
+  // Note: blob_file_size will be set to value_size in order for the first blob
+  // to trigger close
+  constexpr size_t value_size = 8;
+
+  Options options;
+  options.cf_paths.emplace_back(
+      test::PerThreadDBPath(&fault_injection_env_,
+                            "BlobFileBuilderIOErrorTest_IOError"),
+      0);
+  options.enable_blob_files = true;
+  options.blob_file_size = value_size;
+
+  ImmutableCFOptions immutable_cf_options(options);
+  MutableCFOptions mutable_cf_options(options);
+
+  constexpr int job_id = 1;
+  constexpr uint32_t column_family_id = 123;
+  constexpr char column_family_name[] = "foobar";
+  constexpr Env::IOPriority io_priority = Env::IO_HIGH;
+  constexpr Env::WriteLifeTimeHint write_hint = Env::WLTH_MEDIUM;
+
+  std::vector<BlobFileAddition> blob_file_additions;
+
+  BlobFileBuilder builder(TestFileNumberGenerator(), &fault_injection_env_,
+                          &fs_, &immutable_cf_options, &mutable_cf_options,
+                          &file_options_, job_id, column_family_id,
+                          column_family_name, io_priority, write_hint,
+                          &blob_file_additions);
+
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
+    fault_injection_env_.SetFilesystemActive(false,
+                                             Status::IOError(sync_point_));
+  });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  constexpr char key[] = "1";
+  constexpr char value[] = "deadbeef";
+
+  std::string blob_index;
+
+  ASSERT_TRUE(builder.Add(key, value, &blob_index).IsIOError());
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+}  // namespace ROCKSDB_NAMESPACE
+
+int main(int argc, char** argv) {
+  ::testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}

db/blob/blob_file_meta.cc

@@ -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;

db/blob/blob_file_meta.h

@@ -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_;
 };

db/blob/blob_index.h

@@ -3,7 +3,6 @@
 //  COPYING file in the root directory) and Apache 2.0 License
 //  (found in the LICENSE.Apache file in the root directory).
 #pragma once
-#ifndef ROCKSDB_LITE
 
 #include <sstream>
 #include <string>
@@ -176,4 +175,3 @@ class BlobIndex {
 };
 
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/blob_log_format.cc

@@ -3,15 +3,13 @@
 //  COPYING file in the root directory) and Apache 2.0 License
 //  (found in the LICENSE.Apache file in the root directory).
 //
-#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 +142,4 @@ Status BlobLogRecord::CheckBlobCRC() const {
   return Status::OK();
 }
 
-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/blob_log_format.h

@@ -7,9 +7,6 @@
 
 #pragma once
 
-#ifndef ROCKSDB_LITE
-
-#include <limits>
 #include <memory>
 #include <utility>
 
@@ -19,11 +16,9 @@
 #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 +123,4 @@ struct BlobLogRecord {
   Status CheckBlobCRC() const;
 };
 
-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/blob_log_reader.cc

@@ -3,9 +3,8 @@
 //  COPYING file in the root directory) and Apache 2.0 License
 //  (found in the LICENSE.Apache file in the root directory).
 //
-#ifndef ROCKSDB_LITE
 
-#include "utilities/blob_db/blob_log_reader.h"
+#include "db/blob/blob_log_reader.h"
 
 #include <algorithm>
 
@@ -14,17 +13,19 @@
 #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) {
+  assert(file_);
+
   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);
@@ -39,9 +40,12 @@ Status Reader::ReadSlice(uint64_t size, Slice* slice, char* buf) {
   return s;
 }
 
-Status Reader::ReadHeader(BlobLogHeader* header) {
-  assert(file_.get() != nullptr);
+Status BlobLogReader::ReadHeader(BlobLogHeader* header) {
   assert(next_byte_ == 0);
+
+  static_assert(BlobLogHeader::kSize <= sizeof(header_buf_),
+                "Buffer is smaller than BlobLogHeader::kSize");
+
   Status s = ReadSlice(BlobLogHeader::kSize, &buffer_, header_buf_);
   if (!s.ok()) {
     return s;
@@ -54,8 +58,11 @@ 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) {
+  static_assert(BlobLogRecord::kHeaderSize <= sizeof(header_buf_),
+                "Buffer is smaller than BlobLogRecord::kHeaderSize");
+
   Status s = ReadSlice(BlobLogRecord::kHeaderSize, &buffer_, header_buf_);
   if (!s.ok()) {
     return s;
@@ -101,6 +108,20 @@ Status Reader::ReadRecord(BlobLogRecord* record, ReadLevel level,
   return s;
 }
 
-}  // namespace blob_db
+Status BlobLogReader::ReadFooter(BlobLogFooter* footer) {
+  static_assert(BlobLogFooter::kSize <= sizeof(header_buf_),
+                "Buffer is smaller than BlobLogFooter::kSize");
+
+  Status s = ReadSlice(BlobLogFooter::kSize, &buffer_, header_buf_);
+  if (!s.ok()) {
+    return s;
+  }
+
+  if (buffer_.size() != BlobLogFooter::kSize) {
+    return Status::Corruption("EOF reached before file footer");
+  }
+
+  return footer->DecodeFrom(buffer_);
+}
+
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/blob_log_reader.h

@@ -5,32 +5,29 @@
 //
 #pragma once
 
-#ifndef ROCKSDB_LITE
-
 #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 +36,14 @@ 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);
 
@@ -59,6 +56,8 @@ class Reader {
   Status ReadRecord(BlobLogRecord* record, ReadLevel level = kReadHeader,
                     uint64_t* blob_offset = nullptr);
 
+  Status ReadFooter(BlobLogFooter* footer);
+
   void ResetNextByte() { next_byte_ = 0; }
 
   uint64_t GetNextByte() const { return next_byte_; }
@@ -77,6 +76,4 @@ class Reader {
   uint64_t next_byte_;
 };
 
-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/blob_log_writer.cc

@@ -2,44 +2,45 @@
 //  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).
-#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() {
+BlobLogWriter::~BlobLogWriter() = default;
+
+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 +57,9 @@ Status Writer::WriteHeader(BlobLogHeader& header) {
   return s;
 }
 
-Status Writer::AppendFooter(BlobLogFooter& footer) {
+Status BlobLogWriter::AppendFooter(BlobLogFooter& footer,
+                                   std::string* checksum_method,
+                                   std::string* checksum_value) {
   assert(block_offset_ != 0);
   assert(last_elem_type_ == kEtFileHdr || last_elem_type_ == kEtRecord);
 
@@ -66,7 +69,34 @@ 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();
+
+      if (s.ok()) {
+        assert(!!checksum_method == !!checksum_value);
+
+        if (checksum_method) {
+          assert(checksum_method->empty());
+
+          std::string method = dest_->GetFileChecksumFuncName();
+          if (method != kUnknownFileChecksumFuncName) {
+            *checksum_method = std::move(method);
+          }
+        }
+        if (checksum_value) {
+          assert(checksum_value->empty());
+
+          std::string value = dest_->GetFileChecksum();
+          if (value != kUnknownFileChecksum) {
+            *checksum_value = std::move(value);
+          }
+        }
+      }
+    }
+
     dest_.reset();
   }
 
@@ -76,9 +106,9 @@ 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 +119,8 @@ 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 +131,8 @@ 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 +140,10 @@ 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 +165,4 @@ Status Writer::EmitPhysicalRecord(const std::string& headerbuf,
   return s;
 }
 
-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/blob_log_writer.h

@@ -4,46 +4,42 @@
 //  (found in the LICENSE.Apache file in the root directory).
 #pragma once
 
-#ifndef ROCKSDB_LITE
-
 #include <cstdint>
 #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();
 
   static void ConstructBlobHeader(std::string* buf, const Slice& key,
                                   const Slice& val, uint64_t expiration);
@@ -58,7 +54,8 @@ class Writer {
                             const Slice& val, uint64_t* key_offset,
                             uint64_t* blob_offset);
 
-  Status AppendFooter(BlobLogFooter& footer);
+  Status AppendFooter(BlobLogFooter& footer, std::string* checksum_method,
+                      std::string* checksum_value);
 
   Status WriteHeader(BlobLogHeader& header);
 
@@ -68,20 +65,14 @@ 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 +80,4 @@ class Writer {
   ElemType last_elem_type_;
 };
 
-}  // namespace blob_db
 }  // namespace ROCKSDB_NAMESPACE
-#endif  // ROCKSDB_LITE

db/blob/db_blob_index_test.cc

@@ -41,12 +41,13 @@ class DBBlobIndexTest : public DBTestBase {
                                        Tier::kImmutableMemtables,
                                        Tier::kL0SstFile, Tier::kLnSstFile};
 
-  DBBlobIndexTest() : DBTestBase("/db_blob_index_test") {}
+  DBBlobIndexTest()
+      : DBTestBase("/db_blob_index_test", /*env_do_fsync=*/true) {}
 
   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,

db/builder.cc

@@ -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);
 }
 
@@ -82,16 +83,20 @@ Status BuildTable(
     uint64_t sample_for_compression, const CompressionOptions& compression_opts,
     bool paranoid_file_checks, InternalStats* internal_stats,
     TableFileCreationReason reason, IOStatus* io_status,
-    EventLogger* event_logger, int job_id, const Env::IOPriority io_priority,
+    const std::shared_ptr<IOTracer>& io_tracer, 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(
@@ -102,12 +107,13 @@ Status BuildTable(
 
   std::string fname = TableFileName(ioptions.cf_paths, meta->fd.GetNumber(),
                                     meta->fd.GetPathId());
+  std::string file_checksum = kUnknownFileChecksum;
+  std::string file_checksum_func_name = kUnknownFileChecksumFuncName;
 #ifndef ROCKSDB_LITE
   EventHelpers::NotifyTableFileCreationStarted(
       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;
@@ -122,19 +128,25 @@ 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,
-            job_id, meta->fd, kInvalidBlobFileNumber, tp, reason, s);
+            job_id, meta->fd, kInvalidBlobFileNumber, tp, reason, s,
+            file_checksum, file_checksum_func_name);
         return s;
       }
       file->SetIOPriority(io_priority);
       file->SetWriteLifeTimeHint(write_hint);
 
       file_writer.reset(new WritableFileWriter(
-          std::move(file), fname, file_options, env, ioptions.statistics,
-          ioptions.listeners, ioptions.file_checksum_gen_factory));
+          std::move(file), fname, file_options, env, io_tracer,
+          ioptions.statistics, ioptions.listeners,
+          ioptions.file_checksum_gen_factory));
 
       builder = NewTableBuilder(
           ioptions, mutable_cf_options, internal_comparator,
@@ -142,7 +154,7 @@ 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 +173,11 @@ 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 +208,10 @@ 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();
@@ -206,20 +226,24 @@ Status BuildTable(
     delete builder;
 
     // Finish and check for file errors
+    TEST_SYNC_POINT("BuildTable:BeforeSyncTable");
     if (s.ok() && !empty) {
       StopWatch sw(env, ioptions.statistics, TABLE_SYNC_MICROS);
       *io_status = file_writer->Sync(ioptions.use_fsync);
     }
-    if (io_status->ok() && !empty) {
+    TEST_SYNC_POINT("BuildTable:BeforeCloseTableFile");
+    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();
+      file_checksum = meta->file_checksum;
+      file_checksum_func_name = meta->file_checksum_func_name;
     }
 
-    if (!io_status->ok()) {
+    if (s.ok()) {
       s = *io_status;
     }
 
@@ -232,21 +256,32 @@ Status BuildTable(
       // No matter whether use_direct_io_for_flush_and_compaction is true,
       // we will regrad this verification as user reads since the goal is
       // to cache it here for further user reads
+      ReadOptions read_options;
       std::unique_ptr<InternalIterator> it(table_cache->NewIterator(
-          ReadOptions(), file_options, internal_comparator, *meta,
+          read_options, file_options, internal_comparator, *meta,
           nullptr /* range_del_agg */,
           mutable_cf_options.prefix_extractor.get(), nullptr,
           (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("Paranoid checksums do not match");
+        }
       }
     }
   }
@@ -266,7 +301,8 @@ Status BuildTable(
   // Output to event logger and fire events.
   EventHelpers::LogAndNotifyTableFileCreationFinished(
       event_logger, ioptions.listeners, dbname, column_family_name, fname,
-      job_id, meta->fd, meta->oldest_blob_file_number, tp, reason, s);
+      job_id, meta->fd, meta->oldest_blob_file_number, tp, reason, s,
+      file_checksum, file_checksum_func_name);
 
   return s;
 }

db/builder.h

@@ -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
@@ -78,11 +79,13 @@ extern Status BuildTable(
     const uint64_t sample_for_compression,
     const CompressionOptions& compression_opts, bool paranoid_file_checks,
     InternalStats* internal_stats, TableFileCreationReason reason,
-    IOStatus* io_status, EventLogger* event_logger = nullptr, int job_id = 0,
+    IOStatus* io_status, const std::shared_ptr<IOTracer>& io_tracer,
+    EventLogger* event_logger = nullptr, int job_id = 0,
     const Env::IOPriority io_priority = Env::IO_HIGH,
     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

db/c.cc

@@ -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;
@@ -698,6 +707,41 @@ rocksdb_t* rocksdb_open_column_families(
   return result;
 }
 
+rocksdb_t* rocksdb_open_column_families_with_ttl(
+    const rocksdb_options_t* db_options, const char* name,
+    int num_column_families, const char* const* column_family_names,
+    const rocksdb_options_t* const* column_family_options,
+    rocksdb_column_family_handle_t** column_family_handles, const int* ttls,
+    char** errptr) {
+  std::vector<int32_t> ttls_vec;
+  std::vector<ColumnFamilyDescriptor> column_families;
+  for (int i = 0; i < num_column_families; i++) {
+    ttls_vec.push_back(ttls[i]);
+
+    column_families.push_back(ColumnFamilyDescriptor(
+        std::string(column_family_names[i]),
+        ColumnFamilyOptions(column_family_options[i]->rep)));
+  }
+
+  ROCKSDB_NAMESPACE::DBWithTTL* db;
+  std::vector<ColumnFamilyHandle*> handles;
+  if (SaveError(errptr, ROCKSDB_NAMESPACE::DBWithTTL::Open(
+                            DBOptions(db_options->rep), std::string(name),
+                            column_families, &handles, &db, ttls_vec))) {
+    return nullptr;
+  }
+
+  for (size_t i = 0; i < handles.size(); i++) {
+    rocksdb_column_family_handle_t* c_handle =
+        new rocksdb_column_family_handle_t;
+    c_handle->rep = handles[i];
+    column_family_handles[i] = c_handle;
+  }
+  rocksdb_t* result = new rocksdb_t;
+  result->rep = db;
+  return result;
+}
+
 rocksdb_t* rocksdb_open_for_read_only_column_families(
     const rocksdb_options_t* db_options, const char* name,
     int num_column_families, const char* const* column_family_names,
@@ -796,6 +840,18 @@ rocksdb_column_family_handle_t* rocksdb_create_column_family(
   return handle;
 }
 
+rocksdb_column_family_handle_t* rocksdb_create_column_family_with_ttl(
+    rocksdb_t* db, const rocksdb_options_t* column_family_options,
+    const char* column_family_name, int ttl, char** errptr) {
+  ROCKSDB_NAMESPACE::DBWithTTL* db_with_ttl =
+      static_cast<ROCKSDB_NAMESPACE::DBWithTTL*>(db->rep);
+  rocksdb_column_family_handle_t* handle = new rocksdb_column_family_handle_t;
+  SaveError(errptr, db_with_ttl->CreateColumnFamilyWithTtl(
+                        ColumnFamilyOptions(column_family_options->rep),
+                        std::string(column_family_name), &(handle->rep), ttl));
+  return handle;
+}
+
 void rocksdb_drop_column_family(
     rocksdb_t* db,
     rocksdb_column_family_handle_t* handle,
@@ -996,6 +1052,55 @@ 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) {
@@ -2401,6 +2506,10 @@ void rocksdb_options_set_max_compaction_bytes(rocksdb_options_t* opt,
   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);
@@ -2476,14 +2585,26 @@ 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) {
@@ -2537,6 +2658,10 @@ 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;
@@ -2551,16 +2676,28 @@ 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*/) {}
@@ -2570,41 +2707,86 @@ 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) {
@@ -2627,41 +2809,78 @@ 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;
 }
@@ -2731,67 +2950,130 @@ 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
@@ -2802,19 +3084,38 @@ 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();
 }
@@ -2828,11 +3129,20 @@ 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) {
@@ -2867,31 +3177,56 @@ 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);
 }
@@ -2921,6 +3256,10 @@ 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,
@@ -3320,11 +3659,20 @@ void rocksdb_readoptions_set_verify_checksums(
   opt->rep.verify_checksums = v;
 }
 
+unsigned char rocksdb_readoptions_get_verify_checksums(
+    rocksdb_readoptions_t* opt) {
+  return opt->rep.verify_checksums;
+}
+
 void rocksdb_readoptions_set_fill_cache(
     rocksdb_readoptions_t* opt, unsigned char v) {
   opt->rep.fill_cache = v;
 }
 
+unsigned char rocksdb_readoptions_get_fill_cache(rocksdb_readoptions_t* opt) {
+  return opt->rep.fill_cache;
+}
+
 void rocksdb_readoptions_set_snapshot(
     rocksdb_readoptions_t* opt,
     const rocksdb_snapshot_t* snap) {
@@ -3361,11 +3709,19 @@ void rocksdb_readoptions_set_read_tier(
   opt->rep.read_tier = static_cast<ROCKSDB_NAMESPACE::ReadTier>(v);
 }
 
+int rocksdb_readoptions_get_read_tier(rocksdb_readoptions_t* opt) {
+  return static_cast<int>(opt->rep.read_tier);
+}
+
 void rocksdb_readoptions_set_tailing(
     rocksdb_readoptions_t* opt, unsigned char v) {
   opt->rep.tailing = v;
 }
 
+unsigned char rocksdb_readoptions_get_tailing(rocksdb_readoptions_t* opt) {
+  return opt->rep.tailing;
+}
+
 void rocksdb_readoptions_set_managed(
     rocksdb_readoptions_t* opt, unsigned char v) {
   opt->rep.managed = v;
@@ -3376,37 +3732,70 @@ void rocksdb_readoptions_set_readahead_size(
   opt->rep.readahead_size = v;
 }
 
+size_t rocksdb_readoptions_get_readahead_size(rocksdb_readoptions_t* opt) {
+  return opt->rep.readahead_size;
+}
+
 void rocksdb_readoptions_set_prefix_same_as_start(
     rocksdb_readoptions_t* opt, unsigned char v) {
   opt->rep.prefix_same_as_start = v;
 }
 
+unsigned char rocksdb_readoptions_get_prefix_same_as_start(
+    rocksdb_readoptions_t* opt) {
+  return opt->rep.prefix_same_as_start;
+}
+
 void rocksdb_readoptions_set_pin_data(rocksdb_readoptions_t* opt,
                                       unsigned char v) {
   opt->rep.pin_data = v;
 }
 
+unsigned char rocksdb_readoptions_get_pin_data(rocksdb_readoptions_t* opt) {
+  return opt->rep.pin_data;
+}
+
 void rocksdb_readoptions_set_total_order_seek(rocksdb_readoptions_t* opt,
                                               unsigned char v) {
   opt->rep.total_order_seek = v;
 }
 
+unsigned char rocksdb_readoptions_get_total_order_seek(
+    rocksdb_readoptions_t* opt) {
+  return opt->rep.total_order_seek;
+}
+
 void rocksdb_readoptions_set_max_skippable_internal_keys(
     rocksdb_readoptions_t* opt,
     uint64_t v) {
   opt->rep.max_skippable_internal_keys = v;
 }
 
+uint64_t rocksdb_readoptions_get_max_skippable_internal_keys(
+    rocksdb_readoptions_t* opt) {
+  return opt->rep.max_skippable_internal_keys;
+}
+
 void rocksdb_readoptions_set_background_purge_on_iterator_cleanup(
     rocksdb_readoptions_t* opt, unsigned char v) {
   opt->rep.background_purge_on_iterator_cleanup = v;
 }
 
+unsigned char rocksdb_readoptions_get_background_purge_on_iterator_cleanup(
+    rocksdb_readoptions_t* opt) {
+  return opt->rep.background_purge_on_iterator_cleanup;
+}
+
 void rocksdb_readoptions_set_ignore_range_deletions(
     rocksdb_readoptions_t* opt, unsigned char v) {
   opt->rep.ignore_range_deletions = v;
 }
 
+unsigned char rocksdb_readoptions_get_ignore_range_deletions(
+    rocksdb_readoptions_t* opt) {
+  return opt->rep.ignore_range_deletions;
+}
+
 rocksdb_writeoptions_t* rocksdb_writeoptions_create() {
   return new rocksdb_writeoptions_t;
 }
@@ -3420,33 +3809,61 @@ void rocksdb_writeoptions_set_sync(
   opt->rep.sync = v;
 }
 
+unsigned char rocksdb_writeoptions_get_sync(rocksdb_writeoptions_t* opt) {
+  return opt->rep.sync;
+}
+
 void rocksdb_writeoptions_disable_WAL(rocksdb_writeoptions_t* opt, int disable) {
   opt->rep.disableWAL = disable;
 }
 
+unsigned char rocksdb_writeoptions_get_disable_WAL(
+    rocksdb_writeoptions_t* opt) {
+  return opt->rep.disableWAL;
+}
+
 void rocksdb_writeoptions_set_ignore_missing_column_families(
     rocksdb_writeoptions_t* opt,
     unsigned char v) {
   opt->rep.ignore_missing_column_families = v;
 }
 
+unsigned char rocksdb_writeoptions_get_ignore_missing_column_families(
+    rocksdb_writeoptions_t* opt) {
+  return opt->rep.ignore_missing_column_families;
+}
+
 void rocksdb_writeoptions_set_no_slowdown(
     rocksdb_writeoptions_t* opt,
     unsigned char v) {
   opt->rep.no_slowdown = v;
 }
 
+unsigned char rocksdb_writeoptions_get_no_slowdown(
+    rocksdb_writeoptions_t* opt) {
+  return opt->rep.no_slowdown;
+}
+
 void rocksdb_writeoptions_set_low_pri(
     rocksdb_writeoptions_t* opt,
     unsigned char v) {
   opt->rep.low_pri = v;
 }
 
+unsigned char rocksdb_writeoptions_get_low_pri(rocksdb_writeoptions_t* opt) {
+  return opt->rep.low_pri;
+}
+
 void rocksdb_writeoptions_set_memtable_insert_hint_per_batch(
     rocksdb_writeoptions_t* opt, unsigned char v) {
   opt->rep.memtable_insert_hint_per_batch = v;
 }
 
+unsigned char rocksdb_writeoptions_get_memtable_insert_hint_per_batch(
+    rocksdb_writeoptions_t* opt) {
+  return opt->rep.memtable_insert_hint_per_batch;
+}
+
 rocksdb_compactoptions_t* rocksdb_compactoptions_create() {
   return new rocksdb_compactoptions_t;
 }
@@ -3460,21 +3877,40 @@ void rocksdb_compactoptions_set_bottommost_level_compaction(
   opt->rep.bottommost_level_compaction = static_cast<BottommostLevelCompaction>(v);
 }
 
+unsigned char rocksdb_compactoptions_get_bottommost_level_compaction(
+    rocksdb_compactoptions_t* opt) {
+  return static_cast<unsigned char>(opt->rep.bottommost_level_compaction);
+}
+
 void rocksdb_compactoptions_set_exclusive_manual_compaction(
     rocksdb_compactoptions_t* opt, unsigned char v) {
   opt->rep.exclusive_manual_compaction = v;
 }
 
+unsigned char rocksdb_compactoptions_get_exclusive_manual_compaction(
+    rocksdb_compactoptions_t* opt) {
+  return opt->rep.exclusive_manual_compaction;
+}
+
 void rocksdb_compactoptions_set_change_level(rocksdb_compactoptions_t* opt,
                                              unsigned char v) {
   opt->rep.change_level = v;
 }
 
+unsigned char rocksdb_compactoptions_get_change_level(
+    rocksdb_compactoptions_t* opt) {
+  return opt->rep.change_level;
+}
+
 void rocksdb_compactoptions_set_target_level(rocksdb_compactoptions_t* opt,
                                              int n) {
   opt->rep.target_level = n;
 }
 
+int rocksdb_compactoptions_get_target_level(rocksdb_compactoptions_t* opt) {
+  return opt->rep.target_level;
+}
+
 rocksdb_flushoptions_t* rocksdb_flushoptions_create() {
   return new rocksdb_flushoptions_t;
 }
@@ -3488,6 +3924,10 @@ void rocksdb_flushoptions_set_wait(
   opt->rep.wait = v;
 }
 
+unsigned char rocksdb_flushoptions_get_wait(rocksdb_flushoptions_t* opt) {
+  return opt->rep.wait;
+}
+
 rocksdb_cache_t* rocksdb_cache_create_lru(size_t capacity) {
   rocksdb_cache_t* c = new rocksdb_cache_t;
   c->rep = NewLRUCache(capacity);
@@ -3502,6 +3942,10 @@ void rocksdb_cache_set_capacity(rocksdb_cache_t* cache, size_t capacity) {
   cache->rep->SetCapacity(capacity);
 }
 
+size_t rocksdb_cache_get_capacity(rocksdb_cache_t* cache) {
+  return cache->rep->GetCapacity();
+}
+
 size_t rocksdb_cache_get_usage(rocksdb_cache_t* cache) {
   return cache->rep->GetUsage();
 }
@@ -3539,10 +3983,20 @@ void rocksdb_env_set_background_threads(rocksdb_env_t* env, int n) {
   env->rep->SetBackgroundThreads(n);
 }
 
+void rocksdb_env_set_bottom_priority_background_threads(rocksdb_env_t* env,
+                                                        int n) {
+  env->rep->SetBackgroundThreads(n, Env::BOTTOM);
+}
+
 void rocksdb_env_set_high_priority_background_threads(rocksdb_env_t* env, int n) {
   env->rep->SetBackgroundThreads(n, Env::HIGH);
 }
 
+void rocksdb_env_set_low_priority_background_threads(rocksdb_env_t* env,
+                                                     int n) {
+  env->rep->SetBackgroundThreads(n, Env::LOW);
+}
+
 void rocksdb_env_join_all_threads(rocksdb_env_t* env) {
   env->rep->WaitForJoin();
 }

db/c_test.c

@@ -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);
@@ -1564,6 +1587,173 @@ int main(int argc, char** argv) {
     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);
@@ -1607,6 +1797,70 @@ int main(int argc, char** argv) {
     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);
@@ -1745,10 +1999,358 @@ int main(int argc, char** argv) {
     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("read_options");
+  {
+    rocksdb_readoptions_t* ro;
+    ro = rocksdb_readoptions_create();
+
+    rocksdb_readoptions_set_verify_checksums(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_verify_checksums(ro));
+
+    rocksdb_readoptions_set_fill_cache(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_fill_cache(ro));
+
+    rocksdb_readoptions_set_read_tier(ro, 2);
+    CheckCondition(2 == rocksdb_readoptions_get_read_tier(ro));
+
+    rocksdb_readoptions_set_tailing(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_tailing(ro));
+
+    rocksdb_readoptions_set_readahead_size(ro, 100);
+    CheckCondition(100 == rocksdb_readoptions_get_readahead_size(ro));
+
+    rocksdb_readoptions_set_prefix_same_as_start(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_prefix_same_as_start(ro));
+
+    rocksdb_readoptions_set_pin_data(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_pin_data(ro));
+
+    rocksdb_readoptions_set_total_order_seek(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_total_order_seek(ro));
+
+    rocksdb_readoptions_set_max_skippable_internal_keys(ro, 200);
+    CheckCondition(200 ==
+                   rocksdb_readoptions_get_max_skippable_internal_keys(ro));
+
+    rocksdb_readoptions_set_background_purge_on_iterator_cleanup(ro, 1);
+    CheckCondition(
+        1 == rocksdb_readoptions_get_background_purge_on_iterator_cleanup(ro));
+
+    rocksdb_readoptions_set_ignore_range_deletions(ro, 1);
+    CheckCondition(1 == rocksdb_readoptions_get_ignore_range_deletions(ro));
+
+    rocksdb_readoptions_destroy(ro);
+  }
+
+  StartPhase("write_options");
+  {
+    rocksdb_writeoptions_t* wo;
+    wo = rocksdb_writeoptions_create();
+
+    rocksdb_writeoptions_set_sync(wo, 1);
+    CheckCondition(1 == rocksdb_writeoptions_get_sync(wo));
+
+    rocksdb_writeoptions_disable_WAL(wo, 1);
+    CheckCondition(1 == rocksdb_writeoptions_get_disable_WAL(wo));
+
+    rocksdb_writeoptions_set_ignore_missing_column_families(wo, 1);
+    CheckCondition(1 ==
+                   rocksdb_writeoptions_get_ignore_missing_column_families(wo));
+
+    rocksdb_writeoptions_set_no_slowdown(wo, 1);
+    CheckCondition(1 == rocksdb_writeoptions_get_no_slowdown(wo));
+
+    rocksdb_writeoptions_set_low_pri(wo, 1);
+    CheckCondition(1 == rocksdb_writeoptions_get_low_pri(wo));
+
+    rocksdb_writeoptions_set_memtable_insert_hint_per_batch(wo, 1);
+    CheckCondition(1 ==
+                   rocksdb_writeoptions_get_memtable_insert_hint_per_batch(wo));
+
+    rocksdb_writeoptions_destroy(wo);
+  }
+
+  StartPhase("compact_options");
+  {
+    rocksdb_compactoptions_t* co;
+    co = rocksdb_compactoptions_create();
+
+    rocksdb_compactoptions_set_exclusive_manual_compaction(co, 1);
+    CheckCondition(1 ==
+                   rocksdb_compactoptions_get_exclusive_manual_compaction(co));
+
+    rocksdb_compactoptions_set_bottommost_level_compaction(co, 1);
+    CheckCondition(1 ==
+                   rocksdb_compactoptions_get_bottommost_level_compaction(co));
+
+    rocksdb_compactoptions_set_change_level(co, 1);
+    CheckCondition(1 == rocksdb_compactoptions_get_change_level(co));
+
+    rocksdb_compactoptions_set_target_level(co, 1);
+    CheckCondition(1 == rocksdb_compactoptions_get_target_level(co));
+
+    rocksdb_compactoptions_destroy(co);
+  }
+
+  StartPhase("flush_options");
+  {
+    rocksdb_flushoptions_t* fo;
+    fo = rocksdb_flushoptions_create();
+
+    rocksdb_flushoptions_set_wait(fo, 1);
+    CheckCondition(1 == rocksdb_flushoptions_get_wait(fo));
+
+    rocksdb_flushoptions_destroy(fo);
+  }
+
+  StartPhase("cache_options");
+  {
+    rocksdb_cache_t* co;
+    co = rocksdb_cache_create_lru(100);
+    CheckCondition(100 == rocksdb_cache_get_capacity(co));
+
+    rocksdb_cache_set_capacity(co, 200);
+    CheckCondition(200 == rocksdb_cache_get_capacity(co));
+
+    rocksdb_cache_destroy(co);
+  }
+
   StartPhase("iterate_upper_bound");
   {
     // Create new empty database

db/column_family.cc

@@ -12,6 +12,7 @@
 #include <algorithm>
 #include <cinttypes>
 #include <limits>
+#include <sstream>
 #include <string>
 #include <vector>
 
@@ -34,6 +35,7 @@
 #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 {
@@ -147,6 +149,16 @@ Status CheckCompressionSupported(const ColumnFamilyOptions& cf_options) {
           "should be nonzero if we're using zstd's dictionary generator.");
     }
   }
+
+  if (!CompressionTypeSupported(cf_options.blob_compression_type)) {
+    std::ostringstream oss;
+    oss << "The specified blob compression type "
+        << CompressionTypeToString(cf_options.blob_compression_type)
+        << " is not available.";
+
+    return Status::InvalidArgument(oss.str());
+  }
+
   return Status::OK();
 }
 
@@ -346,7 +358,7 @@ 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) {
@@ -487,7 +499,8 @@ ColumnFamilyData::ColumnFamilyData(
     Cache* _table_cache, WriteBufferManager* write_buffer_manager,
     const ColumnFamilyOptions& cf_options, const ImmutableDBOptions& db_options,
     const FileOptions& file_options, ColumnFamilySet* column_family_set,
-    BlockCacheTracer* const block_cache_tracer)
+    BlockCacheTracer* const block_cache_tracer,
+    const std::shared_ptr<IOTracer>& io_tracer)
     : id_(id),
       name_(name),
       dummy_versions_(_dummy_versions),
@@ -545,7 +558,7 @@ ColumnFamilyData::ColumnFamilyData(
     internal_stats_.reset(
         new InternalStats(ioptions_.num_levels, db_options.env, this));
     table_cache_.reset(new TableCache(ioptions_, file_options, _table_cache,
-                                      block_cache_tracer));
+                                      block_cache_tracer, io_tracer));
     if (ioptions_.compaction_style == kCompactionStyleLevel) {
       compaction_picker_.reset(
           new LevelCompactionPicker(ioptions_, &internal_comparator_));
@@ -1042,17 +1055,19 @@ void ColumnFamilyData::CreateNewMemtable(
 }
 
 bool ColumnFamilyData::NeedsCompaction() const {
-  return compaction_picker_->NeedsCompaction(current_->storage_info());
+  return !mutable_cf_options_.disable_auto_compactions &&
+         compaction_picker_->NeedsCompaction(current_->storage_info());
 }
 
 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 +1137,16 @@ 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 +1316,7 @@ 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 +1324,7 @@ 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. ");
@@ -1399,12 +1416,13 @@ ColumnFamilySet::ColumnFamilySet(const std::string& dbname,
                                  Cache* table_cache,
                                  WriteBufferManager* _write_buffer_manager,
                                  WriteController* _write_controller,
-                                 BlockCacheTracer* const block_cache_tracer)
+                                 BlockCacheTracer* const block_cache_tracer,
+                                 const std::shared_ptr<IOTracer>& io_tracer)
     : max_column_family_(0),
       dummy_cfd_(new ColumnFamilyData(
           ColumnFamilyData::kDummyColumnFamilyDataId, "", nullptr, nullptr,
           nullptr, ColumnFamilyOptions(), *db_options, file_options, nullptr,
-          block_cache_tracer)),
+          block_cache_tracer, io_tracer)),
       default_cfd_cache_(nullptr),
       db_name_(dbname),
       db_options_(db_options),
@@ -1412,7 +1430,8 @@ ColumnFamilySet::ColumnFamilySet(const std::string& dbname,
       table_cache_(table_cache),
       write_buffer_manager_(_write_buffer_manager),
       write_controller_(_write_controller),
-      block_cache_tracer_(block_cache_tracer) {
+      block_cache_tracer_(block_cache_tracer),
+      io_tracer_(io_tracer) {
   // initialize linked list
   dummy_cfd_->prev_ = dummy_cfd_;
   dummy_cfd_->next_ = dummy_cfd_;
@@ -1478,7 +1497,7 @@ ColumnFamilyData* ColumnFamilySet::CreateColumnFamily(
   assert(column_families_.find(name) == column_families_.end());
   ColumnFamilyData* new_cfd = new ColumnFamilyData(
       id, name, dummy_versions, table_cache_, write_buffer_manager_, options,
-      *db_options_, file_options_, this, block_cache_tracer_);
+      *db_options_, file_options_, this, block_cache_tracer_, io_tracer_);
   column_families_.insert({name, id});
   column_family_data_.insert({id, new_cfd});
   max_column_family_ = std::max(max_column_family_, id);
@@ -1546,7 +1565,7 @@ 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;

db/column_family.h

@@ -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,
@@ -516,7 +518,8 @@ class ColumnFamilyData {
                    const ImmutableDBOptions& db_options,
                    const FileOptions& file_options,
                    ColumnFamilySet* column_family_set,
-                   BlockCacheTracer* const block_cache_tracer);
+                   BlockCacheTracer* const block_cache_tracer,
+                   const std::shared_ptr<IOTracer>& io_tracer);
 
   std::vector<std::string> GetDbPaths() const;
 
@@ -649,7 +652,8 @@ class ColumnFamilySet {
                   const FileOptions& file_options, Cache* table_cache,
                   WriteBufferManager* _write_buffer_manager,
                   WriteController* _write_controller,
-                  BlockCacheTracer* const block_cache_tracer);
+                  BlockCacheTracer* const block_cache_tracer,
+                  const std::shared_ptr<IOTracer>& io_tracer);
   ~ColumnFamilySet();
 
   ColumnFamilyData* GetDefault() const;
@@ -713,6 +717,7 @@ class ColumnFamilySet {
   WriteBufferManager* write_buffer_manager_;
   WriteController* write_controller_;
   BlockCacheTracer* const block_cache_tracer_;
+  std::shared_ptr<IOTracer> io_tracer_;
 };
 
 // We use ColumnFamilyMemTablesImpl to provide WriteBatch a way to access

db/column_family_test.cc

@@ -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();
@@ -2432,6 +2424,9 @@ 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;
@@ -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;

db/compact_files_test.cc

@@ -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;

db/compacted_db_impl.cc

@@ -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();
@@ -147,7 +149,7 @@ Status CompactedDBImpl::Open(const Options& options,
   std::unique_ptr<CompactedDBImpl> db(new CompactedDBImpl(db_options, dbname));
   Status s = db->Init(options);
   if (s.ok()) {
-    db->StartTimedTasks();
+    db->StartStatsDumpScheduler();
     ROCKS_LOG_INFO(db->immutable_db_options_.info_log,
                    "Opened the db as fully compacted mode");
     LogFlush(db->immutable_db_options_.info_log);

db/compaction/compaction.cc

@@ -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();
 }

db/compaction/compaction.h

@@ -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;
 

db/compaction/compaction_iterator.cc

@@ -40,7 +40,7 @@ CompactionIterator::CompactionIterator(
     const CompactionFilter* compaction_filter,
     const std::atomic<bool>* shutting_down,
     const SequenceNumber preserve_deletes_seqnum,
-    const std::atomic<bool>* manual_compaction_paused,
+    const std::atomic<int>* manual_compaction_paused,
     const std::shared_ptr<Logger> info_log)
     : CompactionIterator(
           input, cmp, merge_helper, last_sequence, snapshots,
@@ -62,7 +62,7 @@ CompactionIterator::CompactionIterator(
     const CompactionFilter* compaction_filter,
     const std::atomic<bool>* shutting_down,
     const SequenceNumber preserve_deletes_seqnum,
-    const std::atomic<bool>* manual_compaction_paused,
+    const std::atomic<int>* manual_compaction_paused,
     const std::shared_ptr<Logger> info_log)
     : input_(input),
       cmp_(cmp),
@@ -86,8 +86,10 @@ CompactionIterator::CompactionIterator(
       info_log_(info_log) {
   assert(compaction_filter_ == nullptr || compaction_ != nullptr);
   assert(snapshots_ != nullptr);
-  bottommost_level_ =
-      compaction_ == nullptr ? false : compaction_->bottommost_level();
+  bottommost_level_ = compaction_ == nullptr
+                          ? false
+                          : compaction_->bottommost_level() &&
+                                !compaction_->allow_ingest_behind();
   if (compaction_ != nullptr) {
     level_ptrs_ = std::vector<size_t>(compaction_->number_levels(), 0);
   }
@@ -142,8 +144,7 @@ void CompactionIterator::Next() {
     if (merge_out_iter_.Valid()) {
       key_ = merge_out_iter_.key();
       value_ = merge_out_iter_.value();
-      bool valid_key __attribute__((__unused__));
-      valid_key =  ParseInternalKey(key_, &ikey_);
+      bool valid_key = ParseInternalKey(key_, &ikey_);
       // MergeUntil stops when it encounters a corrupt key and does not
       // include them in the result, so we expect the keys here to be valid.
       assert(valid_key);
@@ -182,7 +183,7 @@ 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 +226,31 @@ 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 +264,7 @@ 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 +277,6 @@ 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 +313,9 @@ 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 +335,9 @@ 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;
         }
       }
     }
@@ -331,8 +351,7 @@ void CompactionIterator::NextFromInput() {
     // If there are no snapshots, then this kv affect visibility at tip.
     // Otherwise, search though all existing snapshots to find the earliest
     // snapshot that is affected by this kv.
-    SequenceNumber last_sequence __attribute__((__unused__));
-    last_sequence = current_user_key_sequence_;
+    SequenceNumber last_sequence = current_user_key_sequence_;
     current_user_key_sequence_ = ikey_.sequence;
     SequenceNumber last_snapshot = current_user_key_snapshot_;
     SequenceNumber prev_snapshot = 0;  // 0 means no previous snapshot
@@ -544,11 +563,14 @@ void CompactionIterator::NextFromInput() {
       // Handle the case where we have a delete key at the bottom most level
       // We can skip outputting the key iff there are no subsequent puts for this
       // key
+      assert(!compaction_ || compaction_->KeyNotExistsBeyondOutputLevel(
+                                 ikey_.user_key, &level_ptrs_));
       ParsedInternalKey next_ikey;
       input_->Next();
       // Skip over all versions of this key that happen to occur in the same snapshot
       // range as the delete
-      while (input_->Valid() && ParseInternalKey(input_->key(), &next_ikey) &&
+      while (!IsPausingManualCompaction() && !IsShuttingDown() &&
+             input_->Valid() && ParseInternalKey(input_->key(), &next_ikey) &&
              cmp_->Equal(ikey_.user_key, next_ikey.user_key) &&
              (prev_snapshot == 0 ||
               DEFINITELY_NOT_IN_SNAPSHOT(next_ikey.sequence, prev_snapshot))) {
@@ -585,8 +607,7 @@ void CompactionIterator::NextFromInput() {
         //       These will be correctly set below.
         key_ = merge_out_iter_.key();
         value_ = merge_out_iter_.value();
-        bool valid_key __attribute__((__unused__));
-        valid_key = ParseInternalKey(key_, &ikey_);
+        bool valid_key = ParseInternalKey(key_, &ikey_);
         // MergeUntil stops when it encounters a corrupt key and does not
         // include them in the result, so we expect the keys here to valid.
         assert(valid_key);

db/compaction/compaction_iterator.h

@@ -59,34 +59,34 @@ class CompactionIterator {
     const Compaction* compaction_;
   };
 
-  CompactionIterator(
-      InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
-      SequenceNumber last_sequence, std::vector<SequenceNumber>* snapshots,
-      SequenceNumber earliest_write_conflict_snapshot,
-      const SnapshotChecker* snapshot_checker, Env* env,
-      bool report_detailed_time, bool expect_valid_internal_key,
-      CompactionRangeDelAggregator* range_del_agg,
-      const Compaction* compaction = nullptr,
-      const CompactionFilter* compaction_filter = nullptr,
-      const std::atomic<bool>* shutting_down = nullptr,
-      const SequenceNumber preserve_deletes_seqnum = 0,
-      const std::atomic<bool>* manual_compaction_paused = nullptr,
-      const std::shared_ptr<Logger> info_log = nullptr);
+  CompactionIterator(InternalIterator* input, const Comparator* cmp,
+                     MergeHelper* merge_helper, SequenceNumber last_sequence,
+                     std::vector<SequenceNumber>* snapshots,
+                     SequenceNumber earliest_write_conflict_snapshot,
+                     const SnapshotChecker* snapshot_checker, Env* env,
+                     bool report_detailed_time, bool expect_valid_internal_key,
+                     CompactionRangeDelAggregator* range_del_agg,
+                     const Compaction* compaction = nullptr,
+                     const CompactionFilter* compaction_filter = nullptr,
+                     const std::atomic<bool>* shutting_down = nullptr,
+                     const SequenceNumber preserve_deletes_seqnum = 0,
+                     const std::atomic<int>* manual_compaction_paused = nullptr,
+                     const std::shared_ptr<Logger> info_log = nullptr);
 
   // Constructor with custom CompactionProxy, used for tests.
-  CompactionIterator(
-      InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
-      SequenceNumber last_sequence, std::vector<SequenceNumber>* snapshots,
-      SequenceNumber earliest_write_conflict_snapshot,
-      const SnapshotChecker* snapshot_checker, Env* env,
-      bool report_detailed_time, bool expect_valid_internal_key,
-      CompactionRangeDelAggregator* range_del_agg,
-      std::unique_ptr<CompactionProxy> compaction,
-      const CompactionFilter* compaction_filter = nullptr,
-      const std::atomic<bool>* shutting_down = nullptr,
-      const SequenceNumber preserve_deletes_seqnum = 0,
-      const std::atomic<bool>* manual_compaction_paused = nullptr,
-      const std::shared_ptr<Logger> info_log = nullptr);
+  CompactionIterator(InternalIterator* input, const Comparator* cmp,
+                     MergeHelper* merge_helper, SequenceNumber last_sequence,
+                     std::vector<SequenceNumber>* snapshots,
+                     SequenceNumber earliest_write_conflict_snapshot,
+                     const SnapshotChecker* snapshot_checker, Env* env,
+                     bool report_detailed_time, bool expect_valid_internal_key,
+                     CompactionRangeDelAggregator* range_del_agg,
+                     std::unique_ptr<CompactionProxy> compaction,
+                     const CompactionFilter* compaction_filter = nullptr,
+                     const std::atomic<bool>* shutting_down = nullptr,
+                     const SequenceNumber preserve_deletes_seqnum = 0,
+                     const std::atomic<int>* manual_compaction_paused = nullptr,
+                     const std::shared_ptr<Logger> info_log = nullptr);
 
   ~CompactionIterator();
 
@@ -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.
@@ -165,7 +166,7 @@ class CompactionIterator {
   std::unique_ptr<CompactionProxy> compaction_;
   const CompactionFilter* compaction_filter_;
   const std::atomic<bool>* shutting_down_;
-  const std::atomic<bool>* manual_compaction_paused_;
+  const std::atomic<int>* manual_compaction_paused_;
   const SequenceNumber preserve_deletes_seqnum_;
   bool bottommost_level_;
   bool valid_ = false;
@@ -234,7 +235,7 @@ class CompactionIterator {
   bool IsPausingManualCompaction() {
     // This is a best-effort facility, so memory_order_relaxed is sufficient.
     return manual_compaction_paused_ &&
-           manual_compaction_paused_->load(std::memory_order_relaxed);
+           manual_compaction_paused_->load(std::memory_order_relaxed) > 0;
   }
 };
 }  // namespace ROCKSDB_NAMESPACE

db/compaction/compaction_iterator_test.cc

@@ -169,13 +169,15 @@ class FakeCompaction : public CompactionIterator::CompactionProxy {
   Slice GetLargestUserKey() const override {
     return "\xff\xff\xff\xff\xff\xff\xff\xff\xff";
   }
-  bool allow_ingest_behind() const override { return false; }
+  bool allow_ingest_behind() const override { return is_allow_ingest_behind; }
 
   bool preserve_deletes() const override { return false; }
 
   bool key_not_exists_beyond_output_level = false;
 
   bool is_bottommost_level = false;
+
+  bool is_allow_ingest_behind = false;
 };
 
 // A simplifed snapshot checker which assumes each snapshot has a global
@@ -237,6 +239,7 @@ class CompactionIteratorTest : public testing::TestWithParam<bool> {
     if (filter || bottommost_level) {
       compaction_proxy_ = new FakeCompaction();
       compaction_proxy_->is_bottommost_level = bottommost_level;
+      compaction_proxy_->is_allow_ingest_behind = AllowIngestBehind();
       compaction.reset(compaction_proxy_);
     }
     bool use_snapshot_checker = UseSnapshotChecker() || GetParam();
@@ -266,6 +269,8 @@ class CompactionIteratorTest : public testing::TestWithParam<bool> {
 
   virtual bool UseSnapshotChecker() const { return false; }
 
+  virtual bool AllowIngestBehind() const { return false; }
+
   void RunTest(
       const std::vector<std::string>& input_keys,
       const std::vector<std::string>& input_values,
@@ -697,6 +702,18 @@ TEST_P(CompactionIteratorTest, RemoveSingleDeletionAtBottomLevel) {
           nullptr /*compaction_filter*/, true /*bottommost_level*/);
 }
 
+TEST_P(CompactionIteratorTest, ConvertToPutAtBottom) {
+  std::shared_ptr<MergeOperator> merge_op =
+      MergeOperators::CreateStringAppendOperator();
+  RunTest({test::KeyStr("a", 4, kTypeMerge), test::KeyStr("a", 3, kTypeMerge),
+           test::KeyStr("a", 2, kTypeMerge), test::KeyStr("b", 1, kTypeValue)},
+          {"a4", "a3", "a2", "b1"},
+          {test::KeyStr("a", 0, kTypeValue), test::KeyStr("b", 0, kTypeValue)},
+          {"a2,a3,a4", "b1"}, kMaxSequenceNumber /*last_committed_seq*/,
+          merge_op.get(), nullptr /*compaction_filter*/,
+          true /*bottomost_level*/);
+}
+
 INSTANTIATE_TEST_CASE_P(CompactionIteratorTestInstance, CompactionIteratorTest,
                         testing::Values(true, false));
 
@@ -968,6 +985,45 @@ TEST_F(CompactionIteratorWithSnapshotCheckerTest, CompactionFilter_FullMerge) {
       compaction_filter.get());
 }
 
+// Tests how CompactionIterator work together with AllowIngestBehind.
+class CompactionIteratorWithAllowIngestBehindTest
+    : public CompactionIteratorTest {
+ public:
+  bool AllowIngestBehind() const override { return true; }
+};
+
+// When allow_ingest_behind is set, compaction iterator is not targeting
+// the bottommost level since there is no guarantee there won't be further
+// data ingested under the compaction output in future.
+TEST_P(CompactionIteratorWithAllowIngestBehindTest, NoConvertToPutAtBottom) {
+  std::shared_ptr<MergeOperator> merge_op =
+      MergeOperators::CreateStringAppendOperator();
+  RunTest({test::KeyStr("a", 4, kTypeMerge), test::KeyStr("a", 3, kTypeMerge),
+           test::KeyStr("a", 2, kTypeMerge), test::KeyStr("b", 1, kTypeValue)},
+          {"a4", "a3", "a2", "b1"},
+          {test::KeyStr("a", 4, kTypeMerge), test::KeyStr("b", 1, kTypeValue)},
+          {"a2,a3,a4", "b1"}, kMaxSequenceNumber /*last_committed_seq*/,
+          merge_op.get(), nullptr /*compaction_filter*/,
+          true /*bottomost_level*/);
+}
+
+TEST_P(CompactionIteratorWithAllowIngestBehindTest,
+       MergeToPutIfEncounteredPutAtBottom) {
+  std::shared_ptr<MergeOperator> merge_op =
+      MergeOperators::CreateStringAppendOperator();
+  RunTest({test::KeyStr("a", 4, kTypeMerge), test::KeyStr("a", 3, kTypeMerge),
+           test::KeyStr("a", 2, kTypeValue), test::KeyStr("b", 1, kTypeValue)},
+          {"a4", "a3", "a2", "b1"},
+          {test::KeyStr("a", 4, kTypeValue), test::KeyStr("b", 1, kTypeValue)},
+          {"a2,a3,a4", "b1"}, kMaxSequenceNumber /*last_committed_seq*/,
+          merge_op.get(), nullptr /*compaction_filter*/,
+          true /*bottomost_level*/);
+}
+
+INSTANTIATE_TEST_CASE_P(CompactionIteratorWithAllowIngestBehindTestInstance,
+                        CompactionIteratorWithAllowIngestBehindTest,
+                        testing::Values(true, false));
+
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {

db/compaction/compaction_job.cc

@@ -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;
   };
 
@@ -127,9 +134,10 @@ struct CompactionJob::SubcompactionState {
   std::vector<Output> outputs;
   std::unique_ptr<WritableFileWriter> outfile;
   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.
@@ -140,13 +148,13 @@ struct CompactionJob::SubcompactionState {
     }
   }
 
-  uint64_t current_output_file_size;
+  uint64_t current_output_file_size = 0;
 
   // State during the subcompaction
-  uint64_t total_bytes;
-  uint64_t num_output_records;
+  uint64_t total_bytes = 0;
+  uint64_t num_output_records = 0;
   CompactionJobStats compaction_job_stats;
-  uint64_t approx_size;
+  uint64_t approx_size = 0;
   // An index that used to speed up ShouldStopBefore().
   size_t grandparent_index = 0;
   // The number of bytes overlapping between the current output and
@@ -155,49 +163,26 @@ struct CompactionJob::SubcompactionState {
   // A flag determine whether the key has been seen in ShouldStopBefore()
   bool seen_key = false;
 
-  SubcompactionState(Compaction* c, Slice* _start, Slice* _end,
-                     uint64_t size = 0)
-      : compaction(c),
-        start(_start),
-        end(_end),
-        outfile(nullptr),
-        builder(nullptr),
-        current_output_file_size(0),
-        total_bytes(0),
-        num_output_records(0),
-        approx_size(size),
-        grandparent_index(0),
-        overlapped_bytes(0),
-        seen_key(false) {
+  SubcompactionState(Compaction* c, Slice* _start, Slice* _end, uint64_t size)
+      : compaction(c), start(_start), end(_end), approx_size(size) {
     assert(compaction != nullptr);
   }
 
-  SubcompactionState(SubcompactionState&& o) { *this = std::move(o); }
-
-  SubcompactionState& operator=(SubcompactionState&& o) {
-    compaction = std::move(o.compaction);
-    start = std::move(o.start);
-    end = std::move(o.end);
-    status = std::move(o.status);
-    outputs = std::move(o.outputs);
-    outfile = std::move(o.outfile);
-    builder = std::move(o.builder);
-    current_output_file_size = std::move(o.current_output_file_size);
-    total_bytes = std::move(o.total_bytes);
-    num_output_records = std::move(o.num_output_records);
-    compaction_job_stats = std::move(o.compaction_job_stats);
-    approx_size = std::move(o.approx_size);
-    grandparent_index = std::move(o.grandparent_index);
-    overlapped_bytes = std::move(o.overlapped_bytes);
-    seen_key = std::move(o.seen_key);
-    return *this;
+  // 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);
   }
 
-  // Because member std::unique_ptrs do not have these.
-  SubcompactionState(const SubcompactionState&) = delete;
-
-  SubcompactionState& operator=(const SubcompactionState&) = delete;
-
   // 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) {
@@ -305,16 +290,21 @@ 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::shared_ptr<IOTracer>& io_tracer,
+    const std::atomic<int>* 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),
-      fs_(db_options.fs.get()),
+      io_tracer_(io_tracer),
+      fs_(db_options.fs, io_tracer),
       file_options_for_read_(
           fs_->OptimizeForCompactionTableRead(file_options, db_options_)),
       versions_(versions),
@@ -423,7 +413,11 @@ void CompactionJob::Prepare() {
     RecordInHistogram(stats_, NUM_SUBCOMPACTIONS_SCHEDULED,
                       compact_->sub_compact_states.size());
   } else {
-    compact_->sub_compact_states.emplace_back(c, nullptr, nullptr);
+    constexpr Slice* start = nullptr;
+    constexpr Slice* end = nullptr;
+    constexpr uint64_t size = 0;
+
+    compact_->sub_compact_states.emplace_back(c, start, end, size);
   }
 }
 
@@ -606,38 +600,42 @@ 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
@@ -646,22 +644,34 @@ Status CompactionJob::Run() {
         // No matter whether use_direct_io_for_flush_and_compaction is true,
         // we will regard this verification as user reads since the goal is
         // to cache it here for further user reads
+        ReadOptions read_options;
         InternalIterator* iter = cfd->table_cache()->NewIterator(
-            ReadOptions(), file_options_, cfd->internal_comparator(),
-            *files_meta[file_idx], /*range_del_agg=*/nullptr, prefix_extractor,
+            read_options, file_options_, cfd->internal_comparator(),
+            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("Paranoid checksums do not match");
+          }
         }
 
         delete iter;
@@ -719,7 +729,6 @@ 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);
   }
@@ -837,11 +846,20 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
 
   CompactionRangeDelAggregator range_del_agg(&cfd->internal_comparator(),
                                              existing_snapshots_);
+  ReadOptions read_options;
+  read_options.verify_checksums = true;
+  read_options.fill_cache = false;
+  // Compaction iterators shouldn't be confined to a single prefix.
+  // Compactions use Seek() for
+  // (a) concurrent compactions,
+  // (b) CompactionFilter::Decision::kRemoveAndSkipUntil.
+  read_options.total_order_seek = true;
 
   // Although the v2 aggregator is what the level iterator(s) know about,
   // the AddTombstones calls will be propagated down to the v1 aggregator.
-  std::unique_ptr<InternalIterator> input(versions_->MakeInputIterator(
-      sub_compact->compaction, &range_del_agg, file_options_for_read_));
+  std::unique_ptr<InternalIterator> input(
+      versions_->MakeInputIterator(read_options, sub_compact->compaction,
+                                   &range_del_agg, file_options_for_read_));
 
   AutoThreadOperationStageUpdater stage_updater(
       ThreadStatus::STAGE_COMPACTION_PROCESS_KV);
@@ -878,7 +896,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
   TEST_SYNC_POINT_CALLBACK(
       "CompactionJob::Run():PausingManualCompaction:1",
       reinterpret_cast<void*>(
-          const_cast<std::atomic<bool>*>(manual_compaction_paused_)));
+          const_cast<std::atomic<int>*>(manual_compaction_paused_)));
 
   Slice* start = sub_compact->start;
   Slice* end = sub_compact->end;
@@ -894,9 +912,10 @@ 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 +928,12 @@ 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 +960,8 @@ 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();
@@ -966,21 +990,30 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
     TEST_SYNC_POINT_CALLBACK(
         "CompactionJob::Run():PausingManualCompaction:2",
         reinterpret_cast<void*>(
-            const_cast<std::atomic<bool>*>(manual_compaction_paused_)));
+            const_cast<std::atomic<int>*>(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;
@@ -1024,7 +1057,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
   }
   if ((status.ok() || status.IsColumnFamilyDropped()) &&
       (manual_compaction_paused_ &&
-       manual_compaction_paused_->load(std::memory_order_relaxed))) {
+       manual_compaction_paused_->load(std::memory_order_relaxed) > 0)) {
     status = Status::Incomplete(Status::SubCode::kManualCompactionPaused);
   }
   if (status.ok()) {
@@ -1132,6 +1165,8 @@ Status CompactionJob::FinishCompactionOutputFile(
 
   ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
   const Comparator* ucmp = cfd->user_comparator();
+  std::string file_checksum = kUnknownFileChecksum;
+  std::string file_checksum_func_name = kUnknownFileChecksumFuncName;
 
   // Check for iterator errors
   Status s = input_status;
@@ -1205,6 +1240,7 @@ 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 +1268,7 @@ 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 +1342,9 @@ 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 +1354,27 @@ 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();
+    file_checksum = meta->file_checksum;
+    file_checksum_func_name = meta->file_checksum_func_name;
   }
-  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;
@@ -1384,7 +1424,8 @@ Status CompactionJob::FinishCompactionOutputFile(
   EventHelpers::LogAndNotifyTableFileCreationFinished(
       event_logger_, cfd->ioptions()->listeners, dbname_, cfd->GetName(), fname,
       job_id_, output_fd, oldest_blob_file_number, tp,
-      TableFileCreationReason::kCompaction, s);
+      TableFileCreationReason::kCompaction, s, file_checksum,
+      file_checksum_func_name);
 
 #ifndef ROCKSDB_LITE
   // Report new file to SstFileManagerImpl
@@ -1451,10 +1492,22 @@ Status CompactionJob::InstallCompactionResults(
 
 void CompactionJob::RecordCompactionIOStats() {
   RecordTick(stats_, COMPACT_READ_BYTES, IOSTATS(bytes_read));
+  RecordTick(stats_, COMPACT_WRITE_BYTES, IOSTATS(bytes_written));
+  CompactionReason compaction_reason =
+      compact_->compaction->compaction_reason();
+  if (compaction_reason == CompactionReason::kFilesMarkedForCompaction) {
+    RecordTick(stats_, COMPACT_READ_BYTES_MARKED, IOSTATS(bytes_read));
+    RecordTick(stats_, COMPACT_WRITE_BYTES_MARKED, IOSTATS(bytes_written));
+  } else if (compaction_reason == CompactionReason::kPeriodicCompaction) {
+    RecordTick(stats_, COMPACT_READ_BYTES_PERIODIC, IOSTATS(bytes_read));
+    RecordTick(stats_, COMPACT_WRITE_BYTES_PERIODIC, IOSTATS(bytes_written));
+  } else if (compaction_reason == CompactionReason::kTtl) {
+    RecordTick(stats_, COMPACT_READ_BYTES_TTL, IOSTATS(bytes_read));
+    RecordTick(stats_, COMPACT_WRITE_BYTES_TTL, IOSTATS(bytes_written));
+  }
   ThreadStatusUtil::IncreaseThreadOperationProperty(
       ThreadStatus::COMPACTION_BYTES_READ, IOSTATS(bytes_read));
   IOSTATS_RESET(bytes_read);
-  RecordTick(stats_, COMPACT_WRITE_BYTES, IOSTATS(bytes_written));
   ThreadStatusUtil::IncreaseThreadOperationProperty(
       ThreadStatus::COMPACTION_BYTES_WRITTEN, IOSTATS(bytes_written));
   IOSTATS_RESET(bytes_written);
@@ -1483,7 +1536,13 @@ 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_.get(), 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,
@@ -1495,7 +1554,8 @@ Status CompactionJob::OpenCompactionOutputFile(
     EventHelpers::LogAndNotifyTableFileCreationFinished(
         event_logger_, cfd->ioptions()->listeners, dbname_, cfd->GetName(),
         fname, job_id_, FileDescriptor(), kInvalidBlobFileNumber,
-        TableProperties(), TableFileCreationReason::kCompaction, s);
+        TableProperties(), TableFileCreationReason::kCompaction, s,
+        kUnknownFileChecksum, kUnknownFileChecksumFuncName);
     return s;
   }
 
@@ -1523,6 +1583,7 @@ 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);
   }
 
@@ -1532,10 +1593,10 @@ Status CompactionJob::OpenCompactionOutputFile(
       sub_compact->compaction->OutputFilePreallocationSize()));
   const auto& listeners =
       sub_compact->compaction->immutable_cf_options()->listeners;
-  sub_compact->outfile.reset(
-      new WritableFileWriter(std::move(writable_file), fname, file_options_,
-                             env_, db_options_.statistics.get(), listeners,
-                             db_options_.file_checksum_gen_factory.get()));
+  sub_compact->outfile.reset(new WritableFileWriter(
+      std::move(writable_file), fname, file_options_, env_, io_tracer_,
+      db_options_.statistics.get(), listeners,
+      db_options_.file_checksum_gen_factory.get()));
 
   // If the Column family flag is to only optimize filters for hits,
   // we can skip creating filters if this is the bottommost_level where
@@ -1552,7 +1613,8 @@ 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;
 }

db/compaction/compaction_job.h

@@ -62,23 +62,23 @@ class VersionSet;
 // if needed.
 class CompactionJob {
  public:
-  CompactionJob(int job_id, Compaction* compaction,
-                const ImmutableDBOptions& db_options,
-                const FileOptions& file_options, VersionSet* versions,
-                const std::atomic<bool>* shutting_down,
-                const SequenceNumber preserve_deletes_seqnum,
-                LogBuffer* log_buffer, FSDirectory* db_directory,
-                FSDirectory* output_directory, Statistics* stats,
-                InstrumentedMutex* db_mutex, ErrorHandler* db_error_handler,
-                std::vector<SequenceNumber> existing_snapshots,
-                SequenceNumber earliest_write_conflict_snapshot,
-                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 = nullptr);
+  CompactionJob(
+      int job_id, Compaction* compaction, const ImmutableDBOptions& db_options,
+      const FileOptions& file_options, VersionSet* versions,
+      const std::atomic<bool>* shutting_down,
+      const SequenceNumber preserve_deletes_seqnum, LogBuffer* log_buffer,
+      FSDirectory* db_directory, FSDirectory* output_directory,
+      Statistics* stats, InstrumentedMutex* db_mutex,
+      ErrorHandler* db_error_handler,
+      std::vector<SequenceNumber> existing_snapshots,
+      SequenceNumber earliest_write_conflict_snapshot,
+      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::shared_ptr<IOTracer>& io_tracer,
+      const std::atomic<int>* manual_compaction_paused = nullptr,
+      const std::string& db_id = "", const std::string& db_session_id = "");
 
   ~CompactionJob();
 
@@ -152,16 +152,19 @@ 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_;
 
   Env* env_;
-  FileSystem* fs_;
+  std::shared_ptr<IOTracer> io_tracer_;
+  FileSystemPtr fs_;
   // env_option optimized for compaction table reads
   FileOptions file_options_for_read_;
   VersionSet* versions_;
   const std::atomic<bool>* shutting_down_;
-  const std::atomic<bool>* manual_compaction_paused_;
+  const std::atomic<int>* manual_compaction_paused_;
   const SequenceNumber preserve_deletes_seqnum_;
   LogBuffer* log_buffer_;
   FSDirectory* db_directory_;

db/compaction/compaction_job_stats_test.cc

@@ -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);
 }

db/compaction/compaction_job_test.cc

@@ -76,12 +76,13 @@ 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_,
-                                 table_cache_.get(), &write_buffer_manager_,
-                                 &write_controller_,
-                                 /*block_cache_tracer=*/nullptr)),
+        versions_(new VersionSet(
+            dbname_, &db_options_, env_options_, table_cache_.get(),
+            &write_buffer_manager_, &write_controller_,
+            /*block_cache_tracer=*/nullptr, /*io_tracer=*/nullptr)),
         shutting_down_(false),
         preserve_deletes_seqnum_(0),
         mock_table_factory_(new mock::MockTableFactory()),
@@ -248,10 +249,10 @@ class CompactionJobTest : public testing::Test {
   void NewDB() {
     DestroyDB(dbname_, Options());
     EXPECT_OK(env_->CreateDirIfMissing(dbname_));
-    versions_.reset(new VersionSet(dbname_, &db_options_, env_options_,
-                                   table_cache_.get(), &write_buffer_manager_,
-                                   &write_controller_,
-                                   /*block_cache_tracer=*/nullptr));
+    versions_.reset(
+        new VersionSet(dbname_, &db_options_, env_options_, table_cache_.get(),
+                       &write_buffer_manager_, &write_controller_,
+                       /*block_cache_tracer=*/nullptr, /*io_tracer=*/nullptr));
     compaction_job_stats_.Reset();
     SetIdentityFile(env_, dbname_);
 
@@ -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());
@@ -332,7 +334,7 @@ class CompactionJobTest : public testing::Test {
         nullptr, nullptr, &mutex_, &error_handler_, snapshots,
         earliest_write_conflict_snapshot, snapshot_checker, table_cache_,
         &event_logger, false, false, dbname_, &compaction_job_stats_,
-        Env::Priority::USER);
+        Env::Priority::USER, nullptr /* IOTracer */);
     VerifyInitializationOfCompactionJobStats(compaction_job_stats_);
 
     compaction_job.Prepare();
@@ -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 =

db/compaction/compaction_picker.cc

@@ -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,
@@ -638,6 +639,7 @@ Compaction* CompactionPicker::CompactRange(
         compact_range_options.max_subcompactions, /* grandparents */ {},
         /* is manual */ true);
     RegisterCompaction(c);
+    vstorage->ComputeCompactionScore(ioptions_, mutable_cf_options);
     return c;
   }
 
@@ -778,8 +780,8 @@ 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),

db/compaction/compaction_picker.h

@@ -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*/,

db/compaction/compaction_picker_fifo.cc

@@ -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;
 }

db/compaction/compaction_picker_fifo.h

@@ -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);
 };

db/compaction/compaction_picker_level.cc

@@ -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

db/compaction/compaction_picker_level.h

@@ -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(

db/compaction/compaction_picker_test.cc

@@ -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_),
@@ -181,7 +183,8 @@ 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);
 }
 
@@ -192,7 +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);
 }
 
@@ -205,7 +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());
@@ -218,7 +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());
@@ -236,7 +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));
@@ -267,7 +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());
@@ -314,7 +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());
@@ -338,7 +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());
@@ -363,7 +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());
@@ -392,7 +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());
@@ -424,7 +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());
@@ -481,7 +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());
@@ -515,7 +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());
 }
@@ -541,7 +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());
 }
@@ -569,7 +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());
@@ -599,7 +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);
 }
@@ -625,7 +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);
 }
@@ -655,7 +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());
 }
@@ -675,7 +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));
@@ -699,7 +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));
@@ -759,7 +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.
@@ -791,7 +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.
@@ -818,7 +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.
@@ -847,7 +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.
@@ -874,7 +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
@@ -891,7 +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));
@@ -910,7 +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));
@@ -937,7 +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));
@@ -967,7 +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));
@@ -990,7 +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);
 }
 
@@ -1011,7 +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));
@@ -1031,7 +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));
@@ -1059,7 +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));
@@ -1091,7 +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));
@@ -1131,7 +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));
@@ -1169,7 +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));
@@ -1206,7 +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);
 }
 
@@ -1236,7 +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);
 }
 
@@ -1269,7 +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);
 }
 
@@ -1563,7 +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));
@@ -1587,7 +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));
@@ -1616,11 +1659,38 @@ TEST_F(CompactionPickerTest, IsTrivialMoveOn) {
   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;
@@ -1637,7 +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());
 }
@@ -1662,7 +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));
@@ -1671,7 +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));
@@ -1680,7 +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 */));
 }
@@ -1705,7 +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));
@@ -1735,7 +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));
@@ -1767,7 +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));
@@ -1799,7 +1876,8 @@ TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap) {
 
   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);
   // Validate that its a compaction to reduce sorted runs
@@ -1816,7 +1894,8 @@ TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap) {
 
   std::unique_ptr<Compaction> compaction2(
       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(compaction2);
 }
 
@@ -1840,7 +1919,8 @@ TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap2) {
 
   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);
   // Validate that its a delete triggered compaction
@@ -1858,7 +1938,8 @@ TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap2) {
 
   std::unique_ptr<Compaction> compaction2(
       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(compaction2);
 }
 
@@ -1898,7 +1979,8 @@ TEST_F(CompactionPickerTest, UniversalMarkedCompactionStartOutputOverlap) {
 
     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);
     // Validate that its a delete triggered compaction
@@ -1928,11 +2010,177 @@ TEST_F(CompactionPickerTest, UniversalMarkedCompactionStartOutputOverlap) {
     random_index = 0;
     std::unique_ptr<Compaction> compaction2(
         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(compaction2);
     DeleteVersionStorage();
   }
 }
+
+TEST_F(CompactionPickerTest, UniversalMarkedL0NoOverlap) {
+  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 and should result in a full compaction
+  NewVersionStorage(1, kCompactionStyleUniversal);
+
+  // Mark file number 4 for compaction
+  Add(0, 4U, "260", "300", 1 * kFileSize, 0, 260, 300, 0, true);
+  Add(0, 5U, "240", "290", 2 * kFileSize, 0, 201, 250);
+  Add(0, 3U, "301", "350", 4 * kFileSize, 0, 101, 150);
+  Add(0, 6U, "501", "750", 8 * kFileSize, 0, 50, 100);
+  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(0, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(4U, compaction->num_input_files(0));
+  ASSERT_TRUE(file_map_[4].first->being_compacted);
+  ASSERT_TRUE(file_map_[5].first->being_compacted);
+  ASSERT_TRUE(file_map_[3].first->being_compacted);
+  ASSERT_TRUE(file_map_[6].first->being_compacted);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedL0WithOverlap) {
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+
+  // This test covers the case where a file is being compacted, and a
+  // delete triggered compaction is then scheduled. The latter should stop
+  // at the first file being compacted
+  NewVersionStorage(1, kCompactionStyleUniversal);
+
+  // Mark file number 4 for compaction
+  Add(0, 4U, "260", "300", 1 * kFileSize, 0, 260, 300, 0, true);
+  Add(0, 5U, "240", "290", 2 * kFileSize, 0, 201, 250);
+  Add(0, 3U, "301", "350", 4 * kFileSize, 0, 101, 150);
+  Add(0, 6U, "501", "750", 8 * kFileSize, 0, 50, 100);
+  UpdateVersionStorageInfo();
+  file_map_[3].first->being_compacted = true;
+
+  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(0, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(2U, compaction->num_input_files(0));
+  ASSERT_TRUE(file_map_[4].first->being_compacted);
+  ASSERT_TRUE(file_map_[5].first->being_compacted);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedL0Overlap2) {
+  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(1, kCompactionStyleUniversal);
+
+  // Mark file number 4 for compaction
+  Add(0, 4U, "260", "300", 1 * kFileSize, 0, 260, 300);
+  Add(0, 5U, "240", "290", 2 * kFileSize, 0, 201, 250, 0, true);
+  Add(0, 3U, "301", "350", 4 * kFileSize, 0, 101, 150);
+  Add(0, 6U, "501", "750", 8 * kFileSize, 0, 50, 100);
+  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(0, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(3U, compaction->num_input_files(0));
+  ASSERT_TRUE(file_map_[5].first->being_compacted);
+  ASSERT_TRUE(file_map_[3].first->being_compacted);
+  ASSERT_TRUE(file_map_[6].first->being_compacted);
+
+  AddVersionStorage();
+  Add(0, 1U, "150", "200", kFileSize, 0, 500, 550);
+  Add(0, 2U, "201", "250", 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_TRUE(compaction2);
+  ASSERT_EQ(3U, compaction->num_input_files(0));
+  ASSERT_TRUE(file_map_[1].first->being_compacted);
+  ASSERT_TRUE(file_map_[2].first->being_compacted);
+  ASSERT_TRUE(file_map_[4].first->being_compacted);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedManualCompaction) {
+  const uint64_t kFileSize = 100000;
+  const int kNumLevels = 7;
+
+  // This test makes sure the `files_marked_for_compaction_` is updated after
+  // creating manual compaction.
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+
+  NewVersionStorage(kNumLevels, kCompactionStyleUniversal);
+
+  // Add 3 files marked for compaction
+  Add(0, 3U, "301", "350", 4 * kFileSize, 0, 101, 150, 0, true);
+  Add(0, 4U, "260", "300", 1 * kFileSize, 0, 260, 300, 0, true);
+  Add(0, 5U, "240", "290", 2 * kFileSize, 0, 201, 250, 0, true);
+  UpdateVersionStorageInfo();
+
+  // All 3 files are marked for compaction
+  ASSERT_EQ(3U, vstorage_->FilesMarkedForCompaction().size());
+
+  bool manual_conflict = false;
+  InternalKey* manual_end = NULL;
+  std::unique_ptr<Compaction> compaction(
+      universal_compaction_picker.CompactRange(
+          cf_name_, mutable_cf_options_, mutable_db_options_, vstorage_.get(),
+          ColumnFamilyData::kCompactAllLevels, 6, CompactRangeOptions(), NULL,
+          NULL, &manual_end, &manual_conflict, port::kMaxUint64));
+
+  ASSERT_TRUE(compaction);
+
+  ASSERT_EQ(CompactionReason::kManualCompaction,
+            compaction->compaction_reason());
+  ASSERT_EQ(kNumLevels - 1, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(3U, compaction->num_input_files(0));
+  ASSERT_TRUE(file_map_[3].first->being_compacted);
+  ASSERT_TRUE(file_map_[4].first->being_compacted);
+  ASSERT_TRUE(file_map_[5].first->being_compacted);
+
+  // After creating the manual compaction, all files should be cleared from
+  // `FilesMarkedForCompaction`. So they won't be picked by others.
+  ASSERT_EQ(0U, vstorage_->FilesMarkedForCompaction().size());
+}
+
 #endif  // ROCKSDB_LITE
 
 }  // namespace ROCKSDB_NAMESPACE

db/compaction/compaction_picker_universal.cc

@@ -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) {}
@@ -115,6 +114,7 @@ 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_;
@@ -277,11 +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();
 }
 
@@ -730,8 +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,
@@ -841,34 +841,48 @@ 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
     // has the same goals.
-    bool compact = false;
+    int start_index = -1;
 
     start_level_inputs.level = 0;
     start_level_inputs.files.clear();
     output_level = 0;
-    for (FileMetaData* f : vstorage_->LevelFiles(0)) {
+    // Find the first file marked for compaction. Ignore the last file
+    for (size_t loop = 0; loop + 1 < sorted_runs_.size(); loop++) {
+      SortedRun* sr = &sorted_runs_[loop];
+      if (sr->being_compacted) {
+        continue;
+      }
+      FileMetaData* f = vstorage_->LevelFiles(0)[loop];
       if (f->marked_for_compaction) {
-        compact = true;
-      }
-      if (compact) {
         start_level_inputs.files.push_back(f);
+        start_index =
+            static_cast<int>(loop);  // Consider this as the first candidate.
+        break;
       }
     }
+    if (start_index < 0) {
+      // Either no file marked, or they're already being compacted
+      return nullptr;
+    }
+
+    for (size_t loop = start_index + 1; loop < sorted_runs_.size(); loop++) {
+      SortedRun* sr = &sorted_runs_[loop];
+      if (sr->being_compacted) {
+        break;
+      }
+
+      FileMetaData* f = vstorage_->LevelFiles(0)[loop];
+      start_level_inputs.files.push_back(f);
+    }
     if (start_level_inputs.size() <= 1) {
       // If only the last file in L0 is marked for compaction, ignore it
       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;
 
@@ -944,8 +958,8 @@ 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,
@@ -1014,14 +1028,14 @@ 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);

db/compaction/compaction_picker_universal.h

@@ -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; }
 

db/compaction/sst_partitioner.cc

@@ -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

db/comparator_db_test.cc

@@ -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);
     }
 

db/corruption_test.cc

@@ -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 {
@@ -66,6 +69,7 @@ class CorruptionTest : public testing::Test {
 
   ~CorruptionTest() override {
     delete db_;
+    db_ = nullptr;
     DestroyDB(dbname_, Options());
   }
 
@@ -101,22 +105,24 @@ 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 +163,6 @@ 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 +181,7 @@ 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 +191,7 @@ 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 +225,11 @@ 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 +287,7 @@ 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 +310,7 @@ 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 +357,14 @@ 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 +404,7 @@ 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 +423,7 @@ 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 +445,7 @@ 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 +460,7 @@ 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 +487,7 @@ 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 +526,7 @@ 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 +538,7 @@ 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 +564,159 @@ TEST_F(CorruptionTest, FileSystemStateCorrupted) {
   }
 }
 
+static const auto& corruption_modes = {mock::MockTableFactory::kCorruptNone,
+                                       mock::MockTableFactory::kCorruptKey,
+                                       mock::MockTableFactory::kCorruptValue};
+
+TEST_F(CorruptionTest, ParanoidFileChecksOnFlush) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  Status s;
+  for (const auto& mode : corruption_modes) {
+    delete db_;
+    db_ = nullptr;
+    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_));
+    assert(db_ != nullptr);
+    Build(10);
+    s = db_->Flush(FlushOptions());
+    if (mode == mock::MockTableFactory::kCorruptNone) {
+      ASSERT_OK(s);
+    } else {
+      ASSERT_NOK(s);
+    }
+  }
+}
+
+TEST_F(CorruptionTest, ParanoidFileChecksOnCompact) {
+  Options options;
+  options.paranoid_file_checks = true;
+  options.create_if_missing = true;
+  Status s;
+  for (const auto& mode : corruption_modes) {
+    delete db_;
+    db_ = nullptr;
+    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_));
+    assert(db_ != nullptr);
+    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) {

db/cuckoo_table_db_test.cc

@@ -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) {

db/db_block_cache_test.cc

@@ -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 {
 
@@ -32,7 +34,8 @@ class DBBlockCacheTest : public DBTestBase {
   const size_t kNumBlocks = 10;
   const size_t kValueSize = 100;
 
-  DBBlockCacheTest() : DBTestBase("/db_block_cache_test") {}
+  DBBlockCacheTest()
+      : DBTestBase("/db_block_cache_test", /*env_do_fsync=*/true) {}
 
   BlockBasedTableOptions GetTableOptions() {
     BlockBasedTableOptions table_options;
@@ -764,7 +767,7 @@ 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 +854,7 @@ 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());

db/db_bloom_filter_test.cc

@@ -23,7 +23,8 @@ using BFP = BloomFilterPolicy;
 
 class DBBloomFilterTest : public DBTestBase {
  public:
-  DBBloomFilterTest() : DBTestBase("/db_bloom_filter_test") {}
+  DBBloomFilterTest()
+      : DBTestBase("/db_bloom_filter_test", /*env_do_fsync=*/true) {}
 };
 
 class DBBloomFilterTestWithParam : public DBTestBase,
@@ -36,7 +37,8 @@ class DBBloomFilterTestWithParam : public DBTestBase,
   uint32_t format_version_;
 
  public:
-  DBBloomFilterTestWithParam() : DBTestBase("/db_bloom_filter_tests") {}
+  DBBloomFilterTestWithParam()
+      : DBTestBase("/db_bloom_filter_tests", /*env_do_fsync=*/true) {}
 
   ~DBBloomFilterTestWithParam() override {}
 
@@ -1039,7 +1041,7 @@ class DBBloomFilterTestVaryPrefixAndFormatVer
 
  public:
   DBBloomFilterTestVaryPrefixAndFormatVer()
-      : DBTestBase("/db_bloom_filter_tests") {}
+      : DBTestBase("/db_bloom_filter_tests", /*env_do_fsync=*/true) {}
 
   ~DBBloomFilterTestVaryPrefixAndFormatVer() override {}
 

db/db_compaction_filter_test.cc

@@ -21,7 +21,8 @@ static std::string NEW_VALUE = "NewValue";
 
 class DBTestCompactionFilter : public DBTestBase {
  public:
-  DBTestCompactionFilter() : DBTestBase("/db_compaction_filter_test") {}
+  DBTestCompactionFilter()
+      : DBTestBase("/db_compaction_filter_test", /*env_do_fsync=*/true) {}
 };
 
 // Param variant of DBTestBase::ChangeCompactOptions
@@ -126,22 +127,6 @@ class SkipEvenFilter : public CompactionFilter {
   const char* Name() const override { return "DeleteFilter"; }
 };
 
-class DelayFilter : public CompactionFilter {
- public:
-  explicit DelayFilter(DBTestBase* d) : db_test(d) {}
-  bool Filter(int /*level*/, const Slice& /*key*/, const Slice& /*value*/,
-              std::string* /*new_value*/,
-              bool* /*value_changed*/) const override {
-    db_test->env_->addon_time_.fetch_add(1000);
-    return true;
-  }
-
-  const char* Name() const override { return "DelayFilter"; }
-
- private:
-  DBTestBase* db_test;
-};
-
 class ConditionalFilter : public CompactionFilter {
  public:
   explicit ConditionalFilter(const std::string* filtered_value)
@@ -248,20 +233,6 @@ class SkipEvenFilterFactory : public CompactionFilterFactory {
   const char* Name() const override { return "SkipEvenFilterFactory"; }
 };
 
-class DelayFilterFactory : public CompactionFilterFactory {
- public:
-  explicit DelayFilterFactory(DBTestBase* d) : db_test(d) {}
-  std::unique_ptr<CompactionFilter> CreateCompactionFilter(
-      const CompactionFilter::Context& /*context*/) override {
-    return std::unique_ptr<CompactionFilter>(new DelayFilter(db_test));
-  }
-
-  const char* Name() const override { return "DelayFilterFactory"; }
-
- private:
-  DBTestBase* db_test;
-};
-
 class ConditionalFilterFactory : public CompactionFilterFactory {
  public:
   explicit ConditionalFilterFactory(const Slice& filtered_value)
@@ -336,8 +307,9 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
     InternalKeyComparator icmp(options.comparator);
     ReadRangeDelAggregator range_del_agg(&icmp,
                                          kMaxSequenceNumber /* upper_bound */);
+    ReadOptions read_options;
     ScopedArenaIterator iter(dbfull()->NewInternalIterator(
-        &arena, &range_del_agg, kMaxSequenceNumber, handles_[1]));
+        read_options, &arena, &range_del_agg, kMaxSequenceNumber, handles_[1]));
     iter->SeekToFirst();
     ASSERT_OK(iter->status());
     while (iter->Valid()) {
@@ -426,8 +398,9 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
     InternalKeyComparator icmp(options.comparator);
     ReadRangeDelAggregator range_del_agg(&icmp,
                                          kMaxSequenceNumber /* upper_bound */);
+    ReadOptions read_options;
     ScopedArenaIterator iter(dbfull()->NewInternalIterator(
-        &arena, &range_del_agg, kMaxSequenceNumber, handles_[1]));
+        read_options, &arena, &range_del_agg, kMaxSequenceNumber, handles_[1]));
     iter->SeekToFirst();
     ASSERT_OK(iter->status());
     while (iter->Valid()) {
@@ -644,8 +617,9 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
     InternalKeyComparator icmp(options.comparator);
     ReadRangeDelAggregator range_del_agg(&icmp,
                                          kMaxSequenceNumber /* snapshots */);
+    ReadOptions read_options;
     ScopedArenaIterator iter(dbfull()->NewInternalIterator(
-        &arena, &range_del_agg, kMaxSequenceNumber));
+        read_options, &arena, &range_del_agg, kMaxSequenceNumber));
     iter->SeekToFirst();
     ASSERT_OK(iter->status());
     while (iter->Valid()) {

db/db_dynamic_level_test.cc

@@ -15,11 +15,13 @@
 #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 {
  public:
-  DBTestDynamicLevel() : DBTestBase("/db_dynamic_level_test") {}
+  DBTestDynamicLevel()
+      : DBTestBase("/db_dynamic_level_test", /*env_do_fsync=*/true) {}
 };
 
 TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase) {
@@ -80,9 +82,9 @@ 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 +160,7 @@ 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 +177,7 @@ 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 +199,7 @@ 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 +218,7 @@ 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 +259,7 @@ 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 +312,15 @@ 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 +384,7 @@ 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 +443,8 @@ 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 +477,8 @@ 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)));
   }
 

db/db_encryption_test.cc

@@ -16,7 +16,8 @@ namespace ROCKSDB_NAMESPACE {
 
 class DBEncryptionTest : public DBTestBase {
  public:
-  DBEncryptionTest() : DBTestBase("/db_encryption_test") {}
+  DBEncryptionTest()
+      : DBTestBase("/db_encryption_test", /*env_do_fsync=*/true) {}
 };
 
 #ifndef ROCKSDB_LITE

db/db_filesnapshot.cc

@@ -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) {
@@ -167,7 +116,7 @@ Status DBImpl::GetSortedWalFiles(VectorLogPtr& files) {
     // long as deletions are disabled (so the below loop must terminate).
     InstrumentedMutexLock l(&mutex_);
     while (disable_delete_obsolete_files_ > 0 &&
-           pending_purge_obsolete_files_ > 0) {
+           (pending_purge_obsolete_files_ > 0 || bg_purge_scheduled_ > 0)) {
       bg_cv_.Wait();
     }
   }

db/db_flush_test.cc

@@ -13,16 +13,16 @@
 #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 {
 
 class DBFlushTest : public DBTestBase {
  public:
-  DBFlushTest() : DBTestBase("/db_flush_test") {}
+  DBFlushTest() : DBTestBase("/db_flush_test", /*env_do_fsync=*/true) {}
 };
 
 class DBFlushDirectIOTest : public DBFlushTest,
@@ -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();
     }

db/db_impl/db_impl.cc

@@ -65,6 +65,7 @@
 #include "monitoring/iostats_context_imp.h"
 #include "monitoring/perf_context_imp.h"
 #include "monitoring/persistent_stats_history.h"
+#include "monitoring/stats_dump_scheduler.h"
 #include "monitoring/thread_status_updater.h"
 #include "monitoring/thread_status_util.h"
 #include "options/cf_options.h"
@@ -87,10 +88,10 @@
 #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"
@@ -150,8 +151,9 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
       own_info_log_(options.info_log == nullptr),
       initial_db_options_(SanitizeOptions(dbname, options)),
       env_(initial_db_options_.env),
-      fs_(initial_db_options_.env->GetFileSystem()),
+      io_tracer_(std::make_shared<IOTracer>()),
       immutable_db_options_(initial_db_options_),
+      fs_(immutable_db_options_.fs, io_tracer_),
       mutable_db_options_(initial_db_options_),
       stats_(immutable_db_options_.statistics.get()),
       mutex_(stats_, env_, DB_MUTEX_WAIT_MICROS,
@@ -196,13 +198,17 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
       unable_to_release_oldest_log_(false),
       num_running_ingest_file_(0),
 #ifndef ROCKSDB_LITE
-      wal_manager_(immutable_db_options_, file_options_, seq_per_batch),
+      wal_manager_(immutable_db_options_, file_options_, io_tracer_,
+                   seq_per_batch),
 #endif  // ROCKSDB_LITE
       event_logger_(immutable_db_options_.info_log.get()),
       bg_work_paused_(0),
       bg_compaction_paused_(0),
       refitting_level_(false),
       opened_successfully_(false),
+#ifndef ROCKSDB_LITE
+      stats_dump_scheduler_(nullptr),
+#endif  // ROCKSDB_LITE
       two_write_queues_(options.two_write_queues),
       manual_wal_flush_(options.manual_wal_flush),
       // last_sequencee_ is always maintained by the main queue that also writes
@@ -229,7 +235,8 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
   // !batch_per_trx_ implies seq_per_batch_ because it is only unset for
   // WriteUnprepared, which should use seq_per_batch_.
   assert(batch_per_txn_ || seq_per_batch_);
-  env_->GetAbsolutePath(dbname, &db_absolute_path_);
+  // TODO: Check for an error here
+  env_->GetAbsolutePath(dbname, &db_absolute_path_).PermitUncheckedError();
 
   // Reserve ten files or so for other uses and give the rest to TableCache.
   // Give a large number for setting of "infinite" open files.
@@ -244,12 +251,14 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
 
   versions_.reset(new VersionSet(dbname_, &immutable_db_options_, file_options_,
                                  table_cache_.get(), write_buffer_manager_,
-                                 &write_controller_, &block_cache_tracer_));
+                                 &write_controller_, &block_cache_tracer_,
+                                 io_tracer_));
   column_family_memtables_.reset(
       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());
@@ -293,7 +302,7 @@ Status DBImpl::Resume() {
 // 4. Schedule compactions if needed for all the CFs. This is needed as the
 //    flush in the prior step might have been a no-op for some CFs, which
 //    means a new super version wouldn't have been installed
-Status DBImpl::ResumeImpl() {
+Status DBImpl::ResumeImpl(DBRecoverContext context) {
   mutex_.AssertHeld();
   WaitForBackgroundWork();
 
@@ -312,8 +321,37 @@ 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
@@ -326,7 +364,7 @@ Status DBImpl::ResumeImpl() {
       autovector<ColumnFamilyData*> cfds;
       SelectColumnFamiliesForAtomicFlush(&cfds);
       mutex_.Unlock();
-      s = AtomicFlushMemTables(cfds, flush_opts, FlushReason::kErrorRecovery);
+      s = AtomicFlushMemTables(cfds, flush_opts, context.flush_reason);
       mutex_.Lock();
     } else {
       for (auto cfd : *versions_->GetColumnFamilySet()) {
@@ -335,7 +373,7 @@ Status DBImpl::ResumeImpl() {
         }
         cfd->Ref();
         mutex_.Unlock();
-        s = FlushMemTable(cfd, flush_opts, FlushReason::kErrorRecovery);
+        s = FlushMemTable(cfd, flush_opts, context.flush_reason);
         mutex_.Lock();
         cfd->UnrefAndTryDelete();
         if (!s.ok()) {
@@ -364,6 +402,13 @@ 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();
@@ -400,14 +445,12 @@ void DBImpl::CancelAllBackgroundWork(bool wait) {
   ROCKS_LOG_INFO(immutable_db_options_.info_log,
                  "Shutdown: canceling all background work");
 
-  if (thread_dump_stats_ != nullptr) {
-    thread_dump_stats_->cancel();
-    thread_dump_stats_.reset();
-  }
-  if (thread_persist_stats_ != nullptr) {
-    thread_persist_stats_->cancel();
-    thread_persist_stats_.reset();
+#ifndef ROCKSDB_LITE
+  if (stats_dump_scheduler_ != nullptr) {
+    stats_dump_scheduler_->Unregister(this);
   }
+#endif  // !ROCKSDB_LITE
+
   InstrumentedMutexLock l(&mutex_);
   if (!shutting_down_.load(std::memory_order_acquire) &&
       has_unpersisted_data_.load(std::memory_order_relaxed) &&
@@ -459,7 +502,7 @@ Status DBImpl::CloseHelper() {
       env_->UnSchedule(this, Env::Priority::BOTTOM);
   int compactions_unscheduled = env_->UnSchedule(this, Env::Priority::LOW);
   int flushes_unscheduled = env_->UnSchedule(this, Env::Priority::HIGH);
-  Status ret;
+  Status ret = Status::OK();
   mutex_.Lock();
   bg_bottom_compaction_scheduled_ -= bottom_compactions_unscheduled;
   bg_compaction_scheduled_ -= compactions_unscheduled;
@@ -571,7 +614,8 @@ Status DBImpl::CloseHelper() {
   versions_.reset();
   mutex_.Unlock();
   if (db_lock_ != nullptr) {
-    env_->UnlockFile(db_lock_);
+    // TODO: Check for unlock error
+    env_->UnlockFile(db_lock_).PermitUncheckedError();
   }
 
   ROCKS_LOG_INFO(immutable_db_options_.info_log, "Shutdown complete");
@@ -590,7 +634,7 @@ Status DBImpl::CloseHelper() {
 
   if (immutable_db_options_.info_log && own_info_log_) {
     Status s = immutable_db_options_.info_log->Close();
-    if (ret.ok()) {
+    if (!s.ok() && ret.ok()) {
       ret = s;
     }
   }
@@ -609,7 +653,7 @@ Status DBImpl::CloseImpl() { return CloseHelper(); }
 DBImpl::~DBImpl() {
   if (!closed_) {
     closed_ = true;
-    CloseHelper();
+    CloseHelper().PermitUncheckedError();
   }
 }
 
@@ -640,28 +684,19 @@ void DBImpl::PrintStatistics() {
   }
 }
 
-void DBImpl::StartTimedTasks() {
-  unsigned int stats_dump_period_sec = 0;
-  unsigned int stats_persist_period_sec = 0;
+void DBImpl::StartStatsDumpScheduler() {
+#ifndef ROCKSDB_LITE
   {
     InstrumentedMutexLock l(&mutex_);
-    stats_dump_period_sec = mutable_db_options_.stats_dump_period_sec;
-    if (stats_dump_period_sec > 0) {
-      if (!thread_dump_stats_) {
-        thread_dump_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
-            [this]() { DBImpl::DumpStats(); }, "dump_st", env_,
-            static_cast<uint64_t>(stats_dump_period_sec) * kMicrosInSecond));
-      }
-    }
-    stats_persist_period_sec = mutable_db_options_.stats_persist_period_sec;
-    if (stats_persist_period_sec > 0) {
-      if (!thread_persist_stats_) {
-        thread_persist_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
-            [this]() { DBImpl::PersistStats(); }, "pst_st", env_,
-            static_cast<uint64_t>(stats_persist_period_sec) * kMicrosInSecond));
-      }
-    }
+    stats_dump_scheduler_ = StatsDumpScheduler::Default();
+    TEST_SYNC_POINT_CALLBACK("DBImpl::StartStatsDumpScheduler:Init",
+                             &stats_dump_scheduler_);
   }
+
+  stats_dump_scheduler_->Register(this,
+                                  mutable_db_options_.stats_dump_period_sec,
+                                  mutable_db_options_.stats_persist_period_sec);
+#endif  // !ROCKSDB_LITE
 }
 
 // esitmate the total size of stats_history_
@@ -687,7 +722,9 @@ void DBImpl::PersistStats() {
   if (shutdown_initiated_) {
     return;
   }
+  TEST_SYNC_POINT("DBImpl::PersistStats:StartRunning");
   uint64_t now_seconds = env_->NowMicros() / kMicrosInSecond;
+
   Statistics* statistics = immutable_db_options_.statistics.get();
   if (!statistics) {
     return;
@@ -778,6 +815,7 @@ void DBImpl::PersistStats() {
                    " bytes, slice count: %" ROCKSDB_PRIszt,
                    stats_history_size, stats_history_.size());
   }
+  TEST_SYNC_POINT("DBImpl::PersistStats:End");
 #endif  // !ROCKSDB_LITE
 }
 
@@ -832,6 +870,7 @@ void DBImpl::DumpStats() {
   if (shutdown_initiated_) {
     return;
   }
+  TEST_SYNC_POINT("DBImpl::DumpStats:StartRunning");
   {
     InstrumentedMutexLock l(&mutex_);
     default_cf_internal_stats_->GetStringProperty(
@@ -909,7 +948,8 @@ 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",
@@ -1038,36 +1078,22 @@ Status DBImpl::SetDBOptions(
       }
 
       if (new_options.stats_dump_period_sec !=
-          mutable_db_options_.stats_dump_period_sec) {
-        if (thread_dump_stats_) {
+              mutable_db_options_.stats_dump_period_sec ||
+          new_options.stats_persist_period_sec !=
+              mutable_db_options_.stats_persist_period_sec) {
+        if (stats_dump_scheduler_) {
           mutex_.Unlock();
-          thread_dump_stats_->cancel();
+          stats_dump_scheduler_->Unregister(this);
           mutex_.Lock();
         }
-        if (new_options.stats_dump_period_sec > 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));
-        } else {
-          thread_dump_stats_.reset();
-        }
-      }
-      if (new_options.stats_persist_period_sec !=
-          mutable_db_options_.stats_persist_period_sec) {
-        if (thread_persist_stats_) {
+        if (new_options.stats_dump_period_sec > 0 ||
+            new_options.stats_persist_period_sec > 0) {
           mutex_.Unlock();
-          thread_persist_stats_->cancel();
+          stats_dump_scheduler_->Register(this,
+                                          new_options.stats_dump_period_sec,
+                                          new_options.stats_persist_period_sec);
           mutex_.Lock();
         }
-        if (new_options.stats_persist_period_sec > 0) {
-          thread_persist_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
-              [this]() { DBImpl::PersistStats(); }, "pst_st", env_,
-              static_cast<uint64_t>(new_options.stats_persist_period_sec) *
-                  kMicrosInSecond));
-        } else {
-          thread_persist_stats_.reset();
-        }
       }
       write_controller_.set_max_delayed_write_rate(
           new_options.delayed_write_rate);
@@ -1309,23 +1335,25 @@ bool DBImpl::SetPreserveDeletesSequenceNumber(SequenceNumber seqnum) {
   }
 }
 
-InternalIterator* DBImpl::NewInternalIterator(
-    Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
-    ColumnFamilyHandle* column_family, bool allow_unprepared_value) {
+InternalIterator* DBImpl::NewInternalIterator(const ReadOptions& read_options,
+                                              Arena* arena,
+                                              RangeDelAggregator* range_del_agg,
+                                              SequenceNumber sequence,
+                                              ColumnFamilyHandle* column_family,
+                                              bool allow_unprepared_value) {
   ColumnFamilyData* cfd;
   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();
   }
 
   mutex_.Lock();
   SuperVersion* super_version = cfd->GetSuperVersion()->Ref();
   mutex_.Unlock();
-  ReadOptions roptions;
-  return NewInternalIterator(roptions, cfd, super_version, arena, range_del_agg,
-                             sequence, allow_unprepared_value);
+  return NewInternalIterator(read_options, cfd, super_version, arena,
+                             range_del_agg, sequence, allow_unprepared_value);
 }
 
 void DBImpl::SchedulePurge() {
@@ -1523,19 +1551,32 @@ Status DBImpl::Get(const ReadOptions& read_options,
   return s;
 }
 
+namespace {
+class GetWithTimestampReadCallback : public ReadCallback {
+ public:
+  explicit GetWithTimestampReadCallback(SequenceNumber seq)
+      : ReadCallback(seq) {}
+  bool IsVisibleFullCheck(SequenceNumber seq) override {
+    return seq <= max_visible_seq_;
+  }
+};
+}  // namespace
+
 Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
                        GetImplOptions& get_impl_options) {
   assert(get_impl_options.value != nullptr ||
          get_impl_options.merge_operands != nullptr);
 
-#ifndef NDEBUG
   assert(get_impl_options.column_family);
-  ColumnFamilyHandle* cf = get_impl_options.column_family;
-  const Comparator* const ucmp = cf->GetComparator();
+  const Comparator* ucmp = get_impl_options.column_family->GetComparator();
   assert(ucmp);
-  if (ucmp->timestamp_size() > 0) {
+  size_t ts_sz = ucmp->timestamp_size();
+  GetWithTimestampReadCallback read_cb(0);  // Will call Refresh
+
+#ifndef NDEBUG
+  if (ts_sz > 0) {
     assert(read_options.timestamp);
-    assert(read_options.timestamp->size() == ucmp->timestamp_size());
+    assert(read_options.timestamp->size() == ts_sz);
   } else {
     assert(!read_options.timestamp);
   }
@@ -1545,8 +1586,8 @@ 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_) {
@@ -1601,6 +1642,12 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
       snapshot = get_impl_options.callback->max_visible_seq();
     }
   }
+  // If timestamp is used, we use read callback to ensure <key,t,s> is returned
+  // only if t <= read_opts.timestamp and s <= snapshot.
+  if (ts_sz > 0 && !get_impl_options.callback) {
+    read_cb.Refresh(snapshot);
+    get_impl_options.callback = &read_cb;
+  }
   TEST_SYNC_POINT("DBImpl::GetImpl:3");
   TEST_SYNC_POINT("DBImpl::GetImpl:4");
 
@@ -1618,9 +1665,6 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
   bool skip_memtable = (read_options.read_tier == kPersistedTier &&
                         has_unpersisted_data_.load(std::memory_order_relaxed));
   bool done = false;
-  const Comparator* comparator =
-      get_impl_options.column_family->GetComparator();
-  size_t ts_sz = comparator->timestamp_size();
   std::string* timestamp = ts_sz > 0 ? get_impl_options.timestamp : nullptr;
   if (!skip_memtable) {
     // Get value associated with key
@@ -1747,7 +1791,7 @@ 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(),
@@ -2149,10 +2193,11 @@ 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);
@@ -2516,7 +2561,7 @@ 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;
 }
@@ -2553,7 +2598,7 @@ 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");
@@ -2649,12 +2694,6 @@ 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"));
-  }
   Iterator* result = nullptr;
   if (read_options.read_tier == kPersistedTier) {
     return NewErrorIterator(Status::NotSupported(
@@ -2669,8 +2708,9 @@ 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
@@ -2691,10 +2731,11 @@ 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;
 }
@@ -2707,6 +2748,24 @@ 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:
   // +-------------------------------+
@@ -2755,10 +2814,10 @@ ArenaWrappedDBIter* DBImpl::NewIteratorImpl(const ReadOptions& read_options,
       sv->version_number, read_callback, this, cfd, allow_blob,
       read_options.snapshot != nullptr ? false : allow_refresh);
 
-  InternalIterator* internal_iter =
-      NewInternalIterator(read_options, cfd, sv, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), snapshot,
-                          /* allow_unprepared_value */ true);
+  InternalIterator* internal_iter = NewInternalIterator(
+      db_iter->GetReadOptions(), cfd, sv, db_iter->GetArena(),
+      db_iter->GetRangeDelAggregator(), snapshot,
+      /* allow_unprepared_value */ true);
   db_iter->SetIterUnderDBIter(internal_iter);
 
   return db_iter;
@@ -2784,7 +2843,7 @@ 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);
@@ -2804,7 +2863,8 @@ 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));
     }
@@ -2911,7 +2971,7 @@ 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
@@ -2933,7 +2993,7 @@ 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
@@ -2967,9 +3027,23 @@ FileSystem* DBImpl::GetFileSystem() const {
   return immutable_db_options_.fs.get();
 }
 
+#ifndef ROCKSDB_LITE
+
+Status DBImpl::StartIOTrace(Env* env, const TraceOptions& trace_options,
+                            std::unique_ptr<TraceWriter>&& trace_writer) {
+  return io_tracer_->StartIOTrace(env, trace_options, std::move(trace_writer));
+}
+
+Status DBImpl::EndIOTrace() {
+  io_tracer_->EndIOTrace();
+  return Status::OK();
+}
+
+#endif  // ROCKSDB_LITE
+
 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());
 }
@@ -2983,7 +3057,8 @@ 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) {
@@ -3017,7 +3092,8 @@ 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) {
@@ -3036,7 +3112,8 @@ 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);
 }
 
@@ -3209,7 +3286,7 @@ 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);
 
@@ -3234,7 +3311,7 @@ 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;
@@ -3392,7 +3469,7 @@ 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;
@@ -3482,10 +3559,16 @@ 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
@@ -3601,6 +3684,37 @@ 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*/,
@@ -3673,7 +3787,7 @@ Status DestroyDB(const std::string& dbname, const Options& options,
   // log file and prevents cleanup and directory removal
   soptions.info_log.reset();
   // Ignore error in case directory does not exist
-  env->GetChildren(dbname, &filenames);
+  env->GetChildren(dbname, &filenames).PermitUncheckedError();
 
   FileLock* lock;
   const std::string lockname = LockFileName(dbname);
@@ -3695,7 +3809,7 @@ Status DestroyDB(const std::string& dbname, const Options& options,
         } else {
           del = env->DeleteFile(path_to_delete);
         }
-        if (result.ok() && !del.ok()) {
+        if (!del.ok() && result.ok()) {
           result = del;
         }
       }
@@ -3718,12 +3832,13 @@ Status DestroyDB(const std::string& dbname, const Options& options,
             std::string table_path = path + "/" + fname;
             Status del = DeleteDBFile(&soptions, table_path, dbname,
                                       /*force_bg=*/false, /*force_fg=*/false);
-            if (result.ok() && !del.ok()) {
+            if (!del.ok() && result.ok()) {
               result = del;
             }
           }
         }
-        env->DeleteDir(path);
+        // TODO: Should we return an error if we cannot delete the directory?
+        env->DeleteDir(path).PermitUncheckedError();
       }
     }
 
@@ -3746,12 +3861,13 @@ Status DestroyDB(const std::string& dbname, const Options& options,
           Status del =
               DeleteDBFile(&soptions, archivedir + "/" + file, archivedir,
                            /*force_bg=*/false, /*force_fg=*/!wal_in_db_path);
-          if (result.ok() && !del.ok()) {
+          if (!del.ok() && result.ok()) {
             result = del;
           }
         }
       }
-      env->DeleteDir(archivedir);
+      // Ignore error in case dir contains other files
+      env->DeleteDir(archivedir).PermitUncheckedError();
     }
 
     // Delete log files in the WAL dir
@@ -3762,22 +3878,26 @@ Status DestroyDB(const std::string& dbname, const Options& options,
               DeleteDBFile(&soptions, LogFileName(soptions.wal_dir, number),
                            soptions.wal_dir, /*force_bg=*/false,
                            /*force_fg=*/!wal_in_db_path);
-          if (result.ok() && !del.ok()) {
+          if (!del.ok() && result.ok()) {
             result = del;
           }
         }
       }
-      env->DeleteDir(soptions.wal_dir);
+      // Ignore error in case dir contains other files
+      env->DeleteDir(soptions.wal_dir).PermitUncheckedError();
     }
 
-    env->UnlockFile(lock);  // Ignore error since state is already gone
-    env->DeleteFile(lockname);
+    // Ignore error since state is already gone
+    env->UnlockFile(lock).PermitUncheckedError();
+    env->DeleteFile(lockname).PermitUncheckedError();
 
     // sst_file_manager holds a ref to the logger. Make sure the logger is
     // gone before trying to remove the directory.
     soptions.sst_file_manager.reset();
 
-    env->DeleteDir(dbname);  // Ignore error in case dir contains other files
+    // Ignore error in case dir contains other files
+    env->DeleteDir(dbname).PermitUncheckedError();
+    ;
   }
   return result;
 }
@@ -3819,7 +3939,7 @@ Status DBImpl::WriteOptionsFile(bool need_mutex_lock,
   std::string file_name =
       TempOptionsFileName(GetName(), versions_->NewFileNumber());
   Status s = PersistRocksDBOptions(db_options, cf_names, cf_opts, file_name,
-                                   GetFileSystem());
+                                   fs_.get());
 
   if (s.ok()) {
     s = RenameTempFileToOptionsFile(file_name);
@@ -3912,7 +4032,8 @@ Status DBImpl::RenameTempFileToOptionsFile(const std::string& file_name) {
   }
 
   if (0 == disable_delete_obsolete_files_) {
-    DeleteObsoleteOptionsFiles();
+    // TODO: Should we check for errors here?
+    DeleteObsoleteOptionsFiles().PermitUncheckedError();
   }
   return s;
 #else
@@ -4166,7 +4287,7 @@ Status DBImpl::IngestExternalFiles(
     auto* cfd = static_cast<ColumnFamilyHandleImpl*>(arg.column_family)->cfd();
     ingestion_jobs.emplace_back(
         env_, versions_.get(), cfd, immutable_db_options_, file_options_,
-        &snapshots_, arg.options, &directories_, &event_logger_);
+        &snapshots_, arg.options, &directories_, &event_logger_, io_tracer_);
   }
   std::vector<std::pair<bool, Status>> exec_results;
   for (size_t i = 0; i != num_cfs; ++i) {
@@ -4180,7 +4301,8 @@ 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);
   }
@@ -4191,7 +4313,8 @@ 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);
   }
@@ -4367,6 +4490,14 @@ 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
@@ -4426,11 +4557,11 @@ 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_,
-                                   import_options, metadata.files);
+                                   import_options, metadata.files, io_tracer_);
 
   SuperVersionContext dummy_sv_ctx(/* create_superversion */ true);
   VersionEdit dummy_edit;

db/db_impl/db_impl.h

@@ -55,6 +55,7 @@
 #include "rocksdb/write_buffer_manager.h"
 #include "table/scoped_arena_iterator.h"
 #include "trace_replay/block_cache_tracer.h"
+#include "trace_replay/io_tracer.h"
 #include "trace_replay/trace_replay.h"
 #include "util/autovector.h"
 #include "util/hash.h"
@@ -69,6 +70,10 @@ class ArenaWrappedDBIter;
 class InMemoryStatsHistoryIterator;
 class MemTable;
 class PersistentStatsHistoryIterator;
+class StatsDumpScheduler;
+#ifndef NDEBUG
+class StatsDumpTestScheduler;
+#endif  // !NDEBUG
 class TableCache;
 class TaskLimiterToken;
 class Version;
@@ -350,12 +355,20 @@ 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 +376,6 @@ 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 +398,9 @@ 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.
@@ -437,6 +450,13 @@ class DBImpl : public DB {
   using DB::EndBlockCacheTrace;
   Status EndBlockCacheTrace() override;
 
+  using DB::StartIOTrace;
+  Status StartIOTrace(Env* env, const TraceOptions& options,
+                      std::unique_ptr<TraceWriter>&& trace_writer) override;
+
+  using DB::EndIOTrace;
+  Status EndIOTrace() override;
+
   using DB::GetPropertiesOfAllTables;
   virtual Status GetPropertiesOfAllTables(
       ColumnFamilyHandle* column_family,
@@ -477,6 +497,7 @@ 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,
@@ -585,8 +606,10 @@ class DBImpl : public DB {
   // the value and so will require PrepareValue() to be called before value();
   // allow_unprepared_value = false is convenient when this optimization is not
   // useful, e.g. when reading the whole column family.
+  // @param read_options Must outlive the returned iterator.
   InternalIterator* NewInternalIterator(
-      Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
+      const ReadOptions& read_options, Arena* arena,
+      RangeDelAggregator* range_del_agg, SequenceNumber sequence,
       ColumnFamilyHandle* column_family = nullptr,
       bool allow_unprepared_value = false);
 
@@ -712,10 +735,14 @@ class DBImpl : public DB {
 
   const WriteController& write_controller() { return write_controller_; }
 
-  InternalIterator* NewInternalIterator(
-      const ReadOptions&, ColumnFamilyData* cfd, SuperVersion* super_version,
-      Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
-      bool allow_unprepared_value);
+  // @param read_options Must outlive the returned iterator.
+  InternalIterator* NewInternalIterator(const ReadOptions& read_options,
+                                        ColumnFamilyData* cfd,
+                                        SuperVersion* super_version,
+                                        Arena* arena,
+                                        RangeDelAggregator* range_del_agg,
+                                        SequenceNumber sequence,
+                                        bool allow_unprepared_value);
 
   // hollow transactions shell used for recovery.
   // these will then be passed to TransactionDB so that
@@ -843,8 +870,8 @@ 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,
@@ -965,9 +992,7 @@ class DBImpl : public DB {
   int TEST_BGCompactionsAllowed() const;
   int TEST_BGFlushesAllowed() const;
   size_t TEST_GetWalPreallocateBlockSize(uint64_t write_buffer_size) const;
-  void TEST_WaitForDumpStatsRun(std::function<void()> callback) const;
-  void TEST_WaitForPersistStatsRun(std::function<void()> callback) const;
-  bool TEST_IsPersistentStatsEnabled() const;
+  void TEST_WaitForStatsDumpRun(std::function<void()> callback) const;
   size_t TEST_EstimateInMemoryStatsHistorySize() const;
 
   VersionSet* TEST_GetVersionSet() const { return versions_.get(); }
@@ -975,18 +1000,33 @@ class DBImpl : public DB {
   const std::unordered_set<uint64_t>& TEST_GetFilesGrabbedForPurge() const {
     return files_grabbed_for_purge_;
   }
+
+#ifndef ROCKSDB_LITE
+  StatsDumpTestScheduler* TEST_GetStatsDumpScheduler() const;
+#endif  // !ROCKSDB_LITE
+
 #endif  // NDEBUG
 
+  // persist stats to column family "_persistent_stats"
+  void PersistStats();
+
+  // dump rocksdb.stats to LOG
+  void DumpStats();
+
  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_;
   const DBOptions initial_db_options_;
   Env* const env_;
-  std::shared_ptr<FileSystem> fs_;
+  std::shared_ptr<IOTracer> io_tracer_;
   const ImmutableDBOptions immutable_db_options_;
+  FileSystemPtr fs_;
   MutableDBOptions mutable_db_options_;
   Statistics* stats_;
   std::unordered_map<std::string, RecoveredTransaction*>
@@ -1161,6 +1201,10 @@ 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 +1230,7 @@ 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;
@@ -1335,7 +1379,7 @@ class DBImpl : public DB {
   // Required: DB mutex held
   Status PersistentStatsProcessFormatVersion();
 
-  Status ResumeImpl();
+  Status ResumeImpl(DBRecoverContext context);
 
   void MaybeIgnoreError(Status* s) const;
 
@@ -1523,6 +1567,10 @@ class DBImpl : public DB {
                                 uint64_t* log_used,
                                 SequenceNumber* last_sequence, size_t seq_inc);
 
+  // Used by WriteImpl to update bg_error_ if paranoid check is enabled.
+  // Caller must hold mutex_.
+  void WriteStatusCheckOnLocked(const Status& status);
+
   // Used by WriteImpl to update bg_error_ if paranoid check is enabled.
   void WriteStatusCheck(const Status& status);
 
@@ -1604,18 +1652,12 @@ class DBImpl : public DB {
                               LogBuffer* log_buffer);
 
   // Schedule background tasks
-  void StartTimedTasks();
+  void StartStatsDumpScheduler();
 
   void PrintStatistics();
 
   size_t EstimateInMemoryStatsHistorySize() const;
 
-  // persist stats to column family "_persistent_stats"
-  void PersistStats();
-
-  // dump rocksdb.stats to LOG
-  void DumpStats();
-
   // Return the minimum empty level that could hold the total data in the
   // input level. Return the input level, if such level could not be found.
   int FindMinimumEmptyLevelFitting(ColumnFamilyData* cfd,
@@ -1694,8 +1736,8 @@ 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 +1821,8 @@ 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_;
 
@@ -1796,7 +1840,14 @@ class DBImpl : public DB {
   InstrumentedMutex log_write_mutex_;
 
   std::atomic<bool> shutting_down_;
-  std::atomic<bool> manual_compaction_paused_;
+
+  // If zero, manual compactions are allowed to proceed. If non-zero, manual
+  // compactions may still be running, but will quickly fail with
+  // `Status::Incomplete`. The value indicates how many threads have paused
+  // manual compactions. It is accessed in read mode outside the DB mutex in
+  // compaction code paths.
+  std::atomic<int> manual_compaction_paused_;
+
   // This condition variable is signaled on these conditions:
   // * whenever bg_compaction_scheduled_ goes down to 0
   // * if AnyManualCompaction, whenever a compaction finishes, even if it hasn't
@@ -2059,13 +2110,12 @@ class DBImpl : public DB {
   // Only to be set during initialization
   std::unique_ptr<PreReleaseCallback> recoverable_state_pre_release_callback_;
 
-  // handle for scheduling stats dumping at fixed intervals
-  // REQUIRES: mutex locked
-  std::unique_ptr<ROCKSDB_NAMESPACE::RepeatableThread> thread_dump_stats_;
-
-  // handle for scheduling stats snapshoting at fixed intervals
-  // REQUIRES: mutex locked
-  std::unique_ptr<ROCKSDB_NAMESPACE::RepeatableThread> thread_persist_stats_;
+#ifndef ROCKSDB_LITE
+  // Scheduler to run DumpStats() and PersistStats(). Currently, it always use
+  // a global instance from StatsDumpScheduler::Default(). Only in unittest, it
+  // can be overrided by StatsDumpTestSchduler.
+  StatsDumpScheduler* stats_dump_scheduler_;
+#endif
 
   // When set, we use a separate queue for writes that don't write to memtable.
   // In 2PC these are the writes at Prepare phase.

db/db_impl/db_impl_compaction_flush.cc

@@ -6,11 +6,10 @@
 // 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/builder.h"
+#include "db/db_impl/db_impl.h"
 #include "db/error_handler.h"
 #include "db/event_helpers.h"
 #include "file/sst_file_manager_impl.h"
@@ -126,7 +125,12 @@ IOStatus DBImpl::SyncClosedLogs(JobContext* job_context) {
     // "number < current_log_number".
     MarkLogsSynced(current_log_number - 1, true, io_s);
     if (!io_s.ok()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+      if (total_log_size_ > 0) {
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+      } else {
+        // If the WAL is empty, we use different error reason
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
+      }
       TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Failed");
       return io_s;
     }
@@ -154,7 +158,8 @@ 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,
+      io_tracer_, db_id_, db_session_id_);
   FileMetaData file_meta;
 
   TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:BeforePickMemtables");
@@ -167,7 +172,7 @@ Status DBImpl::FlushMemTableToOutputFile(
 #endif  // ROCKSDB_LITE
 
   Status s;
-  IOStatus io_s;
+  IOStatus io_s = IOStatus::OK();
   if (logfile_number_ > 0 &&
       versions_->GetColumnFamilySet()->NumberOfColumnFamilies() > 1) {
     // If there are more than one column families, we need to make sure that
@@ -178,6 +183,10 @@ Status DBImpl::FlushMemTableToOutputFile(
     // SyncClosedLogs() may unlock and re-lock the db_mutex.
     io_s = SyncClosedLogs(job_context);
     s = io_s;
+    if (!io_s.ok() && !io_s.IsShutdownInProgress() &&
+        !io_s.IsColumnFamilyDropped()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+    }
   } else {
     TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Skip");
   }
@@ -193,7 +202,9 @@ 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,
@@ -210,11 +221,27 @@ 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
+      if (!versions_->io_status().ok()) {
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite);
+      } else if (total_log_size_ > 0) {
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+      } else {
+        // If the WAL is empty, we use different error reason
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
+      }
     } else {
       Status new_bg_error = s;
       error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
     }
+  } else {
+    // If we got here, then we decided not to care about the i_os status (either
+    // from never needing it or ignoring the flush job status
+    io_s.PermitUncheckedError();
   }
   if (s.ok()) {
 #ifndef ROCKSDB_LITE
@@ -345,7 +372,7 @@ 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, io_tracer_, db_id_, db_session_id_));
     jobs.back()->PickMemTable();
   }
 
@@ -574,7 +601,19 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
   // 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);
+      // 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
+      if (!versions_->io_status().ok()) {
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kManifestWrite);
+      } else if (total_log_size_ > 0) {
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+      } else {
+        // If the WAL is empty, we use different error reason
+        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
+      }
     } else {
       Status new_bg_error = s;
       error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
@@ -678,7 +717,7 @@ 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()) {
@@ -827,13 +866,20 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
   }
 
   if (options.change_level) {
+    TEST_SYNC_POINT("DBImpl::CompactRange:BeforeRefit:1");
+    TEST_SYNC_POINT("DBImpl::CompactRange:BeforeRefit:2");
+
     ROCKS_LOG_INFO(immutable_db_options_.info_log,
                    "[RefitLevel] waiting for background threads to stop");
+    DisableManualCompaction();
     s = PauseBackgroundWork();
     if (s.ok()) {
+      TEST_SYNC_POINT("DBImpl::CompactRange:PreRefitLevel");
       s = ReFitLevel(cfd, final_output_level, options.target_level);
+      TEST_SYNC_POINT("DBImpl::CompactRange:PostRefitLevel");
+      ContinueBackgroundWork();
     }
-    ContinueBackgroundWork();
+    EnableManualCompaction();
   }
   LogFlush(immutable_db_options_.info_log);
 
@@ -868,7 +914,8 @@ 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;
@@ -939,7 +986,7 @@ Status DBImpl::CompactFilesImpl(
   if (shutting_down_.load(std::memory_order_acquire)) {
     return Status::ShutdownInProgress();
   }
-  if (manual_compaction_paused_.load(std::memory_order_acquire)) {
+  if (manual_compaction_paused_.load(std::memory_order_acquire) > 0) {
     return Status::Incomplete(Status::SubCode::kManualCompactionPaused);
   }
 
@@ -1001,7 +1048,7 @@ 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 +1078,8 @@ 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, io_tracer_,
+      &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 +1136,12 @@ 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) {
@@ -1154,7 +1207,7 @@ void DBImpl::NotifyOnCompactionBegin(ColumnFamilyData* cfd, Compaction* c,
     return;
   }
   if (c->is_manual_compaction() &&
-      manual_compaction_paused_.load(std::memory_order_acquire)) {
+      manual_compaction_paused_.load(std::memory_order_acquire) > 0) {
     return;
   }
   Version* current = cfd->current();
@@ -1164,43 +1217,7 @@ void DBImpl::NotifyOnCompactionBegin(ColumnFamilyData* cfd, Compaction* c,
   TEST_SYNC_POINT("DBImpl::NotifyOnCompactionBegin::UnlockMutex");
   {
     CompactionJobInfo info{};
-    info.cf_name = cfd->GetName();
-    info.status = st;
-    info.thread_id = env_->GetThreadID();
-    info.job_id = job_id;
-    info.base_input_level = c->start_level();
-    info.output_level = c->output_level();
-    info.stats = job_stats;
-    info.table_properties = c->GetOutputTableProperties();
-    info.compaction_reason = c->compaction_reason();
-    info.compression = c->output_compression();
-    for (size_t i = 0; i < c->num_input_levels(); ++i) {
-      for (const auto fmd : *c->inputs(i)) {
-        const FileDescriptor& desc = fmd->fd;
-        const uint64_t file_number = desc.GetNumber();
-        auto fn = TableFileName(c->immutable_cf_options()->cf_paths,
-                                file_number, desc.GetPathId());
-        info.input_files.push_back(fn);
-        info.input_file_infos.push_back(CompactionFileInfo{
-            static_cast<int>(i), file_number, fmd->oldest_blob_file_number});
-        if (info.table_properties.count(fn) == 0) {
-          std::shared_ptr<const TableProperties> tp;
-          auto s = current->GetTableProperties(&tp, fmd, &fn);
-          if (s.ok()) {
-            info.table_properties[fn] = tp;
-          }
-        }
-      }
-    }
-    for (const auto& newf : c->edit()->GetNewFiles()) {
-      const FileMetaData& meta = newf.second;
-      const FileDescriptor& desc = meta.fd;
-      const uint64_t file_number = desc.GetNumber();
-      info.output_files.push_back(TableFileName(
-          c->immutable_cf_options()->cf_paths, file_number, desc.GetPathId()));
-      info.output_file_infos.push_back(CompactionFileInfo{
-          newf.first, file_number, meta.oldest_blob_file_number});
-    }
+    BuildCompactionJobInfo(cfd, c, st, job_stats, job_id, current, &info);
     for (auto listener : immutable_db_options_.listeners) {
       listener->OnCompactionBegin(this, info);
     }
@@ -1228,7 +1245,7 @@ void DBImpl::NotifyOnCompactionCompleted(
     return;
   }
   if (c->is_manual_compaction() &&
-      manual_compaction_paused_.load(std::memory_order_acquire)) {
+      manual_compaction_paused_.load(std::memory_order_acquire) > 0) {
     return;
   }
   Version* current = cfd->current();
@@ -1287,20 +1304,29 @@ Status DBImpl::ReFitLevel(ColumnFamilyData* cfd, int level, int target_level) {
   }
 
   auto* vstorage = cfd->current()->storage_info();
-  if (to_level > level) {
-    if (level == 0) {
-      return Status::NotSupported(
-          "Cannot change from level 0 to other levels.");
-    }
-    // Check levels are empty for a trivial move
-    for (int l = level + 1; l <= to_level; l++) {
-      if (vstorage->NumLevelFiles(l) > 0) {
+  if (to_level != level) {
+    if (to_level > level) {
+      if (level == 0) {
         return Status::NotSupported(
-            "Levels between source and target are not empty for a move.");
+            "Cannot change from level 0 to other levels.");
+      }
+      // Check levels are empty for a trivial move
+      for (int l = level + 1; l <= to_level; l++) {
+        if (vstorage->NumLevelFiles(l) > 0) {
+          return Status::NotSupported(
+              "Levels between source and target are not empty for a move.");
+        }
+      }
+    } else {
+      // to_level < level
+      // Check levels are empty for a trivial move
+      for (int l = to_level; l < level; l++) {
+        if (vstorage->NumLevelFiles(l) > 0) {
+          return Status::NotSupported(
+              "Levels between source and target are not empty for a move.");
+        }
       }
     }
-  }
-  if (to_level != level) {
     ROCKS_LOG_DEBUG(immutable_db_options_.info_log,
                     "[%s] Before refitting:\n%s", cfd->GetName().c_str(),
                     cfd->current()->DebugString().data());
@@ -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
@@ -1589,6 +1615,7 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
       return s;
     }
   }
+
   FlushRequest flush_req;
   {
     WriteContext context;
@@ -1605,7 +1632,11 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
     WaitForPendingWrites();
 
     if (!cfd->mem()->IsEmpty() || !cached_recoverable_state_empty_.load()) {
-      s = SwitchMemtable(cfd, &context);
+      if (flush_reason != FlushReason::kErrorRecoveryRetryFlush) {
+        s = SwitchMemtable(cfd, &context);
+      } else {
+        assert(cfd->imm()->NumNotFlushed() > 0);
+      }
     }
     if (s.ok()) {
       if (cfd->imm()->NumNotFlushed() != 0 || !cfd->mem()->IsEmpty() ||
@@ -1613,7 +1644,8 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
         flush_memtable_id = cfd->imm()->GetLatestMemTableID();
         flush_req.emplace_back(cfd, flush_memtable_id);
       }
-      if (immutable_db_options_.persist_stats_to_disk) {
+      if (immutable_db_options_.persist_stats_to_disk &&
+          flush_reason != FlushReason::kErrorRecoveryRetryFlush) {
         ColumnFamilyData* cfd_stats =
             versions_->GetColumnFamilySet()->GetColumnFamily(
                 kPersistentStatsColumnFamilyName);
@@ -1642,7 +1674,6 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
         }
       }
     }
-
     if (s.ok() && !flush_req.empty()) {
       for (auto& elem : flush_req) {
         ColumnFamilyData* loop_cfd = elem.first;
@@ -1678,8 +1709,10 @@ Status DBImpl::FlushMemTable(ColumnFamilyData* cfd,
       cfds.push_back(iter.first);
       flush_memtable_ids.push_back(&(iter.second));
     }
-    s = WaitForFlushMemTables(cfds, flush_memtable_ids,
-                              (flush_reason == FlushReason::kErrorRecovery));
+    s = WaitForFlushMemTables(
+        cfds, flush_memtable_ids,
+        (flush_reason == FlushReason::kErrorRecovery ||
+         flush_reason == FlushReason::kErrorRecoveryRetryFlush));
     InstrumentedMutexLock lock_guard(&mutex_);
     for (auto* tmp_cfd : cfds) {
       tmp_cfd->UnrefAndTryDelete();
@@ -1737,7 +1770,8 @@ Status DBImpl::AtomicFlushMemTables(
       }
     }
     for (auto cfd : cfds) {
-      if (cfd->mem()->IsEmpty() && cached_recoverable_state_empty_.load()) {
+      if ((cfd->mem()->IsEmpty() && cached_recoverable_state_empty_.load()) ||
+          flush_reason == FlushReason::kErrorRecoveryRetryFlush) {
         continue;
       }
       cfd->Ref();
@@ -1780,8 +1814,10 @@ Status DBImpl::AtomicFlushMemTables(
     for (auto& iter : flush_req) {
       flush_memtable_ids.push_back(&(iter.second));
     }
-    s = WaitForFlushMemTables(cfds, flush_memtable_ids,
-                              (flush_reason == FlushReason::kErrorRecovery));
+    s = WaitForFlushMemTables(
+        cfds, flush_memtable_ids,
+        (flush_reason == FlushReason::kErrorRecovery ||
+         flush_reason == FlushReason::kErrorRecoveryRetryFlush));
     InstrumentedMutexLock lock_guard(&mutex_);
     for (auto* cfd : cfds) {
       cfd->UnrefAndTryDelete();
@@ -1891,6 +1927,13 @@ Status DBImpl::WaitForFlushMemTables(
     if (!error_handler_.GetRecoveryError().ok()) {
       break;
     }
+    // If BGWorkStopped, which indicate that there is a BG error and
+    // 1) soft error but requires no BG work, 2) no in auto_recovery_
+    if (!resuming_from_bg_err && error_handler_.IsBGWorkStopped() &&
+        error_handler_.GetBGError().severity() < Status::Severity::kHardError) {
+      return error_handler_.GetBGError();
+    }
+
     // Number of column families that have been dropped.
     int num_dropped = 0;
     // Number of column families that have finished flush.
@@ -1939,11 +1982,21 @@ Status DBImpl::EnableAutoCompaction(
 }
 
 void DBImpl::DisableManualCompaction() {
-  manual_compaction_paused_.store(true, std::memory_order_release);
+  InstrumentedMutexLock l(&mutex_);
+  manual_compaction_paused_.fetch_add(1, std::memory_order_release);
+  // Wait for any pending manual compactions to finish (typically through
+  // failing with `Status::Incomplete`) prior to returning. This way we are
+  // guaranteed no pending manual compaction will commit while manual
+  // compactions are "disabled".
+  while (HasPendingManualCompaction()) {
+    bg_cv_.Wait();
+  }
 }
 
 void DBImpl::EnableManualCompaction() {
-  manual_compaction_paused_.store(false, std::memory_order_release);
+  InstrumentedMutexLock l(&mutex_);
+  assert(manual_compaction_paused_ > 0);
+  manual_compaction_paused_.fetch_sub(1, std::memory_order_release);
 }
 
 void DBImpl::MaybeScheduleFlushOrCompaction() {
@@ -2502,7 +2555,7 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
     if (shutting_down_.load(std::memory_order_acquire)) {
       status = Status::ShutdownInProgress();
     } else if (is_manual &&
-               manual_compaction_paused_.load(std::memory_order_acquire)) {
+               manual_compaction_paused_.load(std::memory_order_acquire) > 0) {
       status = Status::Incomplete(Status::SubCode::kManualCompactionPaused);
     }
   } else {
@@ -2544,12 +2597,13 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
     if (!c) {
       m->done = true;
       m->manual_end = nullptr;
-      ROCKS_LOG_BUFFER(log_buffer,
-                       "[%s] Manual compaction from level-%d from %s .. "
-                       "%s; nothing to do\n",
-                       m->cfd->GetName().c_str(), m->input_level,
-                       (m->begin ? m->begin->DebugString(true).c_str() : "(begin)"),
-                       (m->end ? m->end->DebugString(true).c_str() : "(end)"));
+      ROCKS_LOG_BUFFER(
+          log_buffer,
+          "[%s] Manual compaction from level-%d from %s .. "
+          "%s; nothing to do\n",
+          m->cfd->GetName().c_str(), m->input_level,
+          (m->begin ? m->begin->DebugString(true).c_str() : "(begin)"),
+          (m->end ? m->end->DebugString(true).c_str() : "(end)"));
     } else {
       // First check if we have enough room to do the compaction
       bool enough_room = EnoughRoomForCompaction(
@@ -2616,7 +2670,8 @@ 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 +2742,6 @@ 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 +2799,6 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
       }
     }
 
-    versions_->SetIOStatusOK();
     status = versions_->LogAndApply(c->column_family_data(),
                                     *c->mutable_cf_options(), c->edit(),
                                     &mutex_, directories_.GetDbDir());
@@ -2821,8 +2874,9 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
         earliest_write_conflict_snapshot, 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, thread_pri,
-        is_manual ? &manual_compaction_paused_ : nullptr);
+        &compaction_job_stats, thread_pri, io_tracer_,
+        is_manual ? &manual_compaction_paused_ : nullptr, db_id_,
+        db_session_id_);
     compaction_job.Prepare();
 
     NotifyOnCompactionBegin(c->column_family_data(), c.get(), status,
@@ -2877,7 +2931,15 @@ 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);
     }

db/db_impl/db_impl_debug.cc

@@ -12,6 +12,7 @@
 #include "db/column_family.h"
 #include "db/db_impl/db_impl.h"
 #include "db/error_handler.h"
+#include "monitoring/stats_dump_scheduler.h"
 #include "monitoring/thread_status_updater.h"
 #include "util/cast_util.h"
 
@@ -47,7 +48,7 @@ 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 +58,7 @@ 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 +89,7 @@ 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 +132,7 @@ 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 +153,7 @@ 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 +243,7 @@ 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();
 }
@@ -271,21 +272,18 @@ size_t DBImpl::TEST_GetWalPreallocateBlockSize(
   return GetWalPreallocateBlockSize(write_buffer_size);
 }
 
-void DBImpl::TEST_WaitForDumpStatsRun(std::function<void()> callback) const {
-  if (thread_dump_stats_ != nullptr) {
-    thread_dump_stats_->TEST_WaitForRun(callback);
+#ifndef ROCKSDB_LITE
+void DBImpl::TEST_WaitForStatsDumpRun(std::function<void()> callback) const {
+  if (stats_dump_scheduler_ != nullptr) {
+    static_cast<StatsDumpTestScheduler*>(stats_dump_scheduler_)
+        ->TEST_WaitForRun(callback);
   }
 }
 
-void DBImpl::TEST_WaitForPersistStatsRun(std::function<void()> callback) const {
-  if (thread_persist_stats_ != nullptr) {
-    thread_persist_stats_->TEST_WaitForRun(callback);
-  }
-}
-
-bool DBImpl::TEST_IsPersistentStatsEnabled() const {
-  return thread_persist_stats_ && thread_persist_stats_->IsRunning();
+StatsDumpTestScheduler* DBImpl::TEST_GetStatsDumpScheduler() const {
+  return static_cast<StatsDumpTestScheduler*>(stats_dump_scheduler_);
 }
+#endif  // !ROCKSDB_LITE
 
 size_t DBImpl::TEST_EstimateInMemoryStatsHistorySize() const {
   return EstimateInMemoryStatsHistorySize();

db/db_impl/db_impl_experimental.cc

@@ -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) {

db/db_impl/db_impl_files.cc

@@ -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;
@@ -391,7 +447,8 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
   // Close WALs before trying to delete them.
   for (const auto w : state.logs_to_free) {
     // TODO: maybe check the return value of Close.
-    w->Close();
+    auto s = w->Close();
+    s.PermitUncheckedError();
   }
 
   bool own_files = OwnTablesAndLogs();
@@ -510,7 +567,6 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
     if (!own_files) {
       continue;
     }
-    Status file_deletion_status;
     if (schedule_only) {
       InstrumentedMutexLock guard_lock(&mutex_);
       SchedulePendingPurge(fname, dir_to_sync, type, number, state.job_id);

db/db_impl/db_impl_open.cc

@@ -6,17 +6,17 @@
 // Copyright (c) 2011 The LevelDB Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
-#include "db/db_impl/db_impl.h"
-
 #include <cinttypes>
 
 #include "db/builder.h"
+#include "db/db_impl/db_impl.h"
 #include "db/error_handler.h"
 #include "env/composite_env_wrapper.h"
 #include "file/read_write_util.h"
 #include "file/sst_file_manager_impl.h"
 #include "file/writable_file_writer.h"
 #include "monitoring/persistent_stats_history.h"
+#include "monitoring/stats_dump_scheduler.h"
 #include "options/options_helper.h"
 #include "rocksdb/wal_filter.h"
 #include "table/block_based/block_based_table_factory.h"
@@ -90,12 +90,22 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
   }
 
   if (result.recycle_log_file_num &&
-      (result.wal_recovery_mode == WALRecoveryMode::kPointInTimeRecovery ||
+      (result.wal_recovery_mode ==
+           WALRecoveryMode::kTolerateCorruptedTailRecords ||
+       result.wal_recovery_mode == WALRecoveryMode::kPointInTimeRecovery ||
        result.wal_recovery_mode == WALRecoveryMode::kAbsoluteConsistency)) {
-    // kPointInTimeRecovery is inconsistent with recycle log file feature since
-    // we define the "end" of the log as the first corrupt record we encounter.
-    // kAbsoluteConsistency doesn't make sense because even a clean
-    // shutdown leaves old junk at the end of the log file.
+    // - kTolerateCorruptedTailRecords is inconsistent with recycle log file
+    //   feature. WAL recycling expects recovery success upon encountering a
+    //   corrupt record at the point where new data ends and recycled data
+    //   remains at the tail. However, `kTolerateCorruptedTailRecords` must fail
+    //   upon encountering any such corrupt record, as it cannot differentiate
+    //   between this and a real corruption, which would cause committed updates
+    //   to be truncated -- a violation of the recovery guarantee.
+    // - kPointInTimeRecovery and kAbsoluteConsistency are incompatible with
+    //   recycle log file feature temporarily due to a bug found introducing a
+    //   hole in the recovered data
+    //   (https://github.com/facebook/rocksdb/pull/7252#issuecomment-673766236).
+    //   Besides this bug, we believe the features are fundamentally compatible.
     result.recycle_log_file_num = 0;
   }
 
@@ -137,13 +147,13 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
     // DeleteScheduler::CleanupDirectory on the same dir later, it will be
     // safe
     std::vector<std::string> filenames;
-    result.env->GetChildren(result.wal_dir, &filenames);
+    result.env->GetChildren(result.wal_dir, &filenames).PermitUncheckedError();
     for (std::string& filename : filenames) {
       if (filename.find(".log.trash", filename.length() -
                                           std::string(".log.trash").length()) !=
           std::string::npos) {
         std::string trash_file = result.wal_dir + "/" + filename;
-        result.env->DeleteFile(trash_file);
+        result.env->DeleteFile(trash_file).PermitUncheckedError();
       }
     }
   }
@@ -253,7 +263,7 @@ 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()) {
@@ -280,8 +290,8 @@ Status DBImpl::NewDB() {
     file->SetPreallocationBlockSize(
         immutable_db_options_.manifest_preallocation_size);
     std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-        std::move(file), manifest, file_options, env_, nullptr /* stats */,
-        immutable_db_options_.listeners));
+        std::move(file), manifest, file_options, env_, io_tracer_,
+        nullptr /* stats */, immutable_db_options_.listeners));
     log::Writer log(std::move(file_writer), 0, false);
     std::string record;
     new_db.EncodeTo(&record);
@@ -293,6 +303,10 @@ 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 +370,7 @@ 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 +385,35 @@ 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 +438,14 @@ 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 +455,16 @@ 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;
@@ -422,8 +472,9 @@ 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(
@@ -483,6 +534,7 @@ 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.
@@ -509,9 +561,8 @@ 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",
@@ -521,15 +572,15 @@ 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);
         }
@@ -581,15 +632,24 @@ 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);
         }
@@ -597,7 +657,6 @@ Status DBImpl::Recover(
       versions_->options_file_number_ = options_file_number;
     }
   }
-
   return s;
 }
 
@@ -704,10 +763,10 @@ 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;
       }
     }
   };
@@ -802,7 +861,8 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
         }
       }
       file_reader.reset(new SequentialFileReader(
-          std::move(file), fname, immutable_db_options_.log_readahead_size));
+          std::move(file), fname, immutable_db_options_.log_readahead_size,
+          io_tracer_));
     }
 
     // Create the log reader.
@@ -830,6 +890,8 @@ 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) &&
@@ -839,7 +901,11 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
                             Status::Corruption("log record too small"));
         continue;
       }
-      WriteBatchInternal::SetContents(&batch, record);
+
+      status = WriteBatchInternal::SetContents(&batch, record);
+      if (!status.ok()) {
+        return status;
+      }
       SequenceNumber sequence = WriteBatchInternal::Sequence(&batch);
 
       if (immutable_db_options_.wal_recovery_mode ==
@@ -994,6 +1060,16 @@ 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;
@@ -1220,7 +1296,8 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
         cfd->GetLatestMutableCFOptions()->paranoid_file_checks;
 
     int64_t _current_time = 0;
-    env_->GetCurrentTime(&_current_time);  // ignore error
+    env_->GetCurrentTime(&_current_time)
+        .PermitUncheckedError();  // ignore error
     const uint64_t current_time = static_cast<uint64_t>(_current_time);
     meta.oldest_ancester_time = current_time;
 
@@ -1253,8 +1330,10 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
           mutable_cf_options.sample_for_compression,
           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);
+          io_tracer_, &event_logger_, job_id, Env::IO_HIGH,
+          nullptr /* table_properties */, -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]"
@@ -1326,9 +1405,10 @@ 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 =
@@ -1346,25 +1426,25 @@ 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);
 
     const auto& listeners = immutable_db_options_.listeners;
-    std::unique_ptr<WritableFileWriter> file_writer(
-        new WritableFileWriter(std::move(lfile), log_fname, opt_file_options,
-                               env_, nullptr /* stats */, listeners));
+    std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
+        std::move(lfile), log_fname, opt_file_options, env_, io_tracer_,
+        nullptr /* stats */, listeners));
     *new_log = new log::Writer(std::move(file_writer), log_file_num,
                                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,
@@ -1465,7 +1545,7 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
               break;
             }
           } else {
-            s = Status::InvalidArgument("Column family not found: ", cf.name);
+            s = Status::InvalidArgument("Column family not found", cf.name);
             break;
           }
         }
@@ -1569,6 +1649,12 @@ 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.
 
@@ -1604,20 +1690,24 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
     paths.erase(std::unique(paths.begin(), paths.end()), paths.end());
     for (auto& path : paths) {
       std::vector<std::string> existing_files;
-      impl->immutable_db_options_.env->GetChildren(path, &existing_files);
+      // TODO: Check for errors here?
+      impl->immutable_db_options_.env->GetChildren(path, &existing_files)
+          .PermitUncheckedError();
       for (auto& file_name : existing_files) {
         uint64_t file_number;
         FileType file_type;
         std::string file_path = path + "/" + file_name;
         if (ParseFileName(file_name, &file_number, &file_type) &&
             file_type == kTableFile) {
+          // TODO: Check for errors from OnAddFile?
           if (known_file_sizes.count(file_name)) {
             // We're assuming that each sst file name exists in at most one of
             // the paths.
             sfm->OnAddFile(file_path, known_file_sizes.at(file_name),
-                           /* compaction */ false);
+                           /* compaction */ false)
+                .PermitUncheckedError();
           } else {
-            sfm->OnAddFile(file_path);
+            sfm->OnAddFile(file_path).PermitUncheckedError();
           }
         }
       }
@@ -1631,6 +1721,7 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
     sfm->ReserveDiskBuffer(max_write_buffer_size,
                            impl->immutable_db_options_.db_paths[0].path);
   }
+
 #endif  // !ROCKSDB_LITE
 
   if (s.ok()) {
@@ -1647,9 +1738,8 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
     }
   }
   if (s.ok()) {
-    impl->StartTimedTasks();
-  }
-  if (!s.ok()) {
+    impl->StartStatsDumpScheduler();
+  } else {
     for (auto* h : *handles) {
       delete h;
     }

db/db_impl/db_impl_readonly.cc

@@ -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();
@@ -85,10 +86,10 @@ Iterator* DBImplReadOnly::NewIterator(const ReadOptions& read_options,
       read_seq,
       super_version->mutable_cf_options.max_sequential_skip_in_iterations,
       super_version->version_number, read_callback);
-  auto internal_iter =
-      NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), read_seq,
-                          /* allow_unprepared_value */ true);
+  auto internal_iter = NewInternalIterator(
+      db_iter->GetReadOptions(), cfd, super_version, db_iter->GetArena(),
+      db_iter->GetRangeDelAggregator(), read_seq,
+      /* allow_unprepared_value */ true);
   db_iter->SetIterUnderDBIter(internal_iter);
   return db_iter;
 }
@@ -111,16 +112,16 @@ 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,
         sv->mutable_cf_options.max_sequential_skip_in_iterations,
         sv->version_number, read_callback);
-    auto* internal_iter =
-        NewInternalIterator(read_options, cfd, sv, db_iter->GetArena(),
-                            db_iter->GetRangeDelAggregator(), read_seq,
-                            /* allow_unprepared_value */ true);
+    auto* internal_iter = NewInternalIterator(
+        db_iter->GetReadOptions(), cfd, sv, db_iter->GetArena(),
+        db_iter->GetRangeDelAggregator(), read_seq,
+        /* allow_unprepared_value */ true);
     db_iter->SetIterUnderDBIter(internal_iter);
     iterators->push_back(db_iter);
   }
@@ -218,7 +219,7 @@ Status DBImplReadOnly::OpenForReadOnlyWithoutCheck(
       auto cfd =
           impl->versions_->GetColumnFamilySet()->GetColumnFamily(cf.name);
       if (cfd == nullptr) {
-        s = Status::InvalidArgument("Column family not found: ", cf.name);
+        s = Status::InvalidArgument("Column family not found", cf.name);
         break;
       }
       handles->push_back(new ColumnFamilyHandleImpl(cfd, impl, &impl->mutex_));
@@ -236,7 +237,7 @@ Status DBImplReadOnly::OpenForReadOnlyWithoutCheck(
     *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) {

db/db_impl/db_impl_secondary.cc

@@ -153,7 +153,8 @@ Status DBImplSecondary::MaybeInitLogReader(
         return status;
       }
       file_reader.reset(new SequentialFileReader(
-          std::move(file), fname, immutable_db_options_.log_readahead_size));
+          std::move(file), fname, immutable_db_options_.log_readahead_size,
+          io_tracer_));
     }
 
     // Create the log reader.
@@ -388,7 +389,7 @@ 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) {
@@ -415,10 +416,10 @@ ArenaWrappedDBIter* DBImplSecondary::NewIteratorImpl(
       snapshot,
       super_version->mutable_cf_options.max_sequential_skip_in_iterations,
       super_version->version_number, read_callback);
-  auto internal_iter =
-      NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), snapshot,
-                          /* allow_unprepared_value */ true);
+  auto internal_iter = NewInternalIterator(
+      db_iter->GetReadOptions(), cfd, super_version, db_iter->GetArena(),
+      db_iter->GetRangeDelAggregator(), snapshot,
+      /* allow_unprepared_value */ true);
   db_iter->SetIterUnderDBIter(internal_iter);
   return db_iter;
 }
@@ -611,7 +612,7 @@ Status DB::OpenAsSecondary(
   impl->versions_.reset(new ReactiveVersionSet(
       dbname, &impl->immutable_db_options_, impl->file_options_,
       impl->table_cache_.get(), impl->write_buffer_manager_,
-      &impl->write_controller_));
+      &impl->write_controller_, impl->io_tracer_));
   impl->column_family_memtables_.reset(
       new ColumnFamilyMemTablesImpl(impl->versions_->GetColumnFamilySet()));
   impl->wal_in_db_path_ = IsWalDirSameAsDBPath(&impl->immutable_db_options_);
@@ -623,7 +624,7 @@ Status DB::OpenAsSecondary(
       auto cfd =
           impl->versions_->GetColumnFamilySet()->GetColumnFamily(cf.name);
       if (nullptr == cfd) {
-        s = Status::InvalidArgument("Column family not found: ", cf.name);
+        s = Status::InvalidArgument("Column family not found", cf.name);
         break;
       }
       handles->push_back(new ColumnFamilyHandleImpl(cfd, impl, &impl->mutex_));
@@ -642,7 +643,7 @@ 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) {

db/db_impl/db_impl_write.cc

@@ -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 {
@@ -469,6 +470,7 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
   WriteThread::Writer w(write_options, my_batch, callback, log_ref,
                         disable_memtable);
   write_thread_.JoinBatchGroup(&w);
+  TEST_SYNC_POINT("DBImplWrite::PipelinedWriteImpl:AfterJoinBatchGroup");
   if (w.state == WriteThread::STATE_GROUP_LEADER) {
     WriteThread::WriteGroup wal_write_group;
     if (w.callback && !w.callback->AllowWriteBatching()) {
@@ -622,8 +624,6 @@ Status DBImpl::UnorderedWriteMemtable(const WriteOptions& write_options,
         0 /*log_number*/, this, true /*concurrent_memtable_writes*/,
         seq_per_batch_, sub_batch_cnt, true /*batch_per_txn*/,
         write_options.memtable_insert_hint_per_batch);
-
-    WriteStatusCheck(w.status);
     if (write_options.disableWAL) {
       has_unpersisted_data_.store(true, std::memory_order_relaxed);
     }
@@ -638,6 +638,7 @@ Status DBImpl::UnorderedWriteMemtable(const WriteOptions& write_options,
     std::lock_guard<std::mutex> lck(switch_mutex_);
     switch_cv_.notify_all();
   }
+  WriteStatusCheck(w.status);
 
   if (!w.FinalStatus().ok()) {
     return w.FinalStatus();
@@ -688,7 +689,7 @@ Status DBImpl::WriteImplWALOnly(
       InstrumentedMutexLock l(&mutex_);
       bool need_log_sync = false;
       status = PreprocessWrite(write_options, &need_log_sync, &write_context);
-      WriteStatusCheck(status);
+      WriteStatusCheckOnLocked(status);
     }
     if (!status.ok()) {
       WriteThread::WriteGroup write_group;
@@ -829,9 +830,22 @@ Status DBImpl::WriteImplWALOnly(
   return status;
 }
 
+void DBImpl::WriteStatusCheckOnLocked(const Status& status) {
+  // Is setting bg_error_ enough here?  This will at least stop
+  // compaction and fail any further writes.
+  // Caller must hold mutex_.
+  assert(!status.IsIOFenced() || !error_handler_.GetBGError().ok());
+  mutex_.AssertHeld();
+  if (immutable_db_options_.paranoid_checks && !status.ok() &&
+      !status.IsBusy() && !status.IsIncomplete()) {
+    error_handler_.SetBGError(status, BackgroundErrorReason::kWriteCallback);
+  }
+}
+
 void DBImpl::WriteStatusCheck(const Status& status) {
   // Is setting bg_error_ enough here?  This will at least stop
   // compaction and fail any further writes.
+  assert(!status.IsIOFenced() || !error_handler_.GetBGError().ok());
   if (immutable_db_options_.paranoid_checks && !status.ok() &&
       !status.IsBusy() && !status.IsIncomplete()) {
     mutex_.Lock();
@@ -843,8 +857,9 @@ void DBImpl::WriteStatusCheck(const Status& status) {
 void DBImpl::IOStatusCheck(const IOStatus& io_status) {
   // Is setting bg_error_ enough here?  This will at least stop
   // compaction and fail any further writes.
-  if (immutable_db_options_.paranoid_checks && !io_status.ok() &&
-      !io_status.IsBusy() && !io_status.IsIncomplete()) {
+  if ((immutable_db_options_.paranoid_checks && !io_status.ok() &&
+       !io_status.IsBusy() && !io_status.IsIncomplete()) ||
+      io_status.IsIOFenced()) {
     mutex_.Lock();
     error_handler_.SetBGError(io_status, BackgroundErrorReason::kWriteCallback);
     mutex_.Unlock();
@@ -1545,11 +1560,14 @@ Status DBImpl::TrimMemtableHistory(WriteContext* context) {
   }
   for (auto& cfd : cfds) {
     autovector<MemTable*> to_delete;
-    cfd->imm()->TrimHistory(&to_delete, cfd->mem()->ApproximateMemoryUsage());
+    bool trimmed = cfd->imm()->TrimHistory(
+        &to_delete, cfd->mem()->ApproximateMemoryUsage());
     if (!to_delete.empty()) {
       for (auto m : to_delete) {
         delete m;
       }
+    }
+    if (trimmed) {
       context->superversion_context.NewSuperVersion();
       assert(context->superversion_context.new_superversion.get() != nullptr);
       cfd->InstallSuperVersion(&context->superversion_context, &mutex_);
@@ -1637,6 +1655,7 @@ 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 +1701,11 @@ 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 +1731,10 @@ 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 +1769,11 @@ 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;

db/db_impl/db_secondary_test.cc

@@ -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 {
 
@@ -19,7 +19,7 @@ namespace ROCKSDB_NAMESPACE {
 class DBSecondaryTest : public DBTestBase {
  public:
   DBSecondaryTest()
-      : DBTestBase("/db_secondary_test"),
+      : DBTestBase("/db_secondary_test", /*env_do_fsync=*/true),
         secondary_path_(),
         handles_secondary_(),
         db_secondary_(nullptr) {

db/db_info_dumper.cc

@@ -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,
@@ -50,16 +53,24 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
         Header(options.info_log, "IDENTITY file:  %s\n", file.c_str());
         break;
       case kDescriptorFile:
-        env->GetFileSize(dbname + "/" + file, &file_size);
-        Header(options.info_log, "MANIFEST file:  %s size: %" PRIu64 " Bytes\n",
-               file.c_str(), file_size);
+        if (env->GetFileSize(dbname + "/" + file, &file_size).ok()) {
+          Header(options.info_log,
+                 "MANIFEST file:  %s size: %" PRIu64 " Bytes\n", file.c_str(),
+                 file_size);
+        } else {
+          Error(options.info_log, "Error when reading MANIFEST file: %s/%s\n",
+                dbname.c_str(), file.c_str());
+        }
         break;
       case kLogFile:
-        env->GetFileSize(dbname + "/" + file, &file_size);
-        char str[16];
-        snprintf(str, sizeof(str), "%" PRIu64, file_size);
-        wal_info.append(file).append(" size: ").
-            append(str).append(" ; ");
+        if (env->GetFileSize(dbname + "/" + file, &file_size).ok()) {
+          char str[16];
+          snprintf(str, sizeof(str), "%" PRIu64, file_size);
+          wal_info.append(file).append(" size: ").append(str).append(" ; ");
+        } else {
+          Error(options.info_log, "Error when reading LOG file: %s/%s\n",
+                dbname.c_str(), file.c_str());
+        }
         break;
       case kTableFile:
         if (++file_num < 10) {
@@ -108,11 +119,14 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
     for (const std::string& file : files) {
       if (ParseFileName(file, &number, &type)) {
         if (type == kLogFile) {
-          env->GetFileSize(options.wal_dir + "/" + file, &file_size);
-          char str[16];
-          snprintf(str, sizeof(str), "%" PRIu64, file_size);
-          wal_info.append(file).append(" size: ").
-              append(str).append(" ; ");
+          if (env->GetFileSize(options.wal_dir + "/" + file, &file_size).ok()) {
+            char str[16];
+            snprintf(str, sizeof(str), "%" PRIu64, file_size);
+            wal_info.append(file).append(" size: ").append(str).append(" ; ");
+          } else {
+            Error(options.info_log, "Error when reading LOG file %s/%s\n",
+                  options.wal_dir.c_str(), file.c_str());
+          }
         }
       }
     }

db/db_info_dumper.h

@@ -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

db/db_inplace_update_test.cc

@@ -13,7 +13,8 @@ namespace ROCKSDB_NAMESPACE {
 
 class DBTestInPlaceUpdate : public DBTestBase {
  public:
-  DBTestInPlaceUpdate() : DBTestBase("/db_inplace_update_test") {}
+  DBTestInPlaceUpdate()
+      : DBTestBase("/db_inplace_update_test", /*env_do_fsync=*/true) {}
 };
 
 TEST_F(DBTestInPlaceUpdate, InPlaceUpdate) {

db/db_io_failure_test.cc

@@ -9,12 +9,14 @@
 
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
+#include "util/random.h"
 
 namespace ROCKSDB_NAMESPACE {
 
 class DBIOFailureTest : public DBTestBase {
  public:
-  DBIOFailureTest() : DBTestBase("/db_io_failure_test") {}
+  DBIOFailureTest()
+      : DBTestBase("/db_io_failure_test", /*env_do_fsync=*/true) {}
 };
 
 #ifndef ROCKSDB_LITE
@@ -33,7 +35,7 @@ TEST_F(DBIOFailureTest, DropWrites) {
     // Force out-of-space errors
     env_->drop_writes_.store(true, std::memory_order_release);
     env_->sleep_counter_.Reset();
-    env_->no_slowdown_ = true;
+    env_->SetMockSleep();
     for (int i = 0; i < 5; i++) {
       if (option_config_ != kUniversalCompactionMultiLevel &&
           option_config_ != kUniversalSubcompactions) {
@@ -281,8 +283,8 @@ 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 +332,8 @@ 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

db/db_iter.cc

@@ -223,11 +223,13 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
 
     is_key_seqnum_zero_ = (ikey_.sequence == 0);
 
-    assert(iterate_upper_bound_ == nullptr || iter_.MayBeOutOfUpperBound() ||
+    assert(iterate_upper_bound_ == nullptr ||
+           iter_.UpperBoundCheckResult() != IterBoundCheck::kInbound ||
            user_comparator_.CompareWithoutTimestamp(
                ikey_.user_key, /*a_has_ts=*/true, *iterate_upper_bound_,
                /*b_has_ts=*/false) < 0);
-    if (iterate_upper_bound_ != nullptr && iter_.MayBeOutOfUpperBound() &&
+    if (iterate_upper_bound_ != nullptr &&
+        iter_.UpperBoundCheckResult() != IterBoundCheck::kInbound &&
         user_comparator_.CompareWithoutTimestamp(
             ikey_.user_key, /*a_has_ts=*/true, *iterate_upper_bound_,
             /*b_has_ts=*/false) >= 0) {
@@ -246,11 +248,10 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
     }
 
     assert(ikey_.user_key.size() >= timestamp_size_);
-    Slice ts;
+    Slice ts = timestamp_size_ > 0 ? ExtractTimestampFromUserKey(
+                                         ikey_.user_key, timestamp_size_)
+                                   : Slice();
     bool more_recent = false;
-    if (timestamp_size_ > 0) {
-      ts = ExtractTimestampFromUserKey(ikey_.user_key, timestamp_size_);
-    }
     if (IsVisible(ikey_.sequence, ts, &more_recent)) {
       // If the previous entry is of seqnum 0, the current entry will not
       // possibly be skipped. This condition can potentially be relaxed to
@@ -285,7 +286,20 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
             // 2) return ikey only if ikey.seqnum >= start_seqnum_
             // note that if deletion seqnum is < start_seqnum_ we
             // just skip it like in normal iterator.
-            if (start_seqnum_ > 0 && ikey_.sequence >= start_seqnum_)  {
+            if (start_seqnum_ > 0) {
+              if (ikey_.sequence >= start_seqnum_) {
+                saved_key_.SetInternalKey(ikey_);
+                valid_ = true;
+                return true;
+              } else {
+                saved_key_.SetUserKey(
+                    ikey_.user_key,
+                    !pin_thru_lifetime_ ||
+                        !iter_.iter()->IsKeyPinned() /* copy */);
+                skipping_saved_key = true;
+                PERF_COUNTER_ADD(internal_delete_skipped_count, 1);
+              }
+            } else if (timestamp_lb_) {
               saved_key_.SetInternalKey(ikey_);
               valid_ = true;
               return true;
@@ -300,10 +314,8 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
           case kTypeValue:
           case kTypeBlobIndex:
             if (start_seqnum_ > 0) {
-              // we are taking incremental snapshot here
-              // incremental snapshots aren't supported on DB with range deletes
-              assert(ikey_.type != kTypeBlobIndex);
               if (ikey_.sequence >= start_seqnum_) {
+                assert(ikey_.type != kTypeBlobIndex);
                 saved_key_.SetInternalKey(ikey_);
                 valid_ = true;
                 return true;
@@ -316,6 +328,10 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
                         !iter_.iter()->IsKeyPinned() /* copy */);
                 skipping_saved_key = true;
               }
+            } else if (timestamp_lb_) {
+              saved_key_.SetInternalKey(ikey_);
+              valid_ = true;
+              return true;
             } else {
               saved_key_.SetUserKey(
                   ikey_.user_key, !pin_thru_lifetime_ ||
@@ -1193,7 +1209,8 @@ void DBIter::Seek(const Slice& target) {
 
 #ifndef ROCKSDB_LITE
   if (db_impl_ != nullptr && cfd_ != nullptr) {
-    db_impl_->TraceIteratorSeek(cfd_->GetID(), target);
+    // TODO: What do we do if this returns an error?
+    db_impl_->TraceIteratorSeek(cfd_->GetID(), target).PermitUncheckedError();
   }
 #endif  // ROCKSDB_LITE
 
@@ -1254,7 +1271,9 @@ void DBIter::SeekForPrev(const Slice& target) {
 
 #ifndef ROCKSDB_LITE
   if (db_impl_ != nullptr && cfd_ != nullptr) {
-    db_impl_->TraceIteratorSeekForPrev(cfd_->GetID(), target);
+    // TODO: What do we do if this returns an error?
+    db_impl_->TraceIteratorSeekForPrev(cfd_->GetID(), target)
+        .PermitUncheckedError();
   }
 #endif  // ROCKSDB_LITE
 

db/db_iter.h

@@ -143,7 +143,7 @@ class DBIter final : public Iterator {
   bool Valid() const override { return valid_; }
   Slice key() const override {
     assert(valid_);
-    if (start_seqnum_ > 0) {
+    if (start_seqnum_ > 0 || timestamp_lb_) {
       return saved_key_.GetInternalKey();
     } else {
       const Slice ukey_and_ts = saved_key_.GetUserKey();

db/db_iter_stress_test.cc

@@ -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();