Fixup HISTORY.md for e9ba4ba "validate range tombstone covers positive range"

Moved it from the wrong section (6.10) to the right section (Unreleased).
validate range tombstone covers positive range (#6788 )
2026-07-08 07:05:23 +08:00 · 2020-05-07 14:48:30 -07:00 · 2020-05-07 11:55:30 -07:00 · 2020-05-07 09:32:51 -07:00 · 2020-05-06 18:08:09 -07:00 · 2020-05-06 16:42:37 -07:00
1041 changed files with 44922 additions and 18411 deletions
@@ -0,0 +1,56 @@
+version: 2.1
+
+orbs:
+  win: circleci/windows@2.4.0
+
+executors:
+  windows-2xlarge:
+    machine:
+      image: 'windows-server-2019-vs2019:stable'
+      resource_class: windows.2xlarge
+      shell: bash.exe
+
+jobs:
+  build:
+    executor: windows-2xlarge
+    
+    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
+
+    steps:
+      - checkout
+      - run:
+          name: "Install thirdparty dependencies"
+          command: |
+            mkdir ${THIRDPARTY_HOME}
+            cd ${THIRDPARTY_HOME}
+            echo "Installing CMake..."
+            curl --fail --silent --show-error --output cmake-3.16.4-win64-x64.zip --location https://github.com/Kitware/CMake/releases/download/v3.16.4/cmake-3.16.4-win64-x64.zip
+            unzip -q cmake-3.16.4-win64-x64.zip
+            echo "Building Snappy dependency..."
+            curl --fail --silent --show-error --output snappy-1.1.7.zip --location https://github.com/google/snappy/archive/1.1.7.zip
+            unzip -q snappy-1.1.7.zip
+            cd snappy-1.1.7
+            mkdir build
+            cd build
+            ${CMAKE_BIN} -G "${CMAKE_GENERATOR}" ..
+            msbuild.exe Snappy.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
+      - run:
+          name: "Build RocksDB"
+          command: |
+            mkdir build
+            cd build
+            ${CMAKE_BIN} -G "${CMAKE_GENERATOR}" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DJNI=1 ..
+            cd ..
+            msbuild.exe build/rocksdb.sln -maxCpuCount -property:Configuration=Debug -property:Platform=x64
+      - run:
+          name: "Test RocksDB"
+          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
@@ -0,0 +1,41 @@
+name: Check buck targets and code format
+on: [push, pull_request]
+jobs:
+  check:
+    name: Check TARGETS file and code format
+    runs-on: ubuntu-latest
+    steps:
+    - name: Checkout feature branch
+      uses: actions/checkout@v2
+      with:
+        fetch-depth: 0
+
+    - name: Fetch from upstream
+      run: |
+        git remote add upstream https://github.com/facebook/rocksdb.git && git fetch upstream
+
+    - name: Where am I
+      run: |
+        echo git status && git status
+        echo "git remote -v" && git remote -v
+        echo git branch && git branch
+
+    - name: Setup Python
+      uses: actions/setup-python@v1
+
+    - name: Install Dependencies
+      run: python -m pip install --upgrade pip
+
+    - name: Install argparse
+      run: pip install argparse
+
+    - name: Download clang-format-diff.py
+      uses: wei/wget@v1
+      with:
+        args: https://raw.githubusercontent.com/llvm-mirror/clang/master/tools/clang-format/clang-format-diff.py
+
+    - name: Check format
+      run: make check-format
+
+    - name: Compare buckify output
+      run: make check-buck-targets
@@ -34,6 +34,7 @@ manifest_dump
 sst_dump
 blob_dump
 block_cache_trace_analyzer
+db_with_timestamp_basic_test
 tools/block_cache_analyzer/*.pyc
 column_aware_encoding_exp
 util/build_version.cc
@@ -52,6 +53,8 @@ trace_analyzer
 trace_analyzer_test
 block_cache_trace_analyzer
 .DS_Store
+.vs
+.vscode

 java/out
 java/target
@@ -80,3 +83,5 @@ fbcode/
 fbcode
 buckifier/*.pyc
 buckifier/__pycache__
+
+compile_commands.json
@@ -3,12 +3,14 @@ language: cpp
 os:
  - linux
  - osx
+arch:
+  - amd64
+  - arm64
+  - ppc64le
 compiler:
  - clang
  - gcc
 osx_image: xcode9.4
-jdk:
-  - openjdk7
 cache:
  - ccache

@@ -48,26 +50,153 @@ env:
  - JOB_NAME=examples # 5-7 minutes
  - JOB_NAME=cmake # 3-5 minutes
  - JOB_NAME=cmake-gcc8 # 3-5 minutes
+  - JOB_NAME=cmake-gcc9 # 3-5 minutes
+  - JOB_NAME=cmake-gcc9-c++20 # 3-5 minutes
  - JOB_NAME=cmake-mingw # 3 minutes

 matrix:
  exclude:
-  - os: osx
-    env: TEST_GROUP=1
-  - os: osx
-    env: TEST_GROUP=2
-  - os: osx
-    env: TEST_GROUP=3
-  - os: osx
-    env: TEST_GROUP=4
  - os: osx
    env: JOB_NAME=cmake-gcc8
-  - os : osx
+  - os: osx
+    env: JOB_NAME=cmake-gcc9
+  - os: osx
+    env: JOB_NAME=cmake-gcc9-c++20
+  - os: osx
    env: JOB_NAME=cmake-mingw
-  - os : linux
-    compiler: clang
-  - os : osx
+  - os: osx
+    arch: ppc64le
+  - os: osx
    compiler: gcc
+  - os : linux
+    arch: arm64
+    env: JOB_NAME=cmake-mingw
+  - os: linux
+    arch: ppc64le
+    env: JOB_NAME=cmake-mingw
+  - os: linux
+    compiler: clang
+  # Exclude all but most unique cmake variants for pull requests, but build all in branches
+  - if: type = pull_request
+    os : linux
+    arch: amd64
+    env: JOB_NAME=cmake
+  - if: type = pull_request
+    os : linux
+    arch: amd64
+    env: JOB_NAME=cmake-gcc8
+  - if: type = pull_request
+    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
+    os: linux
+    arch: ppc64le
+    env: TEST_GROUP=platform_dependent
+  # NB: the cmake build is a partial java test
+  - if: type = pull_request
+    os: osx
+    env: TEST_GROUP=1
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: TEST_GROUP=1
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: TEST_GROUP=1
+  - if: type = pull_request
+    os: osx
+    env: TEST_GROUP=2
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: TEST_GROUP=2
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: TEST_GROUP=2
+  - if: type = pull_request
+    os: osx
+    env: TEST_GROUP=3
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: TEST_GROUP=3
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: TEST_GROUP=3
+  - if: type = pull_request
+    os: osx
+    env: TEST_GROUP=4
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: TEST_GROUP=4
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: TEST_GROUP=4
+  - if: type = pull_request AND commit_message !~ /java/
+    os : osx
+    env: JOB_NAME=java_test
+  - if: type = pull_request AND commit_message !~ /java/
+    os : linux
+    arch: arm64
+    env: JOB_NAME=java_test
+  - if: type = pull_request AND commit_message !~ /java/
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=java_test
+  - if: type = pull_request
+    os : osx
+    env: JOB_NAME=lite_build
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=lite_build
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=lite_build
+  - if: type = pull_request
+    os : osx
+    env: JOB_NAME=examples
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=examples
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=examples
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=cmake-gcc8
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=cmake-gcc8
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=cmake-gcc9
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=cmake-gcc9
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=cmake-gcc9-c++20
+  - if: type = pull_request
+    os: linux
+    arch: ppc64le
+    env: JOB_NAME=cmake-gcc9-c++20

 install:
  - if [ "${TRAVIS_OS_NAME}" == osx ]; then
@@ -77,14 +206,37 @@ install:
      sudo apt-get install -y g++-8;
      CC=gcc-8 && CXX=g++-8;
    fi
+  - if [ "${JOB_NAME}" == cmake-gcc9 ] || [ "${JOB_NAME}" == cmake-gcc9-c++20 ]; then
+      sudo apt-get install -y g++-9;
+      CC=gcc-9 && CXX=g++-9;
+    fi
  - if [ "${JOB_NAME}" == cmake-mingw ]; then
      sudo apt-get install -y mingw-w64 ;
    fi
  - if [[ "${JOB_NAME}" == cmake* ]] && [ "${TRAVIS_OS_NAME}" == linux ]; then
-      mkdir cmake-dist && curl -sfSL https://github.com/Kitware/CMake/releases/download/v3.14.5/cmake-3.14.5-Linux-x86_64.tar.gz | tar --strip-components=1 -C cmake-dist -xz && export PATH=$PWD/cmake-dist/bin:$PATH;
+      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:$PATH;
    fi
-  - if [[ "${JOB_NAME}" == java_test ]]; then
-      java -version && echo "JAVA_HOME=${JAVA_HOME}";
+  - |
+    if [[ "${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
+        export JAVA_HOME=$(/usr/libexec/java_home)
+      else
+        sudo apt-get install -y openjdk-8-jdk
+        export PATH=/usr/lib/jvm/java-8-openjdk-$(dpkg --print-architecture)/bin:$PATH
+        export JAVA_HOME=/usr/lib/jvm/java-8-openjdk-$(dpkg --print-architecture)
+      fi
+      echo "JAVA_HOME=${JAVA_HOME}"
+      which java && java -version
+      which javac && javac -version
    fi

 before_script:
@@ -93,7 +245,7 @@ before_script:
  - ulimit -n 8192

 script:
-  - ${CXX} --version
+  - date; ${CXX} --version
  - if [ `command -v ccache` ]; then ccache -C; fi
  - case $TEST_GROUP in
    platform_dependent)
@@ -103,7 +255,7 @@ script:
      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=db_block_cache_test ROCKSDBTESTS_END=db_iter_test make -j4 check_some
      ;;
    2)
-      OPT=-DTRAVIS V=1 make -j4 tools && OPT=-DTRAVIS 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_START=db_iter_test ROCKSDBTESTS_END=options_file_test make -j4 check_some
      ;;
    3)
      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=options_file_test ROCKSDBTESTS_END=write_prepared_transaction_test make -j4 check_some
@@ -127,7 +279,13 @@ script:
      mkdir build && cd build && cmake -DJNI=1 -DWITH_GFLAGS=OFF .. -DCMAKE_C_COMPILER=x86_64-w64-mingw32-gcc -DCMAKE_CXX_COMPILER=x86_64-w64-mingw32-g++ -DCMAKE_SYSTEM_NAME=Windows && make -j4 rocksdb rocksdbjni
      ;;
    cmake*)
-      mkdir build && cd build && cmake -DJNI=1 .. -DCMAKE_BUILD_TYPE=Release && make -j4 rocksdb rocksdbjni
+      case $JOB_NAME in
+        *-c++20)
+          OPT=-DCMAKE_CXX_STANDARD=20
+          ;;
+      esac
+
+      mkdir build && cd build && cmake -DJNI=1 .. -DCMAKE_BUILD_TYPE=Release $OPT && make -j4 rocksdb rocksdbjni
      ;;
    esac
 notifications:
@@ -83,6 +83,10 @@ else()
  option(WITH_FOLLY_DISTRIBUTED_MUTEX "build with folly::DistributedMutex" OFF)
 endif()

+if( NOT DEFINED CMAKE_CXX_STANDARD )
+  set(CMAKE_CXX_STANDARD 11)
+endif()
+
 include(CMakeDependentOption)
 CMAKE_DEPENDENT_OPTION(WITH_GFLAGS "build with GFlags" ON
  "NOT MSVC;NOT MINGW" OFF)
@@ -191,7 +195,6 @@ else()
    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-format -fno-asynchronous-unwind-tables")
    add_definitions(-D_POSIX_C_SOURCE=1)
  endif()
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11")
  if(NOT CMAKE_BUILD_TYPE STREQUAL "Debug")
    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-omit-frame-pointer")
    include(CheckCXXCompilerFlag)
@@ -203,16 +206,25 @@ else()
 endif()

 include(CheckCCompilerFlag)
-if(CMAKE_SYSTEM_PROCESSOR MATCHES "ppc64le")
+if(CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64")
+  CHECK_C_COMPILER_FLAG("-mcpu=power9" HAS_POWER9)
+  if(HAS_POWER9)
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mcpu=power9 -mtune=power9")
+    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mcpu=power9 -mtune=power9")
+  else()
+    CHECK_C_COMPILER_FLAG("-mcpu=power8" HAS_POWER8)
+    if(HAS_POWER8)
+      set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mcpu=power8 -mtune=power8")
+      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mcpu=power8 -mtune=power8")
+    endif(HAS_POWER8)
+  endif(HAS_POWER9)
  CHECK_C_COMPILER_FLAG("-maltivec" HAS_ALTIVEC)
  if(HAS_ALTIVEC)
    message(STATUS " HAS_ALTIVEC yes")
    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -maltivec")
    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -maltivec")
-    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -mcpu=power8")
-    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -mcpu=power8")
  endif(HAS_ALTIVEC)
-endif(CMAKE_SYSTEM_PROCESSOR MATCHES "ppc64le")
+endif(CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64")

 if(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|AARCH64")
        CHECK_C_COMPILER_FLAG("-march=armv8-a+crc+crypto" HAS_ARMV8_CRC)
@@ -235,7 +247,7 @@ else()
  if(MSVC)
    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /arch:AVX2")
  else()
-    if(NOT HAVE_POWER8 AND NOT HAS_ARMV8_CRC)
+    if(NOT CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64" AND NOT HAS_ARMV8_CRC)
      set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -march=native")
    endif()
  endif()
@@ -245,6 +257,7 @@ include(CheckCXXSourceCompiles)
 if(NOT MSVC)
  set(CMAKE_REQUIRED_FLAGS "-msse4.2 -mpclmul")
 endif()
+
 CHECK_CXX_SOURCE_COMPILES("
 #include <cstdint>
 #include <nmmintrin.h>
@@ -377,7 +390,15 @@ if(MSVC)
    message(STATUS "Debug optimization is enabled")
    set(CMAKE_CXX_FLAGS_DEBUG "/Oxt")
  else()
-    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm")
+    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1")
+
+    # Minimal Build is deprecated after MSVC 2015
+    if( MSVC_VERSION GREATER 1900 )
+      set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Gm-")
+    else()
+      set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Gm")
+    endif()
+
  endif()
  if(WITH_RUNTIME_DEBUG)
    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /${RUNTIME_LIBRARY}d")
@@ -481,6 +502,11 @@ if(HAVE_SCHED_GETCPU)
  add_definitions(-DROCKSDB_SCHED_GETCPU_PRESENT)
 endif()

+check_cxx_symbol_exists(getauxval auvx.h HAVE_AUXV_GETAUXVAL)
+if(HAVE_AUXV_GETAUXVAL)
+  add_definitions(-DROCKSDB_AUXV_GETAUXVAL_PRESENT)
+endif()
+
 include_directories(${PROJECT_SOURCE_DIR})
 include_directories(${PROJECT_SOURCE_DIR}/include)
 include_directories(SYSTEM ${PROJECT_SOURCE_DIR}/third-party/gtest-1.8.1/fused-src)
@@ -492,10 +518,14 @@ find_package(Threads REQUIRED)
 # Main library source code

 set(SOURCES
+        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_garbage.cc
+        db/blob/blob_file_meta.cc
        db/builder.cc
        db/c.cc
        db/column_family.cc
@@ -549,6 +579,7 @@ set(SOURCES
        db/trim_history_scheduler.cc
        db/version_builder.cc
        db/version_edit.cc
+        db/version_edit_handler.cc
        db/version_set.cc
        db/wal_manager.cc
        db/write_batch.cc
@@ -577,6 +608,7 @@ set(SOURCES
        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
@@ -601,15 +633,17 @@ set(SOURCES
        options/options.cc
        options/options_helper.cc
        options/options_parser.cc
-        options/options_sanity_check.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_hash_index.cc
        table/block_based/data_block_footer.cc
@@ -617,9 +651,14 @@ set(SOURCES
        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
@@ -665,6 +704,7 @@ set(SOURCES
        util/random.cc
        util/rate_limiter.cc
        util/slice.cc
+        util/file_checksum_helper.cc
        util/status.cc
        util/string_util.cc
        util/thread_local.cc
@@ -733,11 +773,11 @@ if(HAVE_SSE42 AND NOT MSVC)
    PROPERTIES COMPILE_FLAGS "-msse4.2 -mpclmul")
 endif()

-if(HAVE_POWER8)
+if(CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64")
  list(APPEND SOURCES
    util/crc32c_ppc.c
    util/crc32c_ppc_asm.S)
-endif(HAVE_POWER8)
+endif(CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64")

 if(HAS_ARMV8_CRC)
  list(APPEND SOURCES
@@ -821,7 +861,6 @@ if(ROCKSDB_BUILD_SHARED)
                          LINKER_LANGUAGE CXX
                          VERSION ${rocksdb_VERSION}
                          SOVERSION ${rocksdb_VERSION_MAJOR}
-                          CXX_STANDARD 11
                          OUTPUT_NAME "rocksdb")
  endif()
 endif()
@@ -906,18 +945,21 @@ if(NOT WIN32 OR ROCKSDB_INSTALL_ON_WINDOWS)
  )
 endif()

-add_subdirectory(third-party/gtest-1.8.1/fused-src/gtest)
-add_library(testharness STATIC
-  test_util/testharness.cc)
-target_link_libraries(testharness gtest)
-
 # Tests are excluded from Release builds
 CMAKE_DEPENDENT_OPTION(WITH_TESTS "build with tests" ON
  "CMAKE_BUILD_TYPE STREQUAL Debug" OFF)
 if(WITH_TESTS)
+  add_subdirectory(third-party/gtest-1.8.1/fused-src/gtest)
+  add_library(testharness STATIC
+  test_util/testharness.cc)
+  target_link_libraries(testharness gtest)
+
  set(TESTS
        cache/cache_test.cc
        cache/lru_cache_test.cc
+        db/blob/blob_file_addition_test.cc
+        db/blob/blob_file_garbage_test.cc
+        db/blob/db_blob_index_test.cc
        db/column_family_test.cc
        db/compact_files_test.cc
        db/compaction/compaction_job_stats_test.cc
@@ -928,7 +970,7 @@ if(WITH_TESTS)
        db/corruption_test.cc
        db/cuckoo_table_db_test.cc
        db/db_basic_test.cc
-        db/db_blob_index_test.cc
+        db/db_with_timestamp_basic_test.cc
        db/db_block_cache_test.cc
        db/db_bloom_filter_test.cc
        db/db_compaction_filter_test.cc
@@ -954,12 +996,14 @@ if(WITH_TESTS)
        db/db_tailing_iter_test.cc
        db/db_test.cc
        db/db_test2.cc
+        db/db_logical_block_size_cache_test.cc
        db/db_universal_compaction_test.cc
        db/db_wal_test.cc
+        db/db_with_timestamp_compaction_test.cc
        db/db_write_test.cc
        db/dbformat_test.cc
        db/deletefile_test.cc
-        db/error_handler_test.cc
+        db/error_handler_fs_test.cc
        db/obsolete_files_test.cc
        db/external_sst_file_basic_test.cc
        db/external_sst_file_test.cc
@@ -990,12 +1034,15 @@ if(WITH_TESTS)
        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/random_access_file_reader_test.cc
        logging/auto_roll_logger_test.cc
        logging/env_logger_test.cc
        logging/event_logger_test.cc
        memory/arena_test.cc
+        memory/memkind_kmem_allocator_test.cc
        memtable/inlineskiplist_test.cc
        memtable/skiplist_test.cc
        memtable/write_buffer_manager_test.cc
@@ -1016,6 +1063,7 @@ if(WITH_TESTS)
        table/merger_test.cc
        table/sst_file_reader_test.cc
        table/table_test.cc
+        table/block_fetcher_test.cc
        tools/block_cache_analyzer/block_cache_trace_analyzer_test.cc
        tools/ldb_cmd_test.cc
        tools/reduce_levels_test.cc
@@ -1025,6 +1073,7 @@ if(WITH_TESTS)
        util/bloom_test.cc
        util/coding_test.cc
        util/crc32c_test.cc
+        util/defer_test.cc
        util/dynamic_bloom_test.cc
        util/file_reader_writer_test.cc
        util/filelock_test.cc
@@ -1033,10 +1082,13 @@ if(WITH_TESTS)
        util/random_test.cc
        util/rate_limiter_test.cc
        util/repeatable_thread_test.cc
+        util/slice_test.cc
        util/slice_transform_test.cc
        util/timer_queue_test.cc
+        util/timer_test.cc
        util/thread_list_test.cc
        util/thread_local_test.cc
+        util/work_queue_test.cc
        utilities/backupable/backupable_db_test.cc
        utilities/blob_db/blob_db_test.cc
        utilities/cassandra/cassandra_functional_test.cc
@@ -1056,6 +1108,7 @@ if(WITH_TESTS)
        utilities/table_properties_collectors/compact_on_deletion_collector_test.cc
        utilities/transactions/optimistic_transaction_test.cc
        utilities/transactions/transaction_test.cc
+        utilities/transactions/transaction_lock_mgr_test.cc
        utilities/transactions/write_prepared_transaction_test.cc
        utilities/transactions/write_unprepared_transaction_test.cc
        utilities/ttl/ttl_test.cc
@@ -1074,6 +1127,7 @@ if(WITH_TESTS)
      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()
@@ -1088,7 +1142,7 @@ if(WITH_TESTS)
        PROPERTIES EXCLUDE_FROM_DEFAULT_BUILD_RELEASE 1
        EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
        EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
-        )
+  )

  foreach(sourcefile ${TESTS})
      get_filename_component(exename ${sourcefile} NAME_WE)
@@ -1098,7 +1152,7 @@ if(WITH_TESTS)
        EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
        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 ${ROCKSDB_LIB})
      if(NOT "${exename}" MATCHES "db_sanity_test")
        add_test(NAME ${exename} COMMAND ${exename}${ARTIFACT_SUFFIX})
@@ -1166,9 +1220,15 @@ if(WITH_BENCHMARK_TOOLS)
    ${ROCKSDB_LIB})
 endif()

+option(WITH_CORE_TOOLS "build with ldb and sst_dump" ON)
 option(WITH_TOOLS "build with tools" ON)
-if(WITH_TOOLS)
+if(WITH_CORE_TOOLS OR WITH_TOOLS)
  add_subdirectory(tools)
+  add_custom_target(core_tools
+    DEPENDS ${core_tool_deps})
+endif()
+
+if(WITH_TOOLS)
  add_subdirectory(db_stress_tool)
  add_custom_target(tools
    DEPENDS ${tool_deps})
@@ -1,7 +1,98 @@
 # Rocksdb Change Log
 ## Unreleased
+### Bug Fixes
+* Fix consistency checking error swallowing in some cases when options.force_consistency_checks = true.
+
+### Public API Change
+* Flush(..., column_family) may return Status::ColumnFamilyDropped() instead of Status::InvalidArgument() if column_family is dropped while processing the flush request.
+* BlobDB now explicitly disallows using the default column family's storage directories as blob directory.
+* DeleteRange now returns `Status::InvalidArgument` if the range's end key comes before its start key according to the user comparator. Previously the behavior was undefined.
+
+## 6.10 (5/2/2020)
+### Bug Fixes
+* Fix wrong result being read from ingested file. May happen when a key in the file happen to be prefix of another key also in the file. The issue can further cause more data corruption. The issue exists with rocksdb >= 5.0.0 since DB::IngestExternalFile() was introduced.
+* Finish implementation of BlockBasedTableOptions::IndexType::kBinarySearchWithFirstKey. It's now ready for use. Significantly reduces read amplification in some setups, especially for iterator seeks.
+* Fix a bug by updating CURRENT file so that it points to the correct MANIFEST file after best-efforts recovery.
+* Fixed a bug where ColumnFamilyHandle objects were not cleaned up in case an error happened during BlobDB's open after the base DB had been opened.
+* Fix a potential undefined behavior caused by trying to dereference nullable pointer (timestamp argument) in DB::MultiGet.
+* Fix a bug caused by not including user timestamp in MultiGet LookupKey construction. This can lead to wrong query result since the trailing bytes of a user key, if not shorter than timestamp, will be mistaken for user timestamp.
+* Fix a bug caused by using wrong compare function when sorting the input keys of MultiGet with timestamps.
+* Upgraded version of bzip library (1.0.6 -> 1.0.8) used with RocksJava to address potential vulnerabilities if an attacker can manipulate compressed data saved and loaded by RocksDB (not normal). See issue #6703.
+
+### Public API Change
+* Add a ConfigOptions argument to the APIs dealing with converting options to and from strings and files.  The ConfigOptions is meant to replace some of the options (such as input_strings_escaped and ignore_unknown_options) and allow for more parameters to be passed in the future without changing the function signature.
+* Add NewFileChecksumGenCrc32cFactory to the file checksum public API, such that the builtin Crc32c based file checksum generator factory can be used by applications.
+* Add IsDirectory to Env and FS to indicate if a path is a directory.
+
+### New Features
+* Added support for pipelined & parallel compression optimization for `BlockBasedTableBuilder`. This optimization makes block building, block compression and block appending a pipeline, and uses multiple threads to accelerate block compression. Users can set `CompressionOptions::parallel_threads` greater than 1 to enable compression parallelism. This feature is experimental for now.
+* Provide an allocator for memkind to be used with block cache. This is to work with memory technologies (Intel DCPMM is one such technology currently available) that require different libraries for allocation and management (such as PMDK and memkind). The high capacities available make it possible to provision large caches (up to several TBs in size) beyond what is achievable with DRAM.
+* Option `max_background_flushes` can be set dynamically using DB::SetDBOptions().
+* Added functionality in sst_dump tool to check the compressed file size for different compression levels and print the time spent on compressing files with each compression type. Added arguments `--compression_level_from` and `--compression_level_to` to report size of all compression levels and one compression_type must be specified with it so that it will report compressed sizes of one compression type with different levels.
+* Added statistics for redundant insertions into block cache: rocksdb.block.cache.*add.redundant. (There is currently no coordination to ensure that only one thread loads a table block when many threads are trying to access that same table block.)
+
+### Bug Fixes
+* Fix a bug when making options.bottommost_compression, options.compression_opts and options.bottommost_compression_opts dynamically changeable: the modified values are not written to option files or returned back to users when being queried.
+* Fix a bug where index key comparisons were unaccounted in `PerfContext::user_key_comparison_count` for lookups in files written with `format_version >= 3`.
+* Fix many bloom.filter statistics not being updated in batch MultiGet.
+
+### Performance Improvements
+* Improve performance of batch MultiGet with partitioned filters, by sharing block cache lookups to applicable filter blocks.
+* Reduced memory copies when fetching and uncompressing compressed blocks from sst files.
+
+## 6.9.0 (03/29/2020)
+### Behavior changes
+* Since RocksDB 6.8, ttl-based FIFO compaction can drop a file whose oldest key becomes older than options.ttl while others have not. This fix reverts this and makes ttl-based FIFO compaction use the file's flush time as the criterion. This fix also requires that max_open_files = -1 and compaction_options_fifo.allow_compaction = false to function properly.
+
+### Public API Change
+* Fix spelling so that API now has correctly spelled transaction state name `COMMITTED`, while the old misspelled `COMMITED` is still available as an alias.
+* Updated default format_version in BlockBasedTableOptions from 2 to 4. SST files generated with the new default can be read by RocksDB versions 5.16 and newer, and use more efficient encoding of keys in index blocks.
+* A new parameter `CreateBackupOptions` is added to both `BackupEngine::CreateNewBackup` and `BackupEngine::CreateNewBackupWithMetadata`, you can decrease CPU priority of `BackupEngine`'s background threads by setting `decrease_background_thread_cpu_priority` and `background_thread_cpu_priority` in `CreateBackupOptions`.
+* Updated the public API of SST file checksum. Introduce the FileChecksumGenFactory to create the FileChecksumGenerator for each SST file, such that the FileChecksumGenerator is not shared and it can be more general for checksum implementations. Changed the FileChecksumGenerator interface from Value, Extend, and GetChecksum to Update, Finalize, and GetChecksum. Finalize should be only called once after all data is processed to generate the final checksum. Temproal data should be maintained by the FileChecksumGenerator object itself and finally it can return the checksum string.
+
+### Bug Fixes
+* Fix a bug where range tombstone blocks in ingested files were cached incorrectly during ingestion. If range tombstones were read from those incorrectly cached blocks, the keys they covered would be exposed.
+* Fix a data race that might cause crash when calling DB::GetCreationTimeOfOldestFile() by a small chance. The bug was introduced in 6.6 Release.
+* Fix a bug where a boolean value optimize_filters_for_hits was for max threads when calling load table handles after a flush or compaction. The value is correct to 1. The bug should not cause user visible problems.
+* Fix a bug which might crash the service when write buffer manager fails to insert the dummy handle to the block cache.
+
+### Performance Improvements
+* In CompactRange, for levels starting from 0, if the level does not have any file with any key falling in the specified range, the level is skipped. So instead of always compacting from level 0, the compaction starts from the first level with keys in the specified range until the last such level.
+* Reduced memory copy when reading sst footer and blobdb in direct IO mode.
+* When restarting a database with large numbers of sst files, large amount of CPU time is spent on getting logical block size of the sst files, which slows down the starting progress, this inefficiency is optimized away with an internal cache for the logical block sizes.
+
+### New Features
+* Basic support for user timestamp in iterator. Seek/SeekToFirst/Next and lower/upper bounds are supported. Reverse iteration is not supported. Merge is not considered.
+* When file lock failure when the lock is held by the current process, return acquiring time and thread ID in the error message.
+* Added a new option, best_efforts_recovery (default: false), to allow database to open in a db dir with missing table files. During best efforts recovery, missing table files are ignored, and database recovers to the most recent state without missing table file. Cross-column-family consistency is not guaranteed even if WAL is enabled.
+* options.bottommost_compression, options.compression_opts and options.bottommost_compression_opts are now dynamically changeable.
+
+## 6.8.0 (02/24/2020)
+### Java API Changes
+* Major breaking changes to Java comparators, toward standardizing on ByteBuffer for performant, locale-neutral operations on keys (#6252).
+* Added overloads of common API methods using direct ByteBuffers for keys and values (#2283).
+
 ### Bug Fixes
 * Fix incorrect results while block-based table uses kHashSearch, together with Prev()/SeekForPrev().
+* Fix a bug that prevents opening a DB after two consecutive crash with TransactionDB, where the first crash recovers from a corrupted WAL with kPointInTimeRecovery but the second cannot.
+* Fixed issue #6316 that can cause a corruption of the MANIFEST file in the middle when writing to it fails due to no disk space.
+* Add DBOptions::skip_checking_sst_file_sizes_on_db_open. It disables potentially expensive checking of all sst file sizes in DB::Open().
+* BlobDB now ignores trivially moved files when updating the mapping between blob files and SSTs. This should mitigate issue #6338 where out of order flush/compaction notifications could trigger an assertion with the earlier code.
+* Batched MultiGet() ignores IO errors while reading data blocks, causing it to potentially continue looking for a key and returning stale results.
+* `WriteBatchWithIndex::DeleteRange` returns `Status::NotSupported`. Previously it returned success even though reads on the batch did not account for range tombstones. The corresponding language bindings now cannot be used. In C, that includes `rocksdb_writebatch_wi_delete_range`, `rocksdb_writebatch_wi_delete_range_cf`, `rocksdb_writebatch_wi_delete_rangev`, and `rocksdb_writebatch_wi_delete_rangev_cf`. In Java, that includes `WriteBatchWithIndex::deleteRange`.
+* Assign new MANIFEST file number when caller tries to create a new MANIFEST by calling LogAndApply(..., new_descriptor_log=true). This bug can cause MANIFEST being overwritten during recovery if options.write_dbid_to_manifest = true and there are WAL file(s).
+
+### Performance Improvements
+* Perfom readahead when reading from option files. Inside DB, options.log_readahead_size will be used as the readahead size. In other cases, a default 512KB is used.
+
+### Public API Change
+* The BlobDB garbage collector now emits the statistics `BLOB_DB_GC_NUM_FILES` (number of blob files obsoleted during GC), `BLOB_DB_GC_NUM_NEW_FILES` (number of new blob files generated during GC), `BLOB_DB_GC_FAILURES` (number of failed GC passes), `BLOB_DB_GC_NUM_KEYS_RELOCATED` (number of blobs relocated during GC), and `BLOB_DB_GC_BYTES_RELOCATED` (total size of blobs relocated during GC). On the other hand, the following statistics, which are not relevant for the new GC implementation, are now deprecated: `BLOB_DB_GC_NUM_KEYS_OVERWRITTEN`, `BLOB_DB_GC_NUM_KEYS_EXPIRED`, `BLOB_DB_GC_BYTES_OVERWRITTEN`, `BLOB_DB_GC_BYTES_EXPIRED`, and `BLOB_DB_GC_MICROS`.
+* Disable recycle_log_file_num when an inconsistent recovery modes are requested: kPointInTimeRecovery and kAbsoluteConsistency
+
+### New Features
+* Added the checksum for each SST file generated by Flush or Compaction. Added sst_file_checksum_func to Options such that user can plugin their own SST file checksum function via override the FileChecksumFunc class. If user does not set the sst_file_checksum_func, SST file checksum calculation will not be enabled. The checksum information inlcuding uint32_t checksum value and a checksum function name (string). The checksum information is stored in FileMetadata in version store and also logged to MANIFEST. A new tool is added to LDB such that user can dump out a list of file checksum information from MANIFEST (stored in an unordered_map).
+* `db_bench` now supports `value_size_distribution_type`, `value_size_min`, `value_size_max` options for generating random variable sized value. Added `blob_db_compression_type` option for BlobDB to enable blob compression.
+* Replace RocksDB namespace "rocksdb" with flag "ROCKSDB_NAMESPACE" which if is not defined, defined as "rocksdb" in header file rocksdb_namespace.h.

 ## 6.7.0 (01/21/2020)
 ### Public API Change
@@ -25,6 +116,8 @@
 * It is now possible to enable periodic compactions for the base DB when using BlobDB.
 * BlobDB now garbage collects non-TTL blobs when `enable_garbage_collection` is set to `true` in `BlobDBOptions`. Garbage collection is performed during compaction: any valid blobs located in the oldest N files (where N is the number of non-TTL blob files multiplied by the value of `BlobDBOptions::garbage_collection_cutoff`) encountered during compaction get relocated to new blob files, and old blob files are dropped once they are no longer needed. Note: we recommend enabling periodic compactions for the base DB when using this feature to deal with the case when some old blob files are kept alive by SSTs that otherwise do not get picked for compaction.
 * `db_bench` now supports the `garbage_collection_cutoff` option for BlobDB.
+* Introduce ReadOptions.auto_prefix_mode. When set to true, iterator will return the same result as total order seek, but may choose to use prefix seek internally based on seek key and iterator upper bound.
+* MultiGet() can use IO Uring to parallelize read from the same SST file. This featuer is by default disabled. It can be enabled with environment variable ROCKSDB_USE_IO_URING.

 ## 6.6.2 (01/13/2020)
 ### Bug Fixes
@@ -8,6 +8,7 @@

 BASH_EXISTS := $(shell which bash)
 SHELL := $(shell which bash)
+PYTHON?=$(shell which python)

 CLEAN_FILES = # deliberately empty, so we can append below.
 CFLAGS += ${EXTRA_CFLAGS}
@@ -203,10 +204,20 @@ LDFLAGS += -lrados
 endif

 AM_LINK = $(AM_V_CCLD)$(CXX) $^ $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS)
-# detect what platform we're building on
-dummy := $(shell (export ROCKSDB_ROOT="$(CURDIR)"; export PORTABLE="$(PORTABLE)"; "$(CURDIR)/build_tools/build_detect_platform" "$(CURDIR)/make_config.mk"))
+# Detect what platform we're building on.
+# Export some common variables that might have been passed as Make variables
+# instead of environment variables.
+dummy := $(shell (export ROCKSDB_ROOT="$(CURDIR)"; \
+                  export COMPILE_WITH_ASAN="$(COMPILE_WITH_ASAN)"; \
+                  export COMPILE_WITH_TSAN="$(COMPILE_WITH_TSAN)"; \
+                  export COMPILE_WITH_UBSAN="$(COMPILE_WITH_UBSAN)"; \
+                  export PORTABLE="$(PORTABLE)"; \
+                  export ROCKSDB_NO_FBCODE="$(ROCKSDB_NO_FBCODE)"; \
+                  export USE_CLANG="$(USE_CLANG)"; \
+                  "$(CURDIR)/build_tools/build_detect_platform" "$(CURDIR)/make_config.mk"))
 # this file is generated by the previous line to set build flags and sources
 include make_config.mk
+export JAVAC_ARGS
 CLEAN_FILES += make_config.mk

 missing_make_config_paths := $(shell				\
@@ -216,7 +227,7 @@ missing_make_config_paths := $(shell				\
 	done | sort | uniq)

 $(foreach path, $(missing_make_config_paths), \
-	$(warning Warning: $(path) dont exist))
+	$(warning Warning: $(path) does not exist))

 ifeq ($(PLATFORM), OS_AIX)
 # no debug info
@@ -439,12 +450,18 @@ BENCHTOOLOBJECTS = $(BENCH_LIB_SOURCES:.cc=.o) $(LIBOBJECTS) $(TESTUTIL)

 ANALYZETOOLOBJECTS = $(ANALYZER_LIB_SOURCES:.cc=.o)

+ifeq ($(DEBUG_LEVEL),0)
 STRESSTOOLOBJECTS = $(STRESS_LIB_SOURCES:.cc=.o) $(LIBOBJECTS) $(TESTUTIL)
+else
+STRESSTOOLOBJECTS = $(STRESS_LIB_SOURCES:.cc=.o) $(LIBOBJECTS) $(TESTUTIL) \
+	$(TESTHARNESS)
+endif

 EXPOBJECTS = $(LIBOBJECTS) $(TESTUTIL)

 TESTS = \
 	db_basic_test \
+	db_with_timestamp_basic_test \
 	db_encryption_test \
 	db_test2 \
 	external_sst_file_basic_test \
@@ -459,15 +476,18 @@ TESTS = \
 	env_basic_test \
 	env_test \
 	env_logger_test \
+	io_posix_test \
 	hash_test \
 	random_test \
 	thread_local_test \
+	work_queue_test \
 	rate_limiter_test \
 	perf_context_test \
 	iostats_context_test \
 	db_wal_test \
 	db_block_cache_test \
 	db_test \
+	db_logical_block_size_cache_test \
 	db_blob_index_test \
 	db_iter_test \
 	db_iter_stress_test \
@@ -492,22 +512,25 @@ TESTS = \
 	db_table_properties_test \
 	db_statistics_test \
 	db_write_test \
-	error_handler_test \
+	error_handler_fs_test \
 	autovector_test \
 	blob_db_test \
 	cleanable_test \
 	column_family_test \
 	table_properties_collector_test \
 	arena_test \
+	memkind_kmem_allocator_test \
 	block_test \
 	data_block_hash_index_test \
 	cache_test \
 	corruption_test \
+	slice_test \
 	slice_transform_test \
 	dbformat_test \
 	fault_injection_test \
 	filelock_test \
 	filename_test \
+	random_access_file_reader_test \
 	file_reader_writer_test \
 	block_based_filter_block_test \
 	full_filter_block_test \
@@ -552,6 +575,7 @@ TESTS = \
 	deletefile_test \
 	obsolete_files_test \
 	table_test \
+	block_fetcher_test \
 	delete_scheduler_test \
 	options_test \
 	options_settable_test \
@@ -576,6 +600,7 @@ TESTS = \
 	compaction_job_stats_test \
 	option_change_migration_test \
 	transaction_test \
+	transaction_lock_mgr_test \
 	ldb_cmd_test \
 	persistent_cache_test \
 	statistics_test \
@@ -595,6 +620,11 @@ TESTS = \
 	db_secondary_test \
 	block_cache_tracer_test \
 	block_cache_trace_analyzer_test \
+	defer_test \
+	blob_file_addition_test \
+	blob_file_garbage_test \
+	timer_test \
+	db_with_timestamp_compaction_test \

 ifeq ($(USE_FOLLY_DISTRIBUTED_MUTEX),1)
 	TESTS += folly_synchronization_distributed_mutex_test
@@ -619,6 +649,7 @@ PARALLEL_TEST = \
 	persistent_cache_test \
 	table_test \
 	transaction_test \
+	transaction_lock_mgr_test \
 	write_prepared_transaction_test \
 	write_unprepared_transaction_test \

@@ -935,24 +966,26 @@ check: all
 	    $(MAKE) T="$$t" TMPD=$(TMPD) check_0;                       \
 	else                                                            \
 	    for t in $(TESTS); do                                       \
-	      echo "===== Running $$t"; ./$$t || exit 1; done;          \
+	      echo "===== Running $$t (`date`)"; ./$$t || exit 1; done;          \
 	fi
 	rm -rf $(TMPD)
 ifneq ($(PLATFORM), OS_AIX)
-	python tools/check_all_python.py
+	$(PYTHON) tools/check_all_python.py
 ifeq ($(filter -DROCKSDB_LITE,$(OPT)),)
-	python tools/ldb_test.py
+	$(PYTHON) tools/ldb_test.py
 	sh tools/rocksdb_dump_test.sh
 endif
 endif
+	$(MAKE) check-format
+	$(MAKE) check-buck-targets

 # TODO add ldb_tests
 check_some: $(SUBSET)
-	for t in $(SUBSET); do echo "===== Running $$t"; ./$$t || exit 1; done
+	for t in $(SUBSET); do echo "===== Running $$t (`date`)"; ./$$t || exit 1; done

 .PHONY: ldb_tests
 ldb_tests: ldb
-	python tools/ldb_test.py
+	$(PYTHON) tools/ldb_test.py

 crash_test: whitebox_crash_test blackbox_crash_test

@@ -961,31 +994,31 @@ crash_test_with_atomic_flush: whitebox_crash_test_with_atomic_flush blackbox_cra
 crash_test_with_txn: whitebox_crash_test_with_txn blackbox_crash_test_with_txn

 blackbox_crash_test: db_stress
-	python -u tools/db_crashtest.py --simple blackbox $(CRASH_TEST_EXT_ARGS)
-	python -u tools/db_crashtest.py blackbox $(CRASH_TEST_EXT_ARGS)
+	$(PYTHON) -u tools/db_crashtest.py --simple blackbox $(CRASH_TEST_EXT_ARGS)
+	$(PYTHON) -u tools/db_crashtest.py blackbox $(CRASH_TEST_EXT_ARGS)

 blackbox_crash_test_with_atomic_flush: db_stress
-	python -u tools/db_crashtest.py --cf_consistency blackbox $(CRASH_TEST_EXT_ARGS)
+	$(PYTHON) -u tools/db_crashtest.py --cf_consistency blackbox $(CRASH_TEST_EXT_ARGS)

 blackbox_crash_test_with_txn: db_stress
-	python -u tools/db_crashtest.py --txn blackbox $(CRASH_TEST_EXT_ARGS)
+	$(PYTHON) -u tools/db_crashtest.py --txn blackbox $(CRASH_TEST_EXT_ARGS)

 ifeq ($(CRASH_TEST_KILL_ODD),)
  CRASH_TEST_KILL_ODD=888887
 endif

 whitebox_crash_test: db_stress
-	python -u tools/db_crashtest.py --simple whitebox --random_kill_odd \
+	$(PYTHON) -u tools/db_crashtest.py --simple whitebox --random_kill_odd \
      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)
-	python -u tools/db_crashtest.py whitebox  --random_kill_odd \
+	$(PYTHON) -u tools/db_crashtest.py whitebox  --random_kill_odd \
      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)

 whitebox_crash_test_with_atomic_flush: db_stress
-	python -u tools/db_crashtest.py --cf_consistency whitebox  --random_kill_odd \
+	$(PYTHON) -u tools/db_crashtest.py --cf_consistency whitebox  --random_kill_odd \
      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)

 whitebox_crash_test_with_txn: db_stress
-	python -u tools/db_crashtest.py --txn whitebox --random_kill_odd \
+	$(PYTHON) -u tools/db_crashtest.py --txn whitebox --random_kill_odd \
      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)

 asan_check:
@@ -1054,7 +1087,7 @@ ifneq ($(PAR_TEST),)
 parloop:
 	ret_bad=0;							\
 	for t in $(PAR_TEST); do		\
-		echo "===== Running $$t in parallel $(NUM_PAR)";\
+		echo "===== Running $$t in parallel $(NUM_PAR) (`date`)";\
 		if [ $(db_test) -eq 1 ]; then \
 			seq $(J) | v="$$t" build_tools/gnu_parallel --gnu --plain 's=$(TMPD)/rdb-{};  export TEST_TMPDIR=$$s;' \
 				'timeout 2m ./db_test --gtest_filter=$$v >> $$s/log-{} 2>1'; \
@@ -1128,16 +1161,21 @@ unity_test: db/db_test.o db/db_test_util.o $(TESTHARNESS) $(TOOLLIBOBJECTS) unit
 rocksdb.h rocksdb.cc: build_tools/amalgamate.py Makefile $(LIB_SOURCES) unity.cc
 	build_tools/amalgamate.py -I. -i./include unity.cc -x include/rocksdb/c.h -H rocksdb.h -o rocksdb.cc

-clean: clean-ext-libraries-all clean-rocks
+clean: clean-ext-libraries-all clean-rocks clean-rocksjava

-clean-not-downloaded: clean-ext-libraries-bin clean-rocks
+clean-not-downloaded: clean-ext-libraries-bin clean-rocks clean-not-downloaded-rocksjava

 clean-rocks:
 	rm -f $(BENCHMARKS) $(TOOLS) $(TESTS) $(PARALLEL_TEST) $(LIBRARY) $(SHARED)
 	rm -rf $(CLEAN_FILES) ios-x86 ios-arm scan_build_report
 	$(FIND) . -name "*.[oda]" -exec rm -f {} \;
 	$(FIND) . -type f -regex ".*\.\(\(gcda\)\|\(gcno\)\)" -exec rm {} \;
-	cd java; $(MAKE) clean
+
+clean-rocksjava:
+	cd java && $(MAKE) clean
+
+clean-not-downloaded-rocksjava:
+	cd java && $(MAKE) clean-not-downloaded

 clean-ext-libraries-all:
 	rm -rf bzip2* snappy* zlib* lz4* zstd*
@@ -1160,6 +1198,12 @@ tags0:
 format:
 	build_tools/format-diff.sh

+check-format:
+	build_tools/format-diff.sh -c
+
+check-buck-targets:
+	buckifier/check_buck_targets.sh
+
 package:
 	bash build_tools/make_package.sh $(SHARED_MAJOR).$(SHARED_MINOR)

@@ -1222,6 +1266,9 @@ db_repl_stress: tools/db_repl_stress.o $(LIBOBJECTS) $(TESTUTIL)

 arena_test: memory/arena_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)
+	
+memkind_kmem_allocator_test: memory/memkind_kmem_allocator_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)	

 autovector_test: util/autovector_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)
@@ -1280,18 +1327,30 @@ histogram_test: monitoring/histogram_test.o $(LIBOBJECTS) $(TESTHARNESS)
 thread_local_test: util/thread_local_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+work_queue_test: util/work_queue_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 corruption_test: db/corruption_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 crc32c_test: util/crc32c_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+slice_test: util/slice_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 slice_transform_test: util/slice_transform_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 db_basic_test: db/db_basic_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+db_with_timestamp_basic_test: db/db_with_timestamp_basic_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
+db_with_timestamp_compaction_test: db/db_with_timestamp_compaction_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 db_encryption_test: db/db_encryption_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1301,7 +1360,10 @@ db_test: db/db_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 db_test2: db/db_test2.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-db_blob_index_test: db/db_blob_index_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+db_logical_block_size_cache_test: db/db_logical_block_size_cache_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
+db_blob_index_test: db/blob/db_blob_index_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 db_block_cache_test: db/db_block_cache_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
@@ -1355,7 +1417,7 @@ db_statistics_test: db/db_statistics_test.o db/db_test_util.o $(LIBOBJECTS) $(TE
 db_write_test: db/db_write_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-error_handler_test: db/error_handler_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
+error_handler_fs_test: db/error_handler_fs_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

 external_sst_file_basic_test: db/external_sst_file_basic_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
@@ -1468,6 +1530,9 @@ env_basic_test: env/env_basic_test.o $(LIBOBJECTS) $(TESTHARNESS)
 env_test: env/env_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+io_posix_test: env/io_posix_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 fault_injection_test: db/fault_injection_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1480,6 +1545,9 @@ delete_scheduler_test: file/delete_scheduler_test.o $(LIBOBJECTS) $(TESTHARNESS)
 filename_test: db/filename_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+random_access_file_reader_test: file/random_access_file_reader_test.o $(LIBOBJECTS) $(TESTHARNESS) $(TESTUTIL)
+	$(AM_LINK)
+
 file_reader_writer_test: util/file_reader_writer_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1501,6 +1569,9 @@ cleanable_test: table/cleanable_test.o $(LIBOBJECTS) $(TESTHARNESS)
 table_test: table/table_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+block_fetcher_test: table/block_fetcher_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 block_test: table/block_based/block_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1639,6 +1710,9 @@ write_callback_test: db/write_callback_test.o $(LIBOBJECTS) $(TESTHARNESS)
 heap_test: util/heap_test.o $(GTEST)
 	$(AM_LINK)

+transaction_lock_mgr_test: utilities/transactions/transaction_lock_mgr_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 transaction_test: utilities/transactions/transaction_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1705,6 +1779,18 @@ block_cache_tracer_test: trace_replay/block_cache_tracer_test.o trace_replay/blo
 block_cache_trace_analyzer_test: tools/block_cache_analyzer/block_cache_trace_analyzer_test.o tools/block_cache_analyzer/block_cache_trace_analyzer.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

+defer_test: util/defer_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
+blob_file_addition_test: db/blob/blob_file_addition_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
+blob_file_garbage_test: db/blob/blob_file_garbage_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
+timer_test: util/timer_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(AM_LINK)
+
 #-------------------------------------------------
 # make install related stuff
 INSTALL_PATH ?= /usr/local
@@ -1784,11 +1870,11 @@ SHA256_CMD = sha256sum
 ZLIB_VER ?= 1.2.11
 ZLIB_SHA256 ?= c3e5e9fdd5004dcb542feda5ee4f0ff0744628baf8ed2dd5d66f8ca1197cb1a1
 ZLIB_DOWNLOAD_BASE ?= http://zlib.net
-BZIP2_VER ?= 1.0.6
-BZIP2_SHA256 ?= a2848f34fcd5d6cf47def00461fcb528a0484d8edef8208d6d2e2909dc61d9cd
-BZIP2_DOWNLOAD_BASE ?= https://downloads.sourceforge.net/project/bzip2
-SNAPPY_VER ?= 1.1.7
-SNAPPY_SHA256 ?= 3dfa02e873ff51a11ee02b9ca391807f0c8ea0529a4924afa645fbf97163f9d4
+BZIP2_VER ?= 1.0.8
+BZIP2_SHA256 ?= ab5a03176ee106d3f0fa90e381da478ddae405918153cca248e682cd0c4a2269
+BZIP2_DOWNLOAD_BASE ?= https://sourceware.org/pub/bzip2
+SNAPPY_VER ?= 1.1.8
+SNAPPY_SHA256 ?= 16b677f07832a612b0836178db7f374e414f94657c138e6993cbfc5dcc58651f
 SNAPPY_DOWNLOAD_BASE ?= https://github.com/google/snappy/archive
 LZ4_VER ?= 1.9.2
 LZ4_SHA256 ?= 658ba6191fa44c92280d4aa2c271b0f4fbc0e34d249578dd05e50e76d0e5efcc
@@ -1834,7 +1920,7 @@ endif
 libz.a:
 	-rm -rf zlib-$(ZLIB_VER)
 ifeq (,$(wildcard ./zlib-$(ZLIB_VER).tar.gz))
-	curl --output zlib-$(ZLIB_VER).tar.gz -L ${ZLIB_DOWNLOAD_BASE}/zlib-$(ZLIB_VER).tar.gz
+	curl --fail --output zlib-$(ZLIB_VER).tar.gz --location ${ZLIB_DOWNLOAD_BASE}/zlib-$(ZLIB_VER).tar.gz
 endif
 	ZLIB_SHA256_ACTUAL=`$(SHA256_CMD) zlib-$(ZLIB_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(ZLIB_SHA256)" != "$$ZLIB_SHA256_ACTUAL" ]; then \
@@ -1848,7 +1934,7 @@ endif
 libbz2.a:
 	-rm -rf bzip2-$(BZIP2_VER)
 ifeq (,$(wildcard ./bzip2-$(BZIP2_VER).tar.gz))
-	curl --output bzip2-$(BZIP2_VER).tar.gz -L ${CURL_SSL_OPTS} ${BZIP2_DOWNLOAD_BASE}/bzip2-$(BZIP2_VER).tar.gz
+	curl --fail --output bzip2-$(BZIP2_VER).tar.gz --location ${CURL_SSL_OPTS} ${BZIP2_DOWNLOAD_BASE}/bzip2-$(BZIP2_VER).tar.gz
 endif
 	BZIP2_SHA256_ACTUAL=`$(SHA256_CMD) bzip2-$(BZIP2_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(BZIP2_SHA256)" != "$$BZIP2_SHA256_ACTUAL" ]; then \
@@ -1862,7 +1948,7 @@ endif
 libsnappy.a:
 	-rm -rf snappy-$(SNAPPY_VER)
 ifeq (,$(wildcard ./snappy-$(SNAPPY_VER).tar.gz))
-	curl --output snappy-$(SNAPPY_VER).tar.gz -L ${CURL_SSL_OPTS} ${SNAPPY_DOWNLOAD_BASE}/$(SNAPPY_VER).tar.gz
+	curl --fail --output snappy-$(SNAPPY_VER).tar.gz --location ${CURL_SSL_OPTS} ${SNAPPY_DOWNLOAD_BASE}/$(SNAPPY_VER).tar.gz
 endif
 	SNAPPY_SHA256_ACTUAL=`$(SHA256_CMD) snappy-$(SNAPPY_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(SNAPPY_SHA256)" != "$$SNAPPY_SHA256_ACTUAL" ]; then \
@@ -1877,7 +1963,7 @@ endif
 liblz4.a:
 	-rm -rf lz4-$(LZ4_VER)
 ifeq (,$(wildcard ./lz4-$(LZ4_VER).tar.gz))
-	curl --output lz4-$(LZ4_VER).tar.gz -L ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
+	curl --fail --output lz4-$(LZ4_VER).tar.gz --location ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
 endif
 	LZ4_SHA256_ACTUAL=`$(SHA256_CMD) lz4-$(LZ4_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(LZ4_SHA256)" != "$$LZ4_SHA256_ACTUAL" ]; then \
@@ -1891,7 +1977,7 @@ endif
 libzstd.a:
 	-rm -rf zstd-$(ZSTD_VER)
 ifeq (,$(wildcard ./zstd-$(ZSTD_VER).tar.gz))
-	curl --output zstd-$(ZSTD_VER).tar.gz -L ${CURL_SSL_OPTS} ${ZSTD_DOWNLOAD_BASE}/v$(ZSTD_VER).tar.gz
+	curl --fail --output zstd-$(ZSTD_VER).tar.gz --location ${CURL_SSL_OPTS} ${ZSTD_DOWNLOAD_BASE}/v$(ZSTD_VER).tar.gz
 endif
 	ZSTD_SHA256_ACTUAL=`$(SHA256_CMD) zstd-$(ZSTD_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(ZSTD_SHA256)" != "$$ZSTD_SHA256_ACTUAL" ]; then \
@@ -1961,35 +2047,35 @@ rocksdbjavastaticreleasedocker: rocksdbjavastatic rocksdbjavastaticdockerx86 roc

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

 rocksdbjavastaticdockerx86_64:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_x64-be --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos6_x64-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_x64-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos6_x64-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticdockerppc64le:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_ppc64le-be --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos7_ppc64le-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_ppc64le-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos7_ppc64le-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticdockerarm64v8:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_arm64v8-be --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos7_arm64v8-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_arm64v8-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:centos7_arm64v8-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

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

 rocksdbjavastaticdockerx86_64musl:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_x64-musl-be --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_x64-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_x64-musl-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_x64-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticdockerppc64lemusl:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_ppc64le-musl-be --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_ppc64le-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_ppc64le-musl-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_ppc64le-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticdockerarm64v8musl:
 	mkdir -p java/target
-	docker run --rm --name rocksdb_linux_arm64v8-musl-be --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_arm64v8-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh
+	docker run --rm --name rocksdb_linux_arm64v8-musl-be --attach stdin --attach stdout --attach stderr --volume $(HOME)/.m2:/root/.m2:ro --volume `pwd`:/rocksdb-host:ro --volume /rocksdb-local-build --volume `pwd`/java/target:/rocksdb-java-target --env DEBUG_LEVEL=$(DEBUG_LEVEL) evolvedbinary/rocksjava:alpine3_arm64v8-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh

 rocksdbjavastaticpublish: rocksdbjavastaticrelease rocksdbjavastaticpublishcentral

@@ -2055,7 +2141,7 @@ jtest_run:

 jtest: rocksdbjava
 	cd java;$(MAKE) sample;$(MAKE) test;
-	python tools/check_all_python.py # TODO peterd: find a better place for this check in CI targets
+	$(PYTHON) tools/check_all_python.py # TODO peterd: find a better place for this check in CI targets

 jdb_bench:
 	cd java;$(MAKE) db_bench;
@@ -24,7 +24,7 @@ The public interface is in `include/`.  Callers should not include or
 rely on the details of any other header files in this package.  Those
 internal APIs may be changed without warning.

-Design discussions are conducted in https://www.facebook.com/groups/rocksdb.dev/
+Design discussions are conducted in https://www.facebook.com/groups/rocksdb.dev/ and https://rocksdb.slack.com/

 ## License

@@ -26,14 +26,13 @@ ROCKSDB_EXTERNAL_DEPS = [
    ("lz4", None, "lz4"),
    ("zstd", None),
    ("tbb", None),
-    ("liburing", None, "uring"),
    ("googletest", None, "gtest"),
 ]

 ROCKSDB_OS_DEPS = [
    (
        "linux",
-        ["third-party//numa:numa"],
+        ["third-party//numa:numa", "third-party//liburing:uring"],
    ),
 ]

@@ -49,6 +48,7 @@ ROCKSDB_OS_PREPROCESSOR_FLAGS = [
            "-DROCKSDB_SCHED_GETCPU_PRESENT",
            "-DROCKSDB_IOURING_PRESENT",
            "-DHAVE_SSE42",
+            "-DLIBURING",
            "-DNUMA",
        ],
    ),
@@ -72,7 +72,6 @@ ROCKSDB_PREPROCESSOR_FLAGS = [
    "-DZSTD_STATIC_LINKING_ONLY",
    "-DGFLAGS=gflags",
    "-DTBB",
-    "-DLIBURING",

    # Added missing flags from output of build_detect_platform
    "-DROCKSDB_BACKTRACE",
@@ -113,10 +112,14 @@ ROCKSDB_OS_DEPS += ([(
 cpp_library(
    name = "rocksdb_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_garbage.cc",
+        "db/blob/blob_file_meta.cc",
        "db/builder.cc",
        "db/c.cc",
        "db/column_family.cc",
@@ -170,6 +173,7 @@ cpp_library(
        "db/trim_history_scheduler.cc",
        "db/version_builder.cc",
        "db/version_edit.cc",
+        "db/version_edit_handler.cc",
        "db/version_set.cc",
        "db/wal_manager.cc",
        "db/write_batch.cc",
@@ -201,6 +205,7 @@ cpp_library(
        "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",
@@ -226,16 +231,18 @@ cpp_library(
        "options/options.cc",
        "options/options_helper.cc",
        "options/options_parser.cc",
-        "options/options_sanity_check.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",
@@ -243,9 +250,14 @@ cpp_library(
        "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",
@@ -284,6 +296,7 @@ cpp_library(
        "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",
@@ -367,6 +380,7 @@ cpp_library(
        "db/db_test_util.cc",
        "table/mock_table.cc",
        "test_util/fault_injection_test_env.cc",
+        "test_util/fault_injection_test_fs.cc",
        "test_util/testharness.cc",
        "test_util/testutil.cc",
        "tools/block_cache_analyzer/block_cache_trace_analyzer.cc",
@@ -423,7 +437,10 @@ cpp_library(
    os_deps = ROCKSDB_OS_DEPS,
    os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
    preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
-    deps = [":rocksdb_lib"],
+    deps = [
+        ":rocksdb_lib",
+        ":rocksdb_test_lib",
+    ],
    external_deps = ROCKSDB_EXTERNAL_DEPS,
 )

@@ -477,6 +494,20 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "blob_file_addition_test",
+        "db/blob/blob_file_addition_test.cc",
+        "serial",
+        [],
+        [],
+    ],
+    [
+        "blob_file_garbage_test",
+        "db/blob/blob_file_garbage_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "block_based_filter_block_test",
        "table/block_based/block_based_filter_block_test.cc",
@@ -498,6 +529,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "block_fetcher_test",
+        "table/block_fetcher_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "block_test",
        "table/block_based/block_test.cc",
@@ -689,7 +727,7 @@ ROCKS_TESTS = [
    ],
    [
        "db_blob_index_test",
-        "db/db_blob_index_test.cc",
+        "db/blob/db_blob_index_test.cc",
        "serial",
        [],
        [],
@@ -785,6 +823,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "db_logical_block_size_cache_test",
+        "db/db_logical_block_size_cache_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "db_memtable_test",
        "db/db_memtable_test.cc",
@@ -890,6 +935,20 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "db_with_timestamp_basic_test",
+        "db/db_with_timestamp_basic_test.cc",
+        "serial",
+        [],
+        [],
+    ],
+    [
+        "db_with_timestamp_compaction_test",
+        "db/db_with_timestamp_compaction_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "db_write_test",
        "db/db_write_test.cc",
@@ -904,6 +963,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "defer_test",
+        "util/defer_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "delete_scheduler_test",
        "file/delete_scheduler_test.cc",
@@ -954,8 +1020,8 @@ ROCKS_TESTS = [
        [],
    ],
    [
-        "error_handler_test",
-        "db/error_handler_test.cc",
+        "error_handler_fs_test",
+        "db/error_handler_fs_test.cc",
        "serial",
        [],
        [],
@@ -1072,6 +1138,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "io_posix_test",
+        "env/io_posix_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "iostats_context_test",
        "monitoring/iostats_context_test.cc",
@@ -1114,6 +1187,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "memkind_kmem_allocator_test",
+        "memory/memkind_kmem_allocator_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "memory_test",
        "utilities/memory/memory_test.cc",
@@ -1247,6 +1327,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "random_access_file_reader_test",
+        "file/random_access_file_reader_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "random_test",
        "util/random_test.cc",
@@ -1310,6 +1397,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "slice_test",
+        "util/slice_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "slice_transform_test",
        "util/slice_transform_test.cc",
@@ -1387,6 +1481,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "timer_test",
+        "util/timer_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "trace_analyzer_test",
        "tools/trace_analyzer_test.cc",
@@ -1394,6 +1495,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "transaction_lock_mgr_test",
+        "utilities/transactions/transaction_lock_mgr_test.cc",
+        "parallel",
+        [],
+        [],
+    ],
    [
        "transaction_test",
        "utilities/transactions/transaction_test.cc",
@@ -1443,6 +1551,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "work_queue_test",
+        "util/work_queue_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "write_batch_test",
        "db/write_batch_test.cc",
@@ -106,3 +106,6 @@ LzLabs is using RocksDB as a storage engine in their multi-database distributed
 ## Nebula Graph

 [Nebula Graph](https://github.com/vesoft-inc/nebula) is a distributed, scalable, lightning-fast, open source graph database capable of hosting super large scale graphs with dozens of billions of vertices (nodes) and trillions of edges, with milliseconds of latency.
+
+## YugabyteDB
+[YugabyteDB](https://www.yugabyte.com/) is an open source, high performance, distributed SQL database that uses RocksDB as its storage layer. For more information, please see https://github.com/yugabyte/yugabyte-db/.
@@ -1,6 +1,6 @@
 version: 1.0.{build}

-image: Visual Studio 2017
+image: Visual Studio 2019

 environment:
  JAVA_HOME: C:\Program Files\Java\jdk1.8.0
@@ -24,22 +24,32 @@ 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%
  - echo "Building Snappy dependency..."
  - cd %THIRDPARTY_HOME%
-  - curl -fsSL -o snappy-1.1.7.zip https://github.com/google/snappy/archive/1.1.7.zip
+  - curl --fail --silent --show-error --output snappy-1.1.7.zip --location https://github.com/google/snappy/archive/1.1.7.zip
  - unzip snappy-1.1.7.zip
  - cd snappy-1.1.7
  - mkdir build
  - cd build
-  - cmake -G "%CMAKE_GENERATOR%" ..
+  - if DEFINED CMAKE_PLATEFORM_NAME (set "PLATEFORM_OPT=-A %CMAKE_PLATEFORM_NAME%")
+  - cmake .. -G "%CMAKE_GENERATOR%" %PLATEFORM_OPT%
  - msbuild Snappy.sln /p:Configuration=Debug /p:Platform=x64
  - msbuild Snappy.sln /p:Configuration=Release /p:Platform=x64
  - echo "Building LZ4 dependency..."
  - cd %THIRDPARTY_HOME%
-  - curl -fsSL -o lz4-1.8.3.zip https://github.com/lz4/lz4/archive/v1.8.3.zip
+  - curl --fail --silent --show-error --output lz4-1.8.3.zip --location https://github.com/lz4/lz4/archive/v1.8.3.zip
  - unzip lz4-1.8.3.zip
  - cd lz4-1.8.3\visual\VS2010
  - ps: $CMD="$Env:DEV_ENV"; & $CMD lz4.sln /upgrade
@@ -47,7 +57,7 @@ install:
  - msbuild lz4.sln /p:Configuration=Release /p:Platform=x64
  - echo "Building ZStd dependency..."
  - cd %THIRDPARTY_HOME%
-  - curl -fsSL -o zstd-1.4.0.zip https://github.com/facebook/zstd/archive/v1.4.0.zip
+  - curl --fail --silent --show-error --output zstd-1.4.0.zip --location https://github.com/facebook/zstd/archive/v1.4.0.zip
  - unzip zstd-1.4.0.zip
  - cd zstd-1.4.0\build\VS2010
  - ps: $CMD="$Env:DEV_ENV"; & $CMD zstd.sln /upgrade
@@ -57,7 +67,8 @@ install:
 before_build:
  - md %APPVEYOR_BUILD_FOLDER%\build
  - cd %APPVEYOR_BUILD_FOLDER%\build
-  - cmake -G "%CMAKE_GENERATOR%" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DLZ4=1 -DZSTD=1 -DXPRESS=1 -DJNI=1 ..
+  - 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
  - cd ..

 build:
@@ -68,7 +79,7 @@ build:
 test:

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

 on_failure:
  - cmd: 7z a build-failed.zip %APPVEYOR_BUILD_FOLDER%\build\ && appveyor PushArtifact build-failed.zip
@@ -173,7 +173,7 @@ def generate_targets(repo_path, deps_map):
        src_mk.get("ANALYZER_LIB_SOURCES", [])
        + src_mk.get('STRESS_LIB_SOURCES', [])
        + ["test_util/testutil.cc"],
-        [":rocksdb_lib"])
+        [":rocksdb_lib", ":rocksdb_test_lib"])

    print("Extra dependencies:\n{0}".format(str(deps_map)))
    # test for every test we found in the Makefile
@@ -0,0 +1,31 @@
+#!/usr/bin/env bash
+# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+# If clang_format_diff.py command is not specfied, we assume we are able to
+# access directly without any path.
+
+TGT_DIFF=`git diff TARGETS | head -n 1`
+
+if [ ! -z "$TGT_DIFF" ]
+then
+  echo "TARGETS file has uncommitted changes. Skip this check."
+  exit 0
+fi
+
+echo Backup original TARGETS file.
+
+cp TARGETS TARGETS.bkp
+
+python buckifier/buckify_rocksdb.py
+
+TGT_DIFF=`git diff TARGETS | head -n 1`
+
+if [ -z "$TGT_DIFF" ]
+then
+  mv TARGETS.bkp TARGETS
+  exit 0
+else
+  echo "Please run 'python buckifier/buckify_rocksdb.py' to update TARGETS file."
+  echo "Do not manually update TARGETS file."
+  mv TARGETS.bkp TARGETS
+  exit 1
+fi
@@ -32,14 +32,13 @@ ROCKSDB_EXTERNAL_DEPS = [
    ("lz4", None, "lz4"),
    ("zstd", None),
    ("tbb", None),
-    ("liburing", None, "uring"),
    ("googletest", None, "gtest"),
 ]

 ROCKSDB_OS_DEPS = [
    (
        "linux",
-        ["third-party//numa:numa"],
+        ["third-party//numa:numa", "third-party//liburing:uring"],
    ),
 ]

@@ -55,6 +54,7 @@ ROCKSDB_OS_PREPROCESSOR_FLAGS = [
            "-DROCKSDB_SCHED_GETCPU_PRESENT",
            "-DROCKSDB_IOURING_PRESENT",
            "-DHAVE_SSE42",
+            "-DLIBURING",
            "-DNUMA",
        ],
    ),
@@ -78,7 +78,6 @@ ROCKSDB_PREPROCESSOR_FLAGS = [
    "-DZSTD_STATIC_LINKING_ONLY",
    "-DGFLAGS=gflags",
    "-DTBB",
-    "-DLIBURING",

    # Added missing flags from output of build_detect_platform
    "-DROCKSDB_BACKTRACE",
@@ -9,6 +9,7 @@
 #   PLATFORM_LDFLAGS            Linker flags
 #   JAVA_LDFLAGS                Linker flags for RocksDBJava
 #   JAVA_STATIC_LDFLAGS         Linker flags for RocksDBJava static build
+#   JAVAC_ARGS                  Arguments for javac
 #   PLATFORM_SHARED_EXT         Extension for shared libraries
 #   PLATFORM_SHARED_LDFLAGS     Flags for building shared library
 #   PLATFORM_SHARED_CFLAGS      Flags for compiling objects for shared library
@@ -27,6 +28,7 @@
 #       -DZSTD                      if the ZSTD library is present
 #       -DNUMA                      if the NUMA library is present
 #       -DTBB                       if the TBB library is present
+#       -DMEMKIND                   if the memkind library is present
 #
 # Using gflags in rocksdb:
 # Our project depends on gflags, which requires users to take some extra steps
@@ -150,18 +152,20 @@ case "$TARGET_OS" in
            PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -latomic"
        fi
        PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lpthread -lrt"
-        # check for liburing
-        $CXX $CFLAGS -x c++ - -luring -o /dev/null 2>/dev/null  <<EOF
-          #include <liburing.h>
-          int main() {
-            struct io_uring ring;
-            io_uring_queue_init(1, &ring, 0);
-            return 0;
-          }
+        if test $ROCKSDB_USE_IO_URING; then
+            # check for liburing
+            $CXX $CFLAGS -x c++ - -luring -o /dev/null 2>/dev/null  <<EOF
+              #include <liburing.h>
+              int main() {
+                struct io_uring ring;
+                io_uring_queue_init(1, &ring, 0);
+                return 0;
+              }
 EOF
-        if [ "$?" = 0 ]; then
-          PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -luring"
-          COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_IOURING_PRESENT"
+            if [ "$?" = 0 ]; then
+                PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -luring"
+                COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_IOURING_PRESENT"
+            fi
        fi
        if test -z "$USE_FOLLY_DISTRIBUTED_MUTEX"; then
          USE_FOLLY_DISTRIBUTED_MUTEX=1
@@ -237,10 +241,15 @@ esac
 PLATFORM_CXXFLAGS="$PLATFORM_CXXFLAGS ${CXXFLAGS}"
 JAVA_LDFLAGS="$PLATFORM_LDFLAGS"
 JAVA_STATIC_LDFLAGS="$PLATFORM_LDFLAGS"
+JAVAC_ARGS="-source 7"

 if [ "$CROSS_COMPILE" = "true" -o "$FBCODE_BUILD" = "true" ]; then
    # Cross-compiling; do not try any compilation tests.
    # Also don't need any compilation tests if compiling on fbcode
+    if [ "$FBCODE_BUILD" = "true" ]; then
+      # Enable backtrace on fbcode since the necessary libraries are present
+      COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_BACKTRACE"
+    fi
    true
 else
    if ! test $ROCKSDB_DISABLE_FALLOCATE; then
@@ -421,7 +430,23 @@ EOF
            COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_MALLOC_USABLE_SIZE"
        fi
    fi
-
+    
+    if ! test $ROCKSDB_DISABLE_MEMKIND; then
+        # Test whether memkind library is installed
+        $CXX $CFLAGS $COMMON_FLAGS -lmemkind -x c++ - -o /dev/null 2>/dev/null  <<EOF
+          #include <memkind.h>
+          int main() {
+            memkind_malloc(MEMKIND_DAX_KMEM, 1024);
+            return 0;
+          }
+EOF
+        if [ "$?" = 0 ]; then
+            COMMON_FLAGS="$COMMON_FLAGS -DMEMKIND"
+            PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lmemkind"
+            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
@@ -506,6 +531,20 @@ EOF
        fi
    fi

+    if ! test $ROCKSDB_DISABLE_AUXV_GETAUXVAL; then
+        # Test whether getauxval is supported
+        $CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null  <<EOF
+          #include <sys/auxv.h>
+          int main() {
+            uint64_t auxv = getauxval(AT_HWCAP);
+            (void)auxv;
+          }
+EOF
+        if [ "$?" = 0 ]; then
+            COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_AUXV_GETAUXVAL_PRESENT"
+        fi
+    fi
+
    if ! test $ROCKSDB_DISABLE_ALIGNED_NEW; then
        # Test whether c++17 aligned-new is supported
        $CXX $PLATFORM_CXXFLAGS -faligned-new -x c++ - -o /dev/null 2>/dev/null <<EOF
@@ -666,6 +705,7 @@ EOF
  fi
 fi

+
 if [ "$FBCODE_BUILD" != "true" -a "$PLATFORM" = OS_LINUX ]; then
  $CXX $COMMON_FLAGS $PLATFORM_SHARED_CFLAGS -x c++ -c - -o test_dl.o 2>/dev/null <<EOF
  void dummy_func() {}
@@ -694,6 +734,7 @@ echo "PLATFORM=$PLATFORM" >> "$OUTPUT"
 echo "PLATFORM_LDFLAGS=$PLATFORM_LDFLAGS" >> "$OUTPUT"
 echo "JAVA_LDFLAGS=$JAVA_LDFLAGS" >> "$OUTPUT"
 echo "JAVA_STATIC_LDFLAGS=$JAVA_STATIC_LDFLAGS" >> "$OUTPUT"
+echo "JAVAC_ARGS=$JAVAC_ARGS" >> "$OUTPUT"
 echo "VALGRIND_VER=$VALGRIND_VER" >> "$OUTPUT"
 echo "PLATFORM_CCFLAGS=$PLATFORM_CCFLAGS" >> "$OUTPUT"
 echo "PLATFORM_CXXFLAGS=$PLATFORM_CXXFLAGS" >> "$OUTPUT"
@@ -2,6 +2,30 @@
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
 # If clang_format_diff.py command is not specfied, we assume we are able to
 # access directly without any path.
+
+print_usage () {
+  echo "Usage:"
+  echo "format-diff.sh [OPTIONS]"
+  echo "-c: check only."
+  echo "-h: print this message."
+}
+
+while getopts ':ch' OPTION; do
+  case "$OPTION" in
+    c)
+      CHECK_ONLY=1
+      ;;
+    h)
+      print_usage
+      exit 1
+      ;;
+    ?)
+      print_usage
+      exit 1
+      ;;
+  esac
+done
+
 if [ -z $CLANG_FORMAT_DIFF ]
 then
 CLANG_FORMAT_DIFF="clang-format-diff.py"
@@ -10,18 +34,28 @@ fi
 # Check clang-format-diff.py
 if ! which $CLANG_FORMAT_DIFF &> /dev/null
 then
-  echo "You didn't have clang-format-diff.py and/or clang-format available in your computer!"
-  echo "You can download clang-format-diff.py by running: "
-  echo "    curl --location http://goo.gl/iUW1u2 -o ${CLANG_FORMAT_DIFF}"
-  echo "You can download clang-format by running:"
-  echo "    brew install clang-format"
-  echo "  Or"
-  echo "    apt install clang-format"
-  echo "  This might work too:"
-  echo "    yum install git-clang-format"
-  echo "Then, move both files (i.e. ${CLANG_FORMAT_DIFF} and clang-format) to some directory within PATH=${PATH}"
-  echo "and make sure ${CLANG_FORMAT_DIFF} is executable."
-  exit 128
+  if [ ! -f ./clang-format-diff.py ]
+  then
+    echo "You didn't have clang-format-diff.py and/or clang-format available in your computer!"
+    echo "You can download clang-format-diff.py by running: "
+    echo "    curl --location http://goo.gl/iUW1u2 -o ${CLANG_FORMAT_DIFF}"
+    echo "You can download clang-format by running:"
+    echo "    brew install clang-format"
+    echo "  Or"
+    echo "    apt install clang-format"
+    echo "  This might work too:"
+    echo "    yum install git-clang-format"
+    echo "Then, move both files (i.e. ${CLANG_FORMAT_DIFF} and clang-format) to some directory within PATH=${PATH}"
+    echo "and make sure ${CLANG_FORMAT_DIFF} is executable."
+    exit 128
+  else
+    if [ -x ./clang-format-diff.py ]
+    then
+      PATH=$PATH:.
+    else
+      CLANG_FORMAT_DIFF="python ./clang-format-diff.py"
+    fi
+  fi
 fi

 # Check argparse, a library that clang-format-diff.py requires.
@@ -87,6 +121,10 @@ if [ -z "$diffs" ]
 then
  echo "Nothing needs to be reformatted!"
  exit 0
+elif [ $CHECK_ONLY ]
+then
+  echo "Your change has unformatted code. Please run make format!"
+  exit 1
 fi

 # Highlight the insertion/deletion from the clang-format-diff.py's output
@@ -121,7 +159,7 @@ if [ -z "$uncommitted_code" ]
 then
  git diff -U0 "$FORMAT_UPSTREAM_MERGE_BASE" | $CLANG_FORMAT_DIFF -i -p 1
 else
-  git diff -U0 HEAD^ | $CLANG_FORMAT_DIFF -i -p 1
+  git diff -U0 HEAD | $CLANG_FORMAT_DIFF -i -p 1
 fi
 echo "Files reformatted!"

@@ -5801,7 +5801,7 @@ sub workdir {
 		    . "-" . $self->seq();
 	    } else {
 		$workdir = $opt::workdir;
-		# Rsync treats /./ special. We dont want that
+		# Rsync treats /./ special. We don't want that
 		$workdir =~ s:/\./:/:g; # Remove /./
 		$workdir =~ s:/+$::; # Remove ending / if any
 		$workdir =~ s:^\./::g; # Remove starting ./ if any
@@ -378,7 +378,7 @@ function send_to_ods {
    echo >&2 "ERROR: Key $key doesn't have a value."
    return
  fi
-  curl -s "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build$git_br&key=$key&value=$value" \
+  curl --silent "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build$git_br&key=$key&value=$value" \
    --connect-timeout 60
 }

@@ -414,7 +414,8 @@ STRESS_CRASH_TEST_COMMANDS="[
                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test || $CONTRUN_NAME=crash_test $TASK_CREATION_TOOL',
                'user':'root',
                $PARSER
-            }
+            },
+            $UPLOAD_DB_DIR,
        ],
        $REPORT
    }
@@ -542,6 +543,7 @@ ASAN_CRASH_TEST_COMMANDS="[
                'user':'root',
                $PARSER
            },
+            $UPLOAD_DB_DIR,
        ],
        $REPORT
    }
@@ -634,6 +636,7 @@ UBSAN_CRASH_TEST_COMMANDS="[
                'user':'root',
                $PARSER
            },
+            $UPLOAD_DB_DIR,
        ],
        $REPORT
    }
@@ -751,6 +754,7 @@ TSAN_CRASH_TEST_COMMANDS="[
                'user':'root',
                $PARSER
            },
+            $UPLOAD_DB_DIR,
        ],
        $REPORT
    }
@@ -814,6 +818,8 @@ run_format_compatible()
  rm -rf /dev/shm/rocksdb
  mkdir /dev/shm/rocksdb

+  export https_proxy="fwdproxy:8080"
+  
  tools/check_format_compatible.sh
 }

@@ -880,7 +886,7 @@ run_regression()

  # parameters: $1 -- key, $2 -- value
  function send_size_to_ods {
-    curl -s "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build&key=rocksdb.build_size.$1&value=$2" \
+    curl --silent "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build&key=rocksdb.build_size.$1&value=$2" \
      --connect-timeout 60
  }

@@ -1059,5 +1065,6 @@ case $1 in
    ;;
  *)
    echo "Invalid determinator command"
+    exit 1
    ;;
 esac
@@ -1,9 +1,9 @@
 #!/bin/bash
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
-set -e
+set -ex

-ROCKSDB_VERSION="5.10.3"
-ZSTD_VERSION="1.1.3"
+ROCKSDB_VERSION="6.7.3"
+ZSTD_VERSION="1.4.4"

 echo "This script configures CentOS with everything needed to build and run RocksDB"

@@ -40,5 +40,6 @@ cd /usr/local/rocksdb
 chown -R vagrant:vagrant /usr/local/rocksdb/
 sudo -u vagrant make static_lib
 cd examples/
-sudo -u vagrant make all
-sudo -u vagrant ./c_simple_example
+sudo -u vagrant LD_LIBRARY_PATH=/usr/local/lib/ make all
+sudo -u vagrant LD_LIBRARY_PATH=/usr/local/lib/ ./c_simple_example
+
@@ -0,0 +1,41 @@
+//  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).
+//
+// 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 "rocksdb/cache.h"
+
+#include "cache/lru_cache.h"
+#include "options/options_helper.h"
+#include "util/string_util.h"
+
+namespace ROCKSDB_NAMESPACE {
+Status Cache::CreateFromString(const ConfigOptions& /*opts*/,
+                               const std::string& value,
+                               std::shared_ptr<Cache>* result) {
+  Status status;
+  std::shared_ptr<Cache> cache;
+  if (value.find('=') == std::string::npos) {
+    cache = NewLRUCache(ParseSizeT(value));
+  } else {
+#ifndef ROCKSDB_LITE
+    LRUCacheOptions cache_opts;
+    if (!ParseOptionHelper(reinterpret_cast<char*>(&cache_opts),
+                           OptionType::kLRUCacheOptions, value)) {
+      status = Status::InvalidArgument("Invalid cache options");
+    }
+    cache = NewLRUCache(cache_opts);
+#else
+    status = Status::NotSupported("Cannot load cache in LITE mode ", value);
+#endif  //! ROCKSDB_LITE
+  }
+  if (status.ok()) {
+    result->swap(cache);
+  }
+  return status;
+}
+}  // namespace ROCKSDB_NAMESPACE
@@ -14,56 +14,63 @@ int main() {
 #include <stdio.h>
 #include <sys/types.h>
 #include <cinttypes>
+#include <limits>

 #include "port/port.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
+#include "util/coding.h"
 #include "util/gflags_compat.h"
+#include "util/hash.h"
 #include "util/mutexlock.h"
 #include "util/random.h"

 using GFLAGS_NAMESPACE::ParseCommandLineFlags;

-static const uint32_t KB = 1024;
+static constexpr uint32_t KiB = uint32_t{1} << 10;
+static constexpr uint32_t MiB = KiB << 10;
+static constexpr uint64_t GiB = MiB << 10;

-DEFINE_int32(threads, 16, "Number of concurrent threads to run.");
-DEFINE_int64(cache_size, 8 * KB * KB,
-             "Number of bytes to use as a cache of uncompressed data.");
-DEFINE_int32(num_shard_bits, 4, "shard_bits.");
+DEFINE_uint32(threads, 16, "Number of concurrent threads to run.");
+DEFINE_uint64(cache_size, 1 * GiB,
+              "Number of bytes to use as a cache of uncompressed data.");
+DEFINE_uint32(num_shard_bits, 6, "shard_bits.");

-DEFINE_int64(max_key, 1 * KB * KB * KB, "Max number of key to place in cache");
-DEFINE_uint64(ops_per_thread, 1200000, "Number of operations per thread.");
+DEFINE_double(resident_ratio, 0.25,
+              "Ratio of keys fitting in cache to keyspace.");
+DEFINE_uint64(ops_per_thread, 0,
+              "Number of operations per thread. (Default: 5 * keyspace size)");
+DEFINE_uint32(value_bytes, 8 * KiB, "Size of each value added.");

-DEFINE_bool(populate_cache, false, "Populate cache before operations");
-DEFINE_int32(insert_percent, 40,
-             "Ratio of insert to total workload (expressed as a percentage)");
-DEFINE_int32(lookup_percent, 50,
-             "Ratio of lookup to total workload (expressed as a percentage)");
-DEFINE_int32(erase_percent, 10,
-             "Ratio of erase to total workload (expressed as a percentage)");
+DEFINE_uint32(skew, 5, "Degree of skew in key selection");
+DEFINE_bool(populate_cache, true, "Populate cache before operations");
+
+DEFINE_uint32(lookup_insert_percent, 87,
+              "Ratio of lookup (+ insert on not found) to total workload "
+              "(expressed as a percentage)");
+DEFINE_uint32(insert_percent, 2,
+              "Ratio of insert to total workload (expressed as a percentage)");
+DEFINE_uint32(lookup_percent, 10,
+              "Ratio of lookup to total workload (expressed as a percentage)");
+DEFINE_uint32(erase_percent, 1,
+              "Ratio of erase to total workload (expressed as a percentage)");

 DEFINE_bool(use_clock_cache, false, "");

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CacheBench;
 namespace {
-void deleter(const Slice& /*key*/, void* value) {
-    delete reinterpret_cast<char *>(value);
-}
-
 // State shared by all concurrent executions of the same benchmark.
 class SharedState {
 public:
  explicit SharedState(CacheBench* cache_bench)
      : cv_(&mu_),
-        num_threads_(FLAGS_threads),
        num_initialized_(0),
        start_(false),
        num_done_(0),
-        cache_bench_(cache_bench) {
-  }
+        cache_bench_(cache_bench) {}

  ~SharedState() {}

@@ -87,13 +94,9 @@ class SharedState {
    num_done_++;
  }

-  bool AllInitialized() const {
-    return num_initialized_ >= num_threads_;
-  }
+  bool AllInitialized() const { return num_initialized_ >= FLAGS_threads; }

-  bool AllDone() const {
-    return num_done_ >= num_threads_;
-  }
+  bool AllDone() const { return num_done_ >= FLAGS_threads; }

  void SetStart() {
    start_ = true;
@@ -107,7 +110,6 @@ class SharedState {
  port::Mutex mu_;
  port::CondVar cv_;

-  const uint64_t num_threads_;
  uint64_t num_initialized_;
  bool start_;
  uint64_t num_done_;
@@ -118,17 +120,69 @@ class SharedState {
 // Per-thread state for concurrent executions of the same benchmark.
 struct ThreadState {
  uint32_t tid;
-  Random rnd;
+  Random64 rnd;
  SharedState* shared;

  ThreadState(uint32_t index, SharedState* _shared)
      : tid(index), rnd(1000 + index), shared(_shared) {}
 };
+
+struct KeyGen {
+  char key_data[27];
+
+  Slice GetRand(Random64& rnd, uint64_t max_key) {
+    uint64_t raw = rnd.Next();
+    // Skew according to setting
+    for (uint32_t i = 0; i < FLAGS_skew; ++i) {
+      raw = std::min(raw, rnd.Next());
+    }
+    uint64_t key = fastrange64(raw, max_key);
+    // Variable size and alignment
+    size_t off = key % 8;
+    key_data[0] = char{42};
+    EncodeFixed64(key_data + 1, key);
+    key_data[9] = char{11};
+    EncodeFixed64(key_data + 10, key);
+    key_data[18] = char{4};
+    EncodeFixed64(key_data + 19, key);
+    return Slice(&key_data[off], sizeof(key_data) - off);
+  }
+};
+
+char* createValue(Random64& rnd) {
+  char* rv = new char[FLAGS_value_bytes];
+  // Fill with some filler data, and take some CPU time
+  for (uint32_t i = 0; i < FLAGS_value_bytes; i += 8) {
+    EncodeFixed64(rv + i, rnd.Next());
+  }
+  return rv;
+}
+
+void deleter(const Slice& /*key*/, void* value) {
+  delete[] static_cast<char*>(value);
+}
 }  // namespace

 class CacheBench {
+  static constexpr uint64_t kHundredthUint64 =
+      std::numeric_limits<uint64_t>::max() / 100U;
+
 public:
-  CacheBench() : num_threads_(FLAGS_threads) {
+  CacheBench()
+      : max_key_(static_cast<uint64_t>(FLAGS_cache_size / FLAGS_resident_ratio /
+                                       FLAGS_value_bytes)),
+        lookup_insert_threshold_(kHundredthUint64 *
+                                 FLAGS_lookup_insert_percent),
+        insert_threshold_(lookup_insert_threshold_ +
+                          kHundredthUint64 * FLAGS_insert_percent),
+        lookup_threshold_(insert_threshold_ +
+                          kHundredthUint64 * FLAGS_lookup_percent),
+        erase_threshold_(lookup_threshold_ +
+                         kHundredthUint64 * FLAGS_erase_percent) {
+    if (erase_threshold_ != 100U * kHundredthUint64) {
+      fprintf(stderr, "Percentages must add to 100.\n");
+      exit(1);
+    }
    if (FLAGS_use_clock_cache) {
      cache_ = NewClockCache(FLAGS_cache_size, FLAGS_num_shard_bits);
      if (!cache_) {
@@ -138,30 +192,31 @@ class CacheBench {
    } else {
      cache_ = NewLRUCache(FLAGS_cache_size, FLAGS_num_shard_bits);
    }
+    if (FLAGS_ops_per_thread == 0) {
+      FLAGS_ops_per_thread = 5 * max_key_;
+    }
  }

  ~CacheBench() {}

  void PopulateCache() {
-    Random rnd(1);
-    for (int64_t i = 0; i < FLAGS_cache_size; i++) {
-      uint64_t rand_key = rnd.Next() % FLAGS_max_key;
-      // Cast uint64* to be char*, data would be copied to cache
-      Slice key(reinterpret_cast<char*>(&rand_key), 8);
-      // do insert
-      cache_->Insert(key, new char[10], 1, &deleter);
+    Random64 rnd(1);
+    KeyGen keygen;
+    for (uint64_t i = 0; i < 2 * FLAGS_cache_size; i += FLAGS_value_bytes) {
+      cache_->Insert(keygen.GetRand(rnd, max_key_), createValue(rnd),
+                     FLAGS_value_bytes, &deleter);
    }
  }

  bool Run() {
-    rocksdb::Env* env = rocksdb::Env::Default();
+    ROCKSDB_NAMESPACE::Env* env = ROCKSDB_NAMESPACE::Env::Default();

    PrintEnv();
    SharedState shared(this);
-    std::vector<ThreadState*> threads(num_threads_);
-    for (uint32_t i = 0; i < num_threads_; i++) {
-      threads[i] = new ThreadState(i, &shared);
-      env->StartThread(ThreadBody, threads[i]);
+    std::vector<std::unique_ptr<ThreadState> > threads(FLAGS_threads);
+    for (uint32_t i = 0; i < FLAGS_threads; i++) {
+      threads[i].reset(new ThreadState(i, &shared));
+      env->StartThread(ThreadBody, threads[i].get());
    }
    {
      MutexLock l(shared.GetMutex());
@@ -192,10 +247,15 @@ class CacheBench {

 private:
  std::shared_ptr<Cache> cache_;
-  uint32_t num_threads_;
+  const uint64_t max_key_;
+  // Cumulative thresholds in the space of a random uint64_t
+  const uint64_t lookup_insert_threshold_;
+  const uint64_t insert_threshold_;
+  const uint64_t lookup_threshold_;
+  const uint64_t erase_threshold_;

  static void ThreadBody(void* v) {
-    ThreadState* thread = reinterpret_cast<ThreadState*>(v);
+    ThreadState* thread = static_cast<ThreadState*>(v);
    SharedState* shared = thread->shared;

    {
@@ -220,44 +280,82 @@ class CacheBench {
  }

  void OperateCache(ThreadState* thread) {
+    // To use looked-up values
+    uint64_t result = 0;
+    // To hold handles for a non-trivial amount of time
+    Cache::Handle* handle = nullptr;
+    KeyGen gen;
    for (uint64_t i = 0; i < FLAGS_ops_per_thread; i++) {
-      uint64_t rand_key = thread->rnd.Next() % FLAGS_max_key;
-      // Cast uint64* to be char*, data would be copied to cache
-      Slice key(reinterpret_cast<char*>(&rand_key), 8);
-      int32_t prob_op = thread->rnd.Uniform(100);
-      if (prob_op >= 0 && prob_op < FLAGS_insert_percent) {
-        // do insert
-        cache_->Insert(key, new char[10], 1, &deleter);
-      } else if (prob_op -= FLAGS_insert_percent &&
-                 prob_op < FLAGS_lookup_percent) {
-        // do lookup
-        auto handle = cache_->Lookup(key);
+      Slice key = gen.GetRand(thread->rnd, max_key_);
+      uint64_t random_op = thread->rnd.Next();
+      if (random_op < lookup_insert_threshold_) {
        if (handle) {
          cache_->Release(handle);
+          handle = nullptr;
        }
-      } else if (prob_op -= FLAGS_lookup_percent &&
-                 prob_op < FLAGS_erase_percent) {
+        // do lookup
+        handle = cache_->Lookup(key);
+        if (handle) {
+          // do something with the data
+          result += NPHash64(static_cast<char*>(cache_->Value(handle)),
+                             FLAGS_value_bytes);
+        } else {
+          // do insert
+          cache_->Insert(key, createValue(thread->rnd), FLAGS_value_bytes,
+                         &deleter, &handle);
+        }
+      } else if (random_op < insert_threshold_) {
+        if (handle) {
+          cache_->Release(handle);
+          handle = nullptr;
+        }
+        // do insert
+        cache_->Insert(key, createValue(thread->rnd), FLAGS_value_bytes,
+                       &deleter, &handle);
+      } else if (random_op < lookup_threshold_) {
+        if (handle) {
+          cache_->Release(handle);
+          handle = nullptr;
+        }
+        // do lookup
+        handle = cache_->Lookup(key);
+        if (handle) {
+          // do something with the data
+          result += NPHash64(static_cast<char*>(cache_->Value(handle)),
+                             FLAGS_value_bytes);
+        }
+      } else if (random_op < erase_threshold_) {
        // do erase
        cache_->Erase(key);
+      } else {
+        // Should be extremely unlikely (noop)
+        assert(random_op >= kHundredthUint64 * 100U);
      }
    }
+    if (handle) {
+      cache_->Release(handle);
+      handle = nullptr;
+    }
  }

  void PrintEnv() const {
    printf("RocksDB version     : %d.%d\n", kMajorVersion, kMinorVersion);
-    printf("Number of threads   : %d\n", FLAGS_threads);
+    printf("Number of threads   : %u\n", FLAGS_threads);
    printf("Ops per thread      : %" PRIu64 "\n", FLAGS_ops_per_thread);
    printf("Cache size          : %" PRIu64 "\n", FLAGS_cache_size);
-    printf("Num shard bits      : %d\n", FLAGS_num_shard_bits);
-    printf("Max key             : %" PRIu64 "\n", FLAGS_max_key);
-    printf("Populate cache      : %d\n", FLAGS_populate_cache);
-    printf("Insert percentage   : %d%%\n", FLAGS_insert_percent);
-    printf("Lookup percentage   : %d%%\n", FLAGS_lookup_percent);
-    printf("Erase percentage    : %d%%\n", FLAGS_erase_percent);
+    printf("Num shard bits      : %u\n", FLAGS_num_shard_bits);
+    printf("Max key             : %" PRIu64 "\n", max_key_);
+    printf("Resident ratio      : %g\n", FLAGS_resident_ratio);
+    printf("Skew degree         : %u\n", FLAGS_skew);
+    printf("Populate cache      : %d\n", int{FLAGS_populate_cache});
+    printf("Lookup+Insert pct   : %u%%\n", FLAGS_lookup_insert_percent);
+    printf("Insert percentage   : %u%%\n", FLAGS_insert_percent);
+    printf("Lookup percentage   : %u%%\n", FLAGS_lookup_percent);
+    printf("Erase percentage    : %u%%\n", FLAGS_erase_percent);
    printf("----------------------------\n");
  }
 };
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ParseCommandLineFlags(&argc, &argv, true);
@@ -267,9 +365,11 @@ int main(int argc, char** argv) {
    exit(1);
  }

-  rocksdb::CacheBench bench;
+  ROCKSDB_NAMESPACE::CacheBench bench;
  if (FLAGS_populate_cache) {
    bench.PopulateCache();
+    printf("Population complete\n");
+    printf("----------------------------\n");
  }
  if (bench.Run()) {
    return 0;
@@ -20,7 +20,7 @@
 #include "util/coding.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // Conversions between numeric keys/values and the types expected by Cache.
 static std::string EncodeKey(int k) {
@@ -765,7 +765,7 @@ INSTANTIATE_TEST_CASE_P(CacheTestInstance, CacheTest, testing::Values(kLRU));
 #endif  // SUPPORT_CLOCK_CACHE
 INSTANTIATE_TEST_CASE_P(CacheTestInstance, LRUCacheTest, testing::Values(kLRU));

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -11,7 +11,7 @@

 #ifndef SUPPORT_CLOCK_CACHE

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 std::shared_ptr<Cache> NewClockCache(
    size_t /*capacity*/, int /*num_shard_bits*/, bool /*strict_capacity_limit*/,
@@ -20,7 +20,7 @@ std::shared_ptr<Cache> NewClockCache(
  return nullptr;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #else

@@ -41,7 +41,7 @@ std::shared_ptr<Cache> NewClockCache(
 #include "util/autovector.h"
 #include "util/mutexlock.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 namespace {

@@ -341,7 +341,8 @@ class ClockCacheShard final : public CacheShard {
  CacheHandle* Insert(const Slice& key, uint32_t hash, void* value,
                      size_t change,
                      void (*deleter)(const Slice& key, void* value),
-                      bool hold_reference, CleanupContext* context);
+                      bool hold_reference, CleanupContext* context,
+                      bool* overwritten);

  // Guards list_, head_, and recycle_. In addition, updating table_ also has
  // to hold the mutex, to avoid the cache being in inconsistent state.
@@ -564,7 +565,8 @@ void ClockCacheShard::SetStrictCapacityLimit(bool strict_capacity_limit) {
 CacheHandle* ClockCacheShard::Insert(
    const Slice& key, uint32_t hash, void* value, size_t charge,
    void (*deleter)(const Slice& key, void* value), bool hold_reference,
-    CleanupContext* context) {
+    CleanupContext* context, bool* overwritten) {
+  assert(overwritten != nullptr && *overwritten == false);
  size_t total_charge =
      CacheHandle::CalcTotalCharge(key, charge, metadata_charge_policy_);
  MutexLock l(&mutex_);
@@ -597,6 +599,7 @@ CacheHandle* ClockCacheShard::Insert(
  handle->flags.store(flags, std::memory_order_relaxed);
  HashTable::accessor accessor;
  if (table_.find(accessor, CacheKey(key, hash))) {
+    *overwritten = true;
    CacheHandle* existing_handle = accessor->second;
    table_.erase(accessor);
    UnsetInCache(existing_handle, context);
@@ -619,8 +622,9 @@ Status ClockCacheShard::Insert(const Slice& key, uint32_t hash, void* value,
  char* key_data = new char[key.size()];
  memcpy(key_data, key.data(), key.size());
  Slice key_copy(key_data, key.size());
+  bool overwritten = false;
  CacheHandle* handle = Insert(key_copy, hash, value, charge, deleter,
-                               out_handle != nullptr, &context);
+                               out_handle != nullptr, &context, &overwritten);
  Status s;
  if (out_handle != nullptr) {
    if (handle == nullptr) {
@@ -629,6 +633,10 @@ Status ClockCacheShard::Insert(const Slice& key, uint32_t hash, void* value,
      *out_handle = reinterpret_cast<Cache::Handle*>(handle);
    }
  }
+  if (overwritten) {
+    assert(s.ok());
+    s = Status::OkOverwritten();
+  }
  Cleanup(context);
  return s;
 }
@@ -756,6 +764,6 @@ std::shared_ptr<Cache> NewClockCache(
      capacity, num_shard_bits, strict_capacity_limit, metadata_charge_policy);
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // SUPPORT_CLOCK_CACHE
@@ -16,7 +16,7 @@

 #include "util/mutexlock.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 LRUHandleTable::LRUHandleTable() : list_(nullptr), length_(0), elems_(0) {
  Resize();
@@ -386,6 +386,7 @@ Status LRUCacheShard::Insert(const Slice& key, uint32_t hash, void* value,
      LRUHandle* old = table_.Insert(e);
      usage_ += total_charge;
      if (old != nullptr) {
+        s = Status::OkOverwritten();
        assert(old->InCache());
        old->SetInCache(false);
        if (!old->HasRefs()) {
@@ -571,4 +572,4 @@ std::shared_ptr<Cache> NewLRUCache(
      std::move(memory_allocator), use_adaptive_mutex, metadata_charge_policy);
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -16,7 +16,7 @@
 #include "port/port.h"
 #include "util/autovector.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // LRU cache implementation. This class is not thread-safe.

@@ -336,4 +336,4 @@ class LRUCache
  int num_shards_ = 0;
 };

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -10,7 +10,7 @@
 #include "port/port.h"
 #include "test_util/testharness.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class LRUCacheTest : public testing::Test {
 public:
@@ -190,7 +190,7 @@ TEST_F(LRUCacheTest, EntriesWithPriority) {
  ValidateLRUList({"e", "f", "g", "Z", "d"}, 2);
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -13,7 +13,7 @@

 #include "util/mutexlock.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 ShardedCache::ShardedCache(size_t capacity, int num_shard_bits,
                           bool strict_capacity_limit,
@@ -159,4 +159,4 @@ int GetDefaultCacheShardBits(size_t capacity) {
  return num_shard_bits;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -16,7 +16,7 @@
 #include "rocksdb/cache.h"
 #include "util/hash.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // Single cache shard interface.
 class CacheShard {
@@ -108,4 +108,4 @@ class ShardedCache : public Cache {

 extern int GetDefaultCacheShardBits(size_t capacity);

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,7 @@
+macro(get_cxx_std_flags FLAGS_VARIABLE)
+  if( CMAKE_CXX_STANDARD_REQUIRED )
+    set(${FLAGS_VARIABLE} ${CMAKE_CXX${CMAKE_CXX_STANDARD}_STANDARD_COMPILE_OPTION})
+  else()
+    set(${FLAGS_VARIABLE} ${CMAKE_CXX${CMAKE_CXX_STANDARD}_EXTENSION_COMPILE_OPTION})
+  endif()
+endmacro()
@@ -17,16 +17,19 @@ if [ -d /mnt/gvfs/third-party -a -z "$CXX" ]; then
 else
  GCOV=$(which gcov)
 fi
+echo -e "Using $GCOV"

 COVERAGE_DIR="$PWD/COVERAGE_REPORT"
 mkdir -p $COVERAGE_DIR

 # Find all gcno files to generate the coverage report

+PYTHON=${1:-`which python`}
+echo -e "Using $PYTHON"
 GCNO_FILES=`find $ROOT -name "*.gcno"`
 $GCOV --preserve-paths --relative-only --no-output $GCNO_FILES 2>/dev/null |
  # Parse the raw gcov report to more human readable form.
-  python $ROOT/coverage/parse_gcov_output.py |
+  $PYTHON $ROOT/coverage/parse_gcov_output.py |
  # Write the output to both stdout and report file.
  tee $COVERAGE_DIR/coverage_report_all.txt &&
 echo -e "Generated coverage report for all files: $COVERAGE_DIR/coverage_report_all.txt\n"
@@ -41,7 +44,7 @@ RECENT_REPORT=$COVERAGE_DIR/coverage_report_recent.txt

 echo -e "Recently updated files: $LATEST_FILES\n" > $RECENT_REPORT
 $GCOV --preserve-paths --relative-only --no-output $GCNO_FILES 2>/dev/null |
-  python $ROOT/coverage/parse_gcov_output.py -interested-files $LATEST_FILES |
+  $PYTHON $ROOT/coverage/parse_gcov_output.py -interested-files $LATEST_FILES |
  tee -a $RECENT_REPORT &&
 echo -e "Generated coverage report for recently updated files: $RECENT_REPORT\n"

@@ -1,10 +1,12 @@
-#!/usr/bin/env python2
+#!/usr/bin/env python
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
+
+from __future__ import print_function
+
+import optparse
 import re
 import sys

-from optparse import OptionParser
-
 # the gcov report follows certain pattern. Each file will have two lines
 # of report, from which we can extract the file name, total lines and coverage
 # percentage.
@@ -48,7 +50,7 @@ def parse_gcov_report(gcov_input):
 def get_option_parser():
    usage = "Parse the gcov output and generate more human-readable code " +\
            "coverage report."
-    parser = OptionParser(usage)
+    parser = optparse.OptionParser(usage)

    parser.add_option(
        "--interested-files", "-i",
@@ -73,8 +75,8 @@ def display_file_coverage(per_file_coverage, total_coverage):
    header_template = \
        "%" + str(max_file_name_length) + "s\t%s\t%s"
    separator = "-" * (max_file_name_length + 10 + 20)
-    print header_template % ("Filename", "Coverage", "Lines")  # noqa: E999 T25377293 Grandfathered in
-    print separator
+    print(header_template % ("Filename", "Coverage", "Lines"))  # noqa: E999 T25377293 Grandfathered in
+    print(separator)

    # -- Print body
    # template for printing coverage report for each file.
@@ -82,12 +84,12 @@ def display_file_coverage(per_file_coverage, total_coverage):

    for fname, coverage_info in per_file_coverage.items():
        coverage, lines = coverage_info
-        print record_template % (fname, coverage, lines)
+        print(record_template % (fname, coverage, lines))

    # -- Print footer
    if total_coverage:
-        print separator
-        print record_template % ("Total", total_coverage[0], total_coverage[1])
+        print(separator)
+        print(record_template % ("Total", total_coverage[0], total_coverage[1]))

 def report_coverage():
    parser = get_option_parser()
@@ -111,7 +113,7 @@ def report_coverage():
        total_coverage = None

    if not len(per_file_coverage):
-        print >> sys.stderr, "Cannot find coverage info for the given files."
+        print("Cannot find coverage info for the given files.", file=sys.stderr)
        return
    display_file_coverage(per_file_coverage, total_coverage)

@@ -16,7 +16,7 @@
 #include "table/iterator_wrapper.h"
 #include "util/user_comparator_wrapper.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 Status ArenaWrappedDBIter::GetProperty(std::string prop_name,
                                       std::string* prop) {
@@ -75,7 +75,7 @@ Status ArenaWrappedDBIter::Refresh() {

    InternalIterator* internal_iter = db_impl_->NewInternalIterator(
        read_options_, cfd_, sv, &arena_, db_iter_->GetRangeDelAggregator(),
-        latest_seq);
+        latest_seq, /* allow_unprepared_value */ true);
    SetIterUnderDBIter(internal_iter);
  } else {
    db_iter_->set_sequence(latest_seq);
@@ -103,4 +103,4 @@ ArenaWrappedDBIter* NewArenaWrappedDbIterator(
  return iter;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -20,7 +20,7 @@
 #include "rocksdb/iterator.h"
 #include "util/autovector.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class Arena;

@@ -45,26 +45,29 @@ class ArenaWrappedDBIter : public Iterator {
  // Set the internal iterator wrapped inside the DB Iterator. Usually it is
  // a merging iterator.
  virtual void SetIterUnderDBIter(InternalIterator* iter) {
-    static_cast<DBIter*>(db_iter_)->SetIter(iter);
+    db_iter_->SetIter(iter);
  }

-  virtual bool Valid() const override { return db_iter_->Valid(); }
-  virtual void SeekToFirst() override { db_iter_->SeekToFirst(); }
-  virtual void SeekToLast() override { db_iter_->SeekToLast(); }
-  virtual void Seek(const Slice& target) override { db_iter_->Seek(target); }
-  virtual void SeekForPrev(const Slice& target) override {
+  bool Valid() const override { return db_iter_->Valid(); }
+  void SeekToFirst() override { db_iter_->SeekToFirst(); }
+  void SeekToLast() override { db_iter_->SeekToLast(); }
+  // 'target' does not contain timestamp, even if user timestamp feature is
+  // enabled.
+  void Seek(const Slice& target) override { db_iter_->Seek(target); }
+  void SeekForPrev(const Slice& target) override {
    db_iter_->SeekForPrev(target);
  }
-  virtual void Next() override { db_iter_->Next(); }
-  virtual void Prev() override { db_iter_->Prev(); }
-  virtual Slice key() const override { return db_iter_->key(); }
-  virtual Slice value() const override { return db_iter_->value(); }
-  virtual Status status() const override { return db_iter_->status(); }
+  void Next() override { db_iter_->Next(); }
+  void Prev() override { db_iter_->Prev(); }
+  Slice key() const override { return db_iter_->key(); }
+  Slice value() const override { return db_iter_->value(); }
+  Status status() const override { return db_iter_->status(); }
+  Slice timestamp() const override { return db_iter_->timestamp(); }
  bool IsBlob() const { return db_iter_->IsBlob(); }

-  virtual Status GetProperty(std::string prop_name, std::string* prop) override;
+  Status GetProperty(std::string prop_name, std::string* prop) override;

-  virtual Status Refresh() override;
+  Status Refresh() override;

  void Init(Env* env, const ReadOptions& read_options,
            const ImmutableCFOptions& cf_options,
@@ -109,4 +112,4 @@ extern ArenaWrappedDBIter* NewArenaWrappedDbIterator(
    ReadCallback* read_callback, DBImpl* db_impl = nullptr,
    ColumnFamilyData* cfd = nullptr, bool allow_blob = false,
    bool allow_refresh = true);
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,16 @@
+//  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 <cstdint>
+
+#include "rocksdb/rocksdb_namespace.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+constexpr uint64_t kInvalidBlobFileNumber = 0;
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,154 @@
+//  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_addition.h"
+
+#include <ostream>
+#include <sstream>
+
+#include "logging/event_logger.h"
+#include "rocksdb/slice.h"
+#include "rocksdb/status.h"
+#include "test_util/sync_point.h"
+#include "util/coding.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+// Tags for custom fields. Note that these get persisted in the manifest,
+// so existing tags should not be modified.
+enum BlobFileAddition::CustomFieldTags : uint32_t {
+  kEndMarker,
+
+  // Add forward compatible fields here
+
+  /////////////////////////////////////////////////////////////////////
+
+  kForwardIncompatibleMask = 1 << 6,
+
+  // Add forward incompatible fields here
+};
+
+void BlobFileAddition::EncodeTo(std::string* output) const {
+  PutVarint64(output, blob_file_number_);
+  PutVarint64(output, total_blob_count_);
+  PutVarint64(output, total_blob_bytes_);
+  PutLengthPrefixedSlice(output, checksum_method_);
+  PutLengthPrefixedSlice(output, checksum_value_);
+
+  // Encode any custom fields here. The format to use is a Varint32 tag (see
+  // CustomFieldTags above) followed by a length prefixed slice. Unknown custom
+  // fields will be ignored during decoding unless they're in the forward
+  // incompatible range.
+
+  TEST_SYNC_POINT_CALLBACK("BlobFileAddition::EncodeTo::CustomFields", output);
+
+  PutVarint32(output, kEndMarker);
+}
+
+Status BlobFileAddition::DecodeFrom(Slice* input) {
+  constexpr char class_name[] = "BlobFileAddition";
+
+  if (!GetVarint64(input, &blob_file_number_)) {
+    return Status::Corruption(class_name, "Error decoding blob file number");
+  }
+
+  if (!GetVarint64(input, &total_blob_count_)) {
+    return Status::Corruption(class_name, "Error decoding total blob count");
+  }
+
+  if (!GetVarint64(input, &total_blob_bytes_)) {
+    return Status::Corruption(class_name, "Error decoding total blob bytes");
+  }
+
+  Slice checksum_method;
+  if (!GetLengthPrefixedSlice(input, &checksum_method)) {
+    return Status::Corruption(class_name, "Error decoding checksum method");
+  }
+  checksum_method_ = checksum_method.ToString();
+
+  Slice checksum_value;
+  if (!GetLengthPrefixedSlice(input, &checksum_value)) {
+    return Status::Corruption(class_name, "Error decoding checksum value");
+  }
+  checksum_value_ = checksum_value.ToString();
+
+  while (true) {
+    uint32_t custom_field_tag = 0;
+    if (!GetVarint32(input, &custom_field_tag)) {
+      return Status::Corruption(class_name, "Error decoding custom field tag");
+    }
+
+    if (custom_field_tag == kEndMarker) {
+      break;
+    }
+
+    if (custom_field_tag & kForwardIncompatibleMask) {
+      return Status::Corruption(
+          class_name, "Forward incompatible custom field encountered");
+    }
+
+    Slice custom_field_value;
+    if (!GetLengthPrefixedSlice(input, &custom_field_value)) {
+      return Status::Corruption(class_name,
+                                "Error decoding custom field value");
+    }
+  }
+
+  return Status::OK();
+}
+
+std::string BlobFileAddition::DebugString() const {
+  std::ostringstream oss;
+
+  oss << *this;
+
+  return oss.str();
+}
+
+std::string BlobFileAddition::DebugJSON() const {
+  JSONWriter jw;
+
+  jw << *this;
+
+  jw.EndObject();
+
+  return jw.Get();
+}
+
+bool operator==(const BlobFileAddition& lhs, const BlobFileAddition& rhs) {
+  return lhs.GetBlobFileNumber() == rhs.GetBlobFileNumber() &&
+         lhs.GetTotalBlobCount() == rhs.GetTotalBlobCount() &&
+         lhs.GetTotalBlobBytes() == rhs.GetTotalBlobBytes() &&
+         lhs.GetChecksumMethod() == rhs.GetChecksumMethod() &&
+         lhs.GetChecksumValue() == rhs.GetChecksumValue();
+}
+
+bool operator!=(const BlobFileAddition& lhs, const BlobFileAddition& rhs) {
+  return !(lhs == rhs);
+}
+
+std::ostream& operator<<(std::ostream& os,
+                         const BlobFileAddition& blob_file_addition) {
+  os << "blob_file_number: " << blob_file_addition.GetBlobFileNumber()
+     << " total_blob_count: " << blob_file_addition.GetTotalBlobCount()
+     << " total_blob_bytes: " << blob_file_addition.GetTotalBlobBytes()
+     << " checksum_method: " << blob_file_addition.GetChecksumMethod()
+     << " checksum_value: " << blob_file_addition.GetChecksumValue();
+
+  return os;
+}
+
+JSONWriter& operator<<(JSONWriter& jw,
+                       const BlobFileAddition& blob_file_addition) {
+  jw << "BlobFileNumber" << blob_file_addition.GetBlobFileNumber()
+     << "TotalBlobCount" << blob_file_addition.GetTotalBlobCount()
+     << "TotalBlobBytes" << blob_file_addition.GetTotalBlobBytes()
+     << "ChecksumMethod" << blob_file_addition.GetChecksumMethod()
+     << "ChecksumValue" << blob_file_addition.GetChecksumValue();
+
+  return jw;
+}
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,67 @@
+//  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 <cassert>
+#include <cstdint>
+#include <iosfwd>
+#include <string>
+
+#include "db/blob/blob_constants.h"
+#include "rocksdb/rocksdb_namespace.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class JSONWriter;
+class Slice;
+class Status;
+
+class BlobFileAddition {
+ public:
+  BlobFileAddition() = default;
+
+  BlobFileAddition(uint64_t blob_file_number, uint64_t total_blob_count,
+                   uint64_t total_blob_bytes, std::string checksum_method,
+                   std::string checksum_value)
+      : blob_file_number_(blob_file_number),
+        total_blob_count_(total_blob_count),
+        total_blob_bytes_(total_blob_bytes),
+        checksum_method_(std::move(checksum_method)),
+        checksum_value_(std::move(checksum_value)) {
+    assert(checksum_method_.empty() == checksum_value_.empty());
+  }
+
+  uint64_t GetBlobFileNumber() const { return blob_file_number_; }
+  uint64_t GetTotalBlobCount() const { return total_blob_count_; }
+  uint64_t GetTotalBlobBytes() const { return total_blob_bytes_; }
+  const std::string& GetChecksumMethod() const { return checksum_method_; }
+  const std::string& GetChecksumValue() const { return checksum_value_; }
+
+  void EncodeTo(std::string* output) const;
+  Status DecodeFrom(Slice* input);
+
+  std::string DebugString() const;
+  std::string DebugJSON() const;
+
+ private:
+  enum CustomFieldTags : uint32_t;
+
+  uint64_t blob_file_number_ = kInvalidBlobFileNumber;
+  uint64_t total_blob_count_ = 0;
+  uint64_t total_blob_bytes_ = 0;
+  std::string checksum_method_;
+  std::string checksum_value_;
+};
+
+bool operator==(const BlobFileAddition& lhs, const BlobFileAddition& rhs);
+bool operator!=(const BlobFileAddition& lhs, const BlobFileAddition& rhs);
+
+std::ostream& operator<<(std::ostream& os,
+                         const BlobFileAddition& blob_file_addition);
+JSONWriter& operator<<(JSONWriter& jw,
+                       const BlobFileAddition& blob_file_addition);
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,206 @@
+//  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_addition.h"
+
+#include <cstdint>
+#include <cstring>
+#include <string>
+
+#include "test_util/sync_point.h"
+#include "test_util/testharness.h"
+#include "util/coding.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class BlobFileAdditionTest : public testing::Test {
+ public:
+  static void TestEncodeDecode(const BlobFileAddition& blob_file_addition) {
+    std::string encoded;
+    blob_file_addition.EncodeTo(&encoded);
+
+    BlobFileAddition decoded;
+    Slice input(encoded);
+    ASSERT_OK(decoded.DecodeFrom(&input));
+
+    ASSERT_EQ(blob_file_addition, decoded);
+  }
+};
+
+TEST_F(BlobFileAdditionTest, Empty) {
+  BlobFileAddition blob_file_addition;
+
+  ASSERT_EQ(blob_file_addition.GetBlobFileNumber(), kInvalidBlobFileNumber);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobCount(), 0);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobBytes(), 0);
+  ASSERT_TRUE(blob_file_addition.GetChecksumMethod().empty());
+  ASSERT_TRUE(blob_file_addition.GetChecksumValue().empty());
+
+  TestEncodeDecode(blob_file_addition);
+}
+
+TEST_F(BlobFileAdditionTest, NonEmpty) {
+  constexpr uint64_t blob_file_number = 123;
+  constexpr uint64_t total_blob_count = 2;
+  constexpr uint64_t total_blob_bytes = 123456;
+  const std::string checksum_method("SHA1");
+  const std::string checksum_value("bdb7f34a59dfa1592ce7f52e99f98c570c525cbd");
+
+  BlobFileAddition blob_file_addition(blob_file_number, total_blob_count,
+                                      total_blob_bytes, checksum_method,
+                                      checksum_value);
+
+  ASSERT_EQ(blob_file_addition.GetBlobFileNumber(), blob_file_number);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobCount(), total_blob_count);
+  ASSERT_EQ(blob_file_addition.GetTotalBlobBytes(), total_blob_bytes);
+  ASSERT_EQ(blob_file_addition.GetChecksumMethod(), checksum_method);
+  ASSERT_EQ(blob_file_addition.GetChecksumValue(), checksum_value);
+
+  TestEncodeDecode(blob_file_addition);
+}
+
+TEST_F(BlobFileAdditionTest, DecodeErrors) {
+  std::string str;
+  Slice slice(str);
+
+  BlobFileAddition blob_file_addition;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "blob file number"));
+  }
+
+  constexpr uint64_t blob_file_number = 123;
+  PutVarint64(&str, blob_file_number);
+  slice = str;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "total blob count"));
+  }
+
+  constexpr uint64_t total_blob_count = 4567;
+  PutVarint64(&str, total_blob_count);
+  slice = str;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "total blob bytes"));
+  }
+
+  constexpr uint64_t total_blob_bytes = 12345678;
+  PutVarint64(&str, total_blob_bytes);
+  slice = str;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "checksum method"));
+  }
+
+  constexpr char checksum_method[] = "SHA1";
+  PutLengthPrefixedSlice(&str, checksum_method);
+  slice = str;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "checksum value"));
+  }
+
+  constexpr char checksum_value[] = "bdb7f34a59dfa1592ce7f52e99f98c570c525cbd";
+  PutLengthPrefixedSlice(&str, checksum_value);
+  slice = str;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "custom field tag"));
+  }
+
+  constexpr uint32_t custom_tag = 2;
+  PutVarint32(&str, custom_tag);
+  slice = str;
+
+  {
+    const Status s = blob_file_addition.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "custom field value"));
+  }
+}
+
+TEST_F(BlobFileAdditionTest, ForwardCompatibleCustomField) {
+  SyncPoint::GetInstance()->SetCallBack(
+      "BlobFileAddition::EncodeTo::CustomFields", [&](void* arg) {
+        std::string* output = static_cast<std::string*>(arg);
+
+        constexpr uint32_t forward_compatible_tag = 2;
+        PutVarint32(output, forward_compatible_tag);
+
+        PutLengthPrefixedSlice(output, "deadbeef");
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  constexpr uint64_t blob_file_number = 678;
+  constexpr uint64_t total_blob_count = 9999;
+  constexpr uint64_t total_blob_bytes = 100000000;
+  const std::string checksum_method("CRC32");
+  const std::string checksum_value("3d87ff57");
+
+  BlobFileAddition blob_file_addition(blob_file_number, total_blob_count,
+                                      total_blob_bytes, checksum_method,
+                                      checksum_value);
+
+  TestEncodeDecode(blob_file_addition);
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+TEST_F(BlobFileAdditionTest, ForwardIncompatibleCustomField) {
+  SyncPoint::GetInstance()->SetCallBack(
+      "BlobFileAddition::EncodeTo::CustomFields", [&](void* arg) {
+        std::string* output = static_cast<std::string*>(arg);
+
+        constexpr uint32_t forward_incompatible_tag = (1 << 6) + 1;
+        PutVarint32(output, forward_incompatible_tag);
+
+        PutLengthPrefixedSlice(output, "foobar");
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  constexpr uint64_t blob_file_number = 456;
+  constexpr uint64_t total_blob_count = 100;
+  constexpr uint64_t total_blob_bytes = 2000000;
+  const std::string checksum_method("CRC32B");
+  const std::string checksum_value("6dbdf23a");
+
+  BlobFileAddition blob_file_addition(blob_file_number, total_blob_count,
+                                      total_blob_bytes, checksum_method,
+                                      checksum_value);
+
+  std::string encoded;
+  blob_file_addition.EncodeTo(&encoded);
+
+  BlobFileAddition decoded_blob_file_addition;
+  Slice input(encoded);
+  const Status s = decoded_blob_file_addition.DecodeFrom(&input);
+
+  ASSERT_TRUE(s.IsCorruption());
+  ASSERT_TRUE(std::strstr(s.getState(), "Forward incompatible"));
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+}  // namespace ROCKSDB_NAMESPACE
+
+int main(int argc, char** argv) {
+  ::testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}
@@ -0,0 +1,134 @@
+//  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_garbage.h"
+
+#include <ostream>
+#include <sstream>
+
+#include "logging/event_logger.h"
+#include "rocksdb/slice.h"
+#include "rocksdb/status.h"
+#include "test_util/sync_point.h"
+#include "util/coding.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+// Tags for custom fields. Note that these get persisted in the manifest,
+// so existing tags should not be modified.
+enum BlobFileGarbage::CustomFieldTags : uint32_t {
+  kEndMarker,
+
+  // Add forward compatible fields here
+
+  /////////////////////////////////////////////////////////////////////
+
+  kForwardIncompatibleMask = 1 << 6,
+
+  // Add forward incompatible fields here
+};
+
+void BlobFileGarbage::EncodeTo(std::string* output) const {
+  PutVarint64(output, blob_file_number_);
+  PutVarint64(output, garbage_blob_count_);
+  PutVarint64(output, garbage_blob_bytes_);
+
+  // Encode any custom fields here. The format to use is a Varint32 tag (see
+  // CustomFieldTags above) followed by a length prefixed slice. Unknown custom
+  // fields will be ignored during decoding unless they're in the forward
+  // incompatible range.
+
+  TEST_SYNC_POINT_CALLBACK("BlobFileGarbage::EncodeTo::CustomFields", output);
+
+  PutVarint32(output, kEndMarker);
+}
+
+Status BlobFileGarbage::DecodeFrom(Slice* input) {
+  constexpr char class_name[] = "BlobFileGarbage";
+
+  if (!GetVarint64(input, &blob_file_number_)) {
+    return Status::Corruption(class_name, "Error decoding blob file number");
+  }
+
+  if (!GetVarint64(input, &garbage_blob_count_)) {
+    return Status::Corruption(class_name, "Error decoding garbage blob count");
+  }
+
+  if (!GetVarint64(input, &garbage_blob_bytes_)) {
+    return Status::Corruption(class_name, "Error decoding garbage blob bytes");
+  }
+
+  while (true) {
+    uint32_t custom_field_tag = 0;
+    if (!GetVarint32(input, &custom_field_tag)) {
+      return Status::Corruption(class_name, "Error decoding custom field tag");
+    }
+
+    if (custom_field_tag == kEndMarker) {
+      break;
+    }
+
+    if (custom_field_tag & kForwardIncompatibleMask) {
+      return Status::Corruption(
+          class_name, "Forward incompatible custom field encountered");
+    }
+
+    Slice custom_field_value;
+    if (!GetLengthPrefixedSlice(input, &custom_field_value)) {
+      return Status::Corruption(class_name,
+                                "Error decoding custom field value");
+    }
+  }
+
+  return Status::OK();
+}
+
+std::string BlobFileGarbage::DebugString() const {
+  std::ostringstream oss;
+
+  oss << *this;
+
+  return oss.str();
+}
+
+std::string BlobFileGarbage::DebugJSON() const {
+  JSONWriter jw;
+
+  jw << *this;
+
+  jw.EndObject();
+
+  return jw.Get();
+}
+
+bool operator==(const BlobFileGarbage& lhs, const BlobFileGarbage& rhs) {
+  return lhs.GetBlobFileNumber() == rhs.GetBlobFileNumber() &&
+         lhs.GetGarbageBlobCount() == rhs.GetGarbageBlobCount() &&
+         lhs.GetGarbageBlobBytes() == rhs.GetGarbageBlobBytes();
+}
+
+bool operator!=(const BlobFileGarbage& lhs, const BlobFileGarbage& rhs) {
+  return !(lhs == rhs);
+}
+
+std::ostream& operator<<(std::ostream& os,
+                         const BlobFileGarbage& blob_file_garbage) {
+  os << "blob_file_number: " << blob_file_garbage.GetBlobFileNumber()
+     << " garbage_blob_count: " << blob_file_garbage.GetGarbageBlobCount()
+     << " garbage_blob_bytes: " << blob_file_garbage.GetGarbageBlobBytes();
+
+  return os;
+}
+
+JSONWriter& operator<<(JSONWriter& jw,
+                       const BlobFileGarbage& blob_file_garbage) {
+  jw << "BlobFileNumber" << blob_file_garbage.GetBlobFileNumber()
+     << "GarbageBlobCount" << blob_file_garbage.GetGarbageBlobCount()
+     << "GarbageBlobBytes" << blob_file_garbage.GetGarbageBlobBytes();
+
+  return jw;
+}
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,57 @@
+//  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 <cstdint>
+#include <iosfwd>
+#include <string>
+
+#include "db/blob/blob_constants.h"
+#include "rocksdb/rocksdb_namespace.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class JSONWriter;
+class Slice;
+class Status;
+
+class BlobFileGarbage {
+ public:
+  BlobFileGarbage() = default;
+
+  BlobFileGarbage(uint64_t blob_file_number, uint64_t garbage_blob_count,
+                  uint64_t garbage_blob_bytes)
+      : blob_file_number_(blob_file_number),
+        garbage_blob_count_(garbage_blob_count),
+        garbage_blob_bytes_(garbage_blob_bytes) {}
+
+  uint64_t GetBlobFileNumber() const { return blob_file_number_; }
+  uint64_t GetGarbageBlobCount() const { return garbage_blob_count_; }
+  uint64_t GetGarbageBlobBytes() const { return garbage_blob_bytes_; }
+
+  void EncodeTo(std::string* output) const;
+  Status DecodeFrom(Slice* input);
+
+  std::string DebugString() const;
+  std::string DebugJSON() const;
+
+ private:
+  enum CustomFieldTags : uint32_t;
+
+  uint64_t blob_file_number_ = kInvalidBlobFileNumber;
+  uint64_t garbage_blob_count_ = 0;
+  uint64_t garbage_blob_bytes_ = 0;
+};
+
+bool operator==(const BlobFileGarbage& lhs, const BlobFileGarbage& rhs);
+bool operator!=(const BlobFileGarbage& lhs, const BlobFileGarbage& rhs);
+
+std::ostream& operator<<(std::ostream& os,
+                         const BlobFileGarbage& blob_file_garbage);
+JSONWriter& operator<<(JSONWriter& jw,
+                       const BlobFileGarbage& blob_file_garbage);
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,173 @@
+//  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_garbage.h"
+
+#include <cstdint>
+#include <cstring>
+#include <string>
+
+#include "test_util/sync_point.h"
+#include "test_util/testharness.h"
+#include "util/coding.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class BlobFileGarbageTest : public testing::Test {
+ public:
+  static void TestEncodeDecode(const BlobFileGarbage& blob_file_garbage) {
+    std::string encoded;
+    blob_file_garbage.EncodeTo(&encoded);
+
+    BlobFileGarbage decoded;
+    Slice input(encoded);
+    ASSERT_OK(decoded.DecodeFrom(&input));
+
+    ASSERT_EQ(blob_file_garbage, decoded);
+  }
+};
+
+TEST_F(BlobFileGarbageTest, Empty) {
+  BlobFileGarbage blob_file_garbage;
+
+  ASSERT_EQ(blob_file_garbage.GetBlobFileNumber(), kInvalidBlobFileNumber);
+  ASSERT_EQ(blob_file_garbage.GetGarbageBlobCount(), 0);
+  ASSERT_EQ(blob_file_garbage.GetGarbageBlobBytes(), 0);
+
+  TestEncodeDecode(blob_file_garbage);
+}
+
+TEST_F(BlobFileGarbageTest, NonEmpty) {
+  constexpr uint64_t blob_file_number = 123;
+  constexpr uint64_t garbage_blob_count = 1;
+  constexpr uint64_t garbage_blob_bytes = 9876;
+
+  BlobFileGarbage blob_file_garbage(blob_file_number, garbage_blob_count,
+                                    garbage_blob_bytes);
+
+  ASSERT_EQ(blob_file_garbage.GetBlobFileNumber(), blob_file_number);
+  ASSERT_EQ(blob_file_garbage.GetGarbageBlobCount(), garbage_blob_count);
+  ASSERT_EQ(blob_file_garbage.GetGarbageBlobBytes(), garbage_blob_bytes);
+
+  TestEncodeDecode(blob_file_garbage);
+}
+
+TEST_F(BlobFileGarbageTest, DecodeErrors) {
+  std::string str;
+  Slice slice(str);
+
+  BlobFileGarbage blob_file_garbage;
+
+  {
+    const Status s = blob_file_garbage.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "blob file number"));
+  }
+
+  constexpr uint64_t blob_file_number = 123;
+  PutVarint64(&str, blob_file_number);
+  slice = str;
+
+  {
+    const Status s = blob_file_garbage.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "garbage blob count"));
+  }
+
+  constexpr uint64_t garbage_blob_count = 4567;
+  PutVarint64(&str, garbage_blob_count);
+  slice = str;
+
+  {
+    const Status s = blob_file_garbage.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "garbage blob bytes"));
+  }
+
+  constexpr uint64_t garbage_blob_bytes = 12345678;
+  PutVarint64(&str, garbage_blob_bytes);
+  slice = str;
+
+  {
+    const Status s = blob_file_garbage.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "custom field tag"));
+  }
+
+  constexpr uint32_t custom_tag = 2;
+  PutVarint32(&str, custom_tag);
+  slice = str;
+
+  {
+    const Status s = blob_file_garbage.DecodeFrom(&slice);
+    ASSERT_TRUE(s.IsCorruption());
+    ASSERT_TRUE(std::strstr(s.getState(), "custom field value"));
+  }
+}
+
+TEST_F(BlobFileGarbageTest, ForwardCompatibleCustomField) {
+  SyncPoint::GetInstance()->SetCallBack(
+      "BlobFileGarbage::EncodeTo::CustomFields", [&](void* arg) {
+        std::string* output = static_cast<std::string*>(arg);
+
+        constexpr uint32_t forward_compatible_tag = 2;
+        PutVarint32(output, forward_compatible_tag);
+
+        PutLengthPrefixedSlice(output, "deadbeef");
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  constexpr uint64_t blob_file_number = 678;
+  constexpr uint64_t garbage_blob_count = 9999;
+  constexpr uint64_t garbage_blob_bytes = 100000000;
+
+  BlobFileGarbage blob_file_garbage(blob_file_number, garbage_blob_count,
+                                    garbage_blob_bytes);
+
+  TestEncodeDecode(blob_file_garbage);
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+TEST_F(BlobFileGarbageTest, ForwardIncompatibleCustomField) {
+  SyncPoint::GetInstance()->SetCallBack(
+      "BlobFileGarbage::EncodeTo::CustomFields", [&](void* arg) {
+        std::string* output = static_cast<std::string*>(arg);
+
+        constexpr uint32_t forward_incompatible_tag = (1 << 6) + 1;
+        PutVarint32(output, forward_incompatible_tag);
+
+        PutLengthPrefixedSlice(output, "foobar");
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  constexpr uint64_t blob_file_number = 456;
+  constexpr uint64_t garbage_blob_count = 100;
+  constexpr uint64_t garbage_blob_bytes = 2000000;
+
+  BlobFileGarbage blob_file_garbage(blob_file_number, garbage_blob_count,
+                                    garbage_blob_bytes);
+
+  std::string encoded;
+  blob_file_garbage.EncodeTo(&encoded);
+
+  BlobFileGarbage decoded_blob_file_addition;
+  Slice input(encoded);
+  const Status s = decoded_blob_file_addition.DecodeFrom(&input);
+
+  ASSERT_TRUE(s.IsCorruption());
+  ASSERT_TRUE(std::strstr(s.getState(), "Forward incompatible"));
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+}  // namespace ROCKSDB_NAMESPACE
+
+int main(int argc, char** argv) {
+  ::testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}
@@ -0,0 +1,48 @@
+//  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_meta.h"
+
+#include <ostream>
+#include <sstream>
+
+namespace ROCKSDB_NAMESPACE {
+
+std::string SharedBlobFileMetaData::DebugString() const {
+  std::ostringstream oss;
+  oss << (*this);
+
+  return oss.str();
+}
+
+std::ostream& operator<<(std::ostream& os,
+                         const SharedBlobFileMetaData& shared_meta) {
+  os << "blob_file_number: " << shared_meta.GetBlobFileNumber()
+     << " total_blob_count: " << shared_meta.GetTotalBlobCount()
+     << " total_blob_bytes: " << shared_meta.GetTotalBlobBytes()
+     << " checksum_method: " << shared_meta.GetChecksumMethod()
+     << " checksum_value: " << shared_meta.GetChecksumValue();
+
+  return os;
+}
+
+std::string BlobFileMetaData::DebugString() const {
+  std::ostringstream oss;
+  oss << (*this);
+
+  return oss.str();
+}
+
+std::ostream& operator<<(std::ostream& os, const BlobFileMetaData& meta) {
+  const auto& shared_meta = meta.GetSharedMeta();
+  assert(shared_meta);
+
+  os << (*shared_meta) << " garbage_blob_count: " << meta.GetGarbageBlobCount()
+     << " garbage_blob_bytes: " << meta.GetGarbageBlobBytes();
+
+  return os;
+}
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -0,0 +1,154 @@
+//  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 "rocksdb/rocksdb_namespace.h"
+
+#include <cassert>
+#include <iosfwd>
+#include <memory>
+#include <string>
+
+namespace ROCKSDB_NAMESPACE {
+
+// SharedBlobFileMetaData represents the immutable part of blob files' metadata,
+// like the blob file number, total number and size of blobs, or checksum
+// method and value. There is supposed to be one object of this class per blob
+// file (shared across all versions that include the blob file in question);
+// hence, the type is neither copyable nor movable. A blob file can be marked
+// obsolete when the corresponding SharedBlobFileMetaData object is destroyed.
+
+class SharedBlobFileMetaData {
+ public:
+  static std::shared_ptr<SharedBlobFileMetaData> Create(
+      uint64_t blob_file_number, uint64_t total_blob_count,
+      uint64_t total_blob_bytes, std::string checksum_method,
+      std::string checksum_value) {
+    return std::shared_ptr<SharedBlobFileMetaData>(new SharedBlobFileMetaData(
+        blob_file_number, total_blob_count, total_blob_bytes,
+        std::move(checksum_method), std::move(checksum_value)));
+  }
+
+  template <typename Deleter>
+  static std::shared_ptr<SharedBlobFileMetaData> Create(
+      uint64_t blob_file_number, uint64_t total_blob_count,
+      uint64_t total_blob_bytes, std::string checksum_method,
+      std::string checksum_value, Deleter deleter) {
+    return std::shared_ptr<SharedBlobFileMetaData>(
+        new SharedBlobFileMetaData(blob_file_number, total_blob_count,
+                                   total_blob_bytes, std::move(checksum_method),
+                                   std::move(checksum_value)),
+        deleter);
+  }
+
+  SharedBlobFileMetaData(const SharedBlobFileMetaData&) = delete;
+  SharedBlobFileMetaData& operator=(const SharedBlobFileMetaData&) = delete;
+
+  SharedBlobFileMetaData(SharedBlobFileMetaData&&) = delete;
+  SharedBlobFileMetaData& operator=(SharedBlobFileMetaData&&) = delete;
+
+  uint64_t GetBlobFileNumber() const { return blob_file_number_; }
+  uint64_t GetTotalBlobCount() const { return total_blob_count_; }
+  uint64_t GetTotalBlobBytes() const { return total_blob_bytes_; }
+  const std::string& GetChecksumMethod() const { return checksum_method_; }
+  const std::string& GetChecksumValue() const { return checksum_value_; }
+
+  std::string DebugString() const;
+
+ private:
+  SharedBlobFileMetaData(uint64_t blob_file_number, uint64_t total_blob_count,
+                         uint64_t total_blob_bytes, std::string checksum_method,
+                         std::string checksum_value)
+      : blob_file_number_(blob_file_number),
+        total_blob_count_(total_blob_count),
+        total_blob_bytes_(total_blob_bytes),
+        checksum_method_(std::move(checksum_method)),
+        checksum_value_(std::move(checksum_value)) {
+    assert(checksum_method_.empty() == checksum_value_.empty());
+  }
+
+  uint64_t blob_file_number_;
+  uint64_t total_blob_count_;
+  uint64_t total_blob_bytes_;
+  std::string checksum_method_;
+  std::string checksum_value_;
+};
+
+std::ostream& operator<<(std::ostream& os,
+                         const SharedBlobFileMetaData& shared_meta);
+
+// BlobFileMetaData contains the part of the metadata for blob files that can
+// vary across versions, like the amount of garbage in the blob file. In
+// addition, BlobFileMetaData objects point to and share the ownership of the
+// SharedBlobFileMetaData object for the corresponding blob file. Similarly to
+// SharedBlobFileMetaData, BlobFileMetaData are not copyable or movable. They
+// are meant to be jointly owned by the versions in which the blob file has the
+// same (immutable *and* mutable) state.
+
+class BlobFileMetaData {
+ public:
+  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));
+  }
+
+  BlobFileMetaData(const BlobFileMetaData&) = delete;
+  BlobFileMetaData& operator=(const BlobFileMetaData&) = delete;
+
+  BlobFileMetaData(BlobFileMetaData&&) = delete;
+  BlobFileMetaData& operator=(BlobFileMetaData&&) = delete;
+
+  const std::shared_ptr<SharedBlobFileMetaData>& GetSharedMeta() const {
+    return shared_meta_;
+  }
+
+  uint64_t GetBlobFileNumber() const {
+    assert(shared_meta_);
+    return shared_meta_->GetBlobFileNumber();
+  }
+  uint64_t GetTotalBlobCount() const {
+    assert(shared_meta_);
+    return shared_meta_->GetTotalBlobCount();
+  }
+  uint64_t GetTotalBlobBytes() const {
+    assert(shared_meta_);
+    return shared_meta_->GetTotalBlobBytes();
+  }
+  const std::string& GetChecksumMethod() const {
+    assert(shared_meta_);
+    return shared_meta_->GetChecksumMethod();
+  }
+  const std::string& GetChecksumValue() const {
+    assert(shared_meta_);
+    return shared_meta_->GetChecksumValue();
+  }
+
+  uint64_t GetGarbageBlobCount() const { return garbage_blob_count_; }
+  uint64_t GetGarbageBlobBytes() const { return garbage_blob_bytes_; }
+
+  std::string DebugString() const;
+
+ private:
+  BlobFileMetaData(std::shared_ptr<SharedBlobFileMetaData> shared_meta,
+                   uint64_t garbage_blob_count, uint64_t garbage_blob_bytes)
+      : shared_meta_(std::move(shared_meta)),
+        garbage_blob_count_(garbage_blob_count),
+        garbage_blob_bytes_(garbage_blob_bytes) {
+    assert(shared_meta_);
+    assert(garbage_blob_count_ <= shared_meta_->GetTotalBlobCount());
+    assert(garbage_blob_bytes_ <= shared_meta_->GetTotalBlobBytes());
+  }
+
+  std::shared_ptr<SharedBlobFileMetaData> shared_meta_;
+  uint64_t garbage_blob_count_;
+  uint64_t garbage_blob_bytes_;
+};
+
+std::ostream& operator<<(std::ostream& os, const BlobFileMetaData& meta);
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -12,7 +12,7 @@
 #include "util/coding.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // BlobIndex is a pointer to the blob and metadata of the blob. The index is
 // stored in base DB as ValueType::kTypeBlobIndex.
@@ -175,5 +175,5 @@ class BlobIndex {
  CompressionType compression_ = kNoCompression;
 };

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -23,7 +23,7 @@
 #include "util/string_util.h"
 #include "utilities/merge_operators.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // kTypeBlobIndex is a value type used by BlobDB only. The base rocksdb
 // should accept the value type on write, and report not supported value
@@ -427,10 +427,10 @@ TEST_F(DBBlobIndexTest, Iterate) {
  }
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -37,7 +37,7 @@
 #include "test_util/sync_point.h"
 #include "util/stop_watch.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class TableFactory;

@@ -81,10 +81,11 @@ Status BuildTable(
    SnapshotChecker* snapshot_checker, const CompressionType compression,
    uint64_t sample_for_compression, const CompressionOptions& compression_opts,
    bool paranoid_file_checks, InternalStats* internal_stats,
-    TableFileCreationReason reason, 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) {
+    TableFileCreationReason reason, IOStatus* io_status,
+    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) {
  assert((column_family_id ==
          TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
         column_family_name.empty());
@@ -131,9 +132,10 @@ Status BuildTable(
      file->SetIOPriority(io_priority);
      file->SetWriteLifeTimeHint(write_hint);

-      file_writer.reset(
-          new WritableFileWriter(std::move(file), fname, file_options, env,
-                                 ioptions.statistics, ioptions.listeners));
+      file_writer.reset(new WritableFileWriter(
+          std::move(file), fname, file_options, env, ioptions.statistics,
+          ioptions.listeners, ioptions.file_checksum_gen_factory));
+
      builder = NewTableBuilder(
          ioptions, mutable_cf_options, internal_comparator,
          int_tbl_prop_collector_factories, column_family_id,
@@ -181,14 +183,15 @@ Status BuildTable(
    }

    // Finish and check for builder errors
-    tp = builder->GetTableProperties();
-    bool empty = builder->NumEntries() == 0 && tp.num_range_deletions == 0;
+    bool empty = builder->IsEmpty();
    s = c_iter.status();
+    TEST_SYNC_POINT("BuildTable:BeforeFinishBuildTable");
    if (!s.ok() || empty) {
      builder->Abandon();
    } else {
      s = builder->Finish();
    }
+    *io_status = builder->io_status();

    if (s.ok() && !empty) {
      uint64_t file_size = builder->FileSize();
@@ -205,11 +208,22 @@ Status BuildTable(
    // Finish and check for file errors
    if (s.ok() && !empty) {
      StopWatch sw(env, ioptions.statistics, TABLE_SYNC_MICROS);
-      s = file_writer->Sync(ioptions.use_fsync);
+      *io_status = file_writer->Sync(ioptions.use_fsync);
    }
-    if (s.ok() && !empty) {
-      s = file_writer->Close();
+    if (io_status->ok() && !empty) {
+      *io_status = file_writer->Close();
    }
+    if (io_status->ok() && !empty) {
+      // Add the checksum information to file metadata.
+      meta->file_checksum = file_writer->GetFileChecksum();
+      meta->file_checksum_func_name = file_writer->GetFileChecksumFuncName();
+    }
+
+    if (!io_status->ok()) {
+      s = *io_status;
+    }
+
+    // TODO Also check the IO status when create the Iterator.

    if (s.ok() && !empty) {
      // Verify that the table is usable
@@ -226,7 +240,8 @@ Status BuildTable(
                                      : internal_stats->GetFileReadHist(0),
          TableReaderCaller::kFlush, /*arena=*/nullptr,
          /*skip_filter=*/false, level, /*smallest_compaction_key=*/nullptr,
-          /*largest_compaction_key*/ nullptr));
+          /*largest_compaction_key*/ nullptr,
+          /*allow_unprepared_value*/ false));
      s = it->status();
      if (s.ok() && paranoid_file_checks) {
        for (it->SeekToFirst(); it->Valid(); it->Next()) {
@@ -256,4 +271,4 @@ Status BuildTable(
  return s;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -22,7 +22,7 @@
 #include "rocksdb/types.h"
 #include "table/scoped_arena_iterator.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 struct Options;
 struct FileMetaData;
@@ -78,11 +78,11 @@ extern Status BuildTable(
    const uint64_t sample_for_compression,
    const CompressionOptions& compression_opts, bool paranoid_file_checks,
    InternalStats* internal_stats, TableFileCreationReason reason,
-    EventLogger* event_logger = nullptr, int job_id = 0,
+    IOStatus* io_status, 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);

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -46,74 +46,74 @@
 #include <unordered_set>
 #include <map>

-using rocksdb::BytewiseComparator;
-using rocksdb::Cache;
-using rocksdb::ColumnFamilyDescriptor;
-using rocksdb::ColumnFamilyHandle;
-using rocksdb::ColumnFamilyOptions;
-using rocksdb::CompactionFilter;
-using rocksdb::CompactionFilterFactory;
-using rocksdb::CompactionFilterContext;
-using rocksdb::CompactionOptionsFIFO;
-using rocksdb::Comparator;
-using rocksdb::CompressionType;
-using rocksdb::WALRecoveryMode;
-using rocksdb::DB;
-using rocksdb::DBOptions;
-using rocksdb::DbPath;
-using rocksdb::Env;
-using rocksdb::EnvOptions;
-using rocksdb::InfoLogLevel;
-using rocksdb::FileLock;
-using rocksdb::FilterPolicy;
-using rocksdb::FlushOptions;
-using rocksdb::IngestExternalFileOptions;
-using rocksdb::Iterator;
-using rocksdb::Logger;
-using rocksdb::MergeOperator;
-using rocksdb::MergeOperators;
-using rocksdb::NewBloomFilterPolicy;
-using rocksdb::NewLRUCache;
-using rocksdb::Options;
-using rocksdb::BlockBasedTableOptions;
-using rocksdb::CuckooTableOptions;
-using rocksdb::RandomAccessFile;
-using rocksdb::Range;
-using rocksdb::ReadOptions;
-using rocksdb::SequentialFile;
-using rocksdb::Slice;
-using rocksdb::SliceParts;
-using rocksdb::SliceTransform;
-using rocksdb::Snapshot;
-using rocksdb::SstFileWriter;
-using rocksdb::Status;
-using rocksdb::WritableFile;
-using rocksdb::WriteBatch;
-using rocksdb::WriteBatchWithIndex;
-using rocksdb::WriteOptions;
-using rocksdb::LiveFileMetaData;
-using rocksdb::BackupEngine;
-using rocksdb::BackupableDBOptions;
-using rocksdb::BackupInfo;
-using rocksdb::BackupID;
-using rocksdb::RestoreOptions;
-using rocksdb::CompactRangeOptions;
-using rocksdb::BottommostLevelCompaction;
-using rocksdb::RateLimiter;
-using rocksdb::NewGenericRateLimiter;
-using rocksdb::PinnableSlice;
-using rocksdb::TransactionDBOptions;
-using rocksdb::TransactionDB;
-using rocksdb::TransactionOptions;
-using rocksdb::OptimisticTransactionDB;
-using rocksdb::OptimisticTransactionOptions;
-using rocksdb::Transaction;
-using rocksdb::Checkpoint;
-using rocksdb::TransactionLogIterator;
-using rocksdb::BatchResult;
-using rocksdb::PerfLevel;
-using rocksdb::PerfContext;
-using rocksdb::MemoryUtil;
+using ROCKSDB_NAMESPACE::BackupableDBOptions;
+using ROCKSDB_NAMESPACE::BackupEngine;
+using ROCKSDB_NAMESPACE::BackupID;
+using ROCKSDB_NAMESPACE::BackupInfo;
+using ROCKSDB_NAMESPACE::BatchResult;
+using ROCKSDB_NAMESPACE::BlockBasedTableOptions;
+using ROCKSDB_NAMESPACE::BottommostLevelCompaction;
+using ROCKSDB_NAMESPACE::BytewiseComparator;
+using ROCKSDB_NAMESPACE::Cache;
+using ROCKSDB_NAMESPACE::Checkpoint;
+using ROCKSDB_NAMESPACE::ColumnFamilyDescriptor;
+using ROCKSDB_NAMESPACE::ColumnFamilyHandle;
+using ROCKSDB_NAMESPACE::ColumnFamilyOptions;
+using ROCKSDB_NAMESPACE::CompactionFilter;
+using ROCKSDB_NAMESPACE::CompactionFilterContext;
+using ROCKSDB_NAMESPACE::CompactionFilterFactory;
+using ROCKSDB_NAMESPACE::CompactionOptionsFIFO;
+using ROCKSDB_NAMESPACE::CompactRangeOptions;
+using ROCKSDB_NAMESPACE::Comparator;
+using ROCKSDB_NAMESPACE::CompressionType;
+using ROCKSDB_NAMESPACE::CuckooTableOptions;
+using ROCKSDB_NAMESPACE::DB;
+using ROCKSDB_NAMESPACE::DBOptions;
+using ROCKSDB_NAMESPACE::DbPath;
+using ROCKSDB_NAMESPACE::Env;
+using ROCKSDB_NAMESPACE::EnvOptions;
+using ROCKSDB_NAMESPACE::FileLock;
+using ROCKSDB_NAMESPACE::FilterPolicy;
+using ROCKSDB_NAMESPACE::FlushOptions;
+using ROCKSDB_NAMESPACE::InfoLogLevel;
+using ROCKSDB_NAMESPACE::IngestExternalFileOptions;
+using ROCKSDB_NAMESPACE::Iterator;
+using ROCKSDB_NAMESPACE::LiveFileMetaData;
+using ROCKSDB_NAMESPACE::Logger;
+using ROCKSDB_NAMESPACE::MemoryUtil;
+using ROCKSDB_NAMESPACE::MergeOperator;
+using ROCKSDB_NAMESPACE::MergeOperators;
+using ROCKSDB_NAMESPACE::NewBloomFilterPolicy;
+using ROCKSDB_NAMESPACE::NewGenericRateLimiter;
+using ROCKSDB_NAMESPACE::NewLRUCache;
+using ROCKSDB_NAMESPACE::OptimisticTransactionDB;
+using ROCKSDB_NAMESPACE::OptimisticTransactionOptions;
+using ROCKSDB_NAMESPACE::Options;
+using ROCKSDB_NAMESPACE::PerfContext;
+using ROCKSDB_NAMESPACE::PerfLevel;
+using ROCKSDB_NAMESPACE::PinnableSlice;
+using ROCKSDB_NAMESPACE::RandomAccessFile;
+using ROCKSDB_NAMESPACE::Range;
+using ROCKSDB_NAMESPACE::RateLimiter;
+using ROCKSDB_NAMESPACE::ReadOptions;
+using ROCKSDB_NAMESPACE::RestoreOptions;
+using ROCKSDB_NAMESPACE::SequentialFile;
+using ROCKSDB_NAMESPACE::Slice;
+using ROCKSDB_NAMESPACE::SliceParts;
+using ROCKSDB_NAMESPACE::SliceTransform;
+using ROCKSDB_NAMESPACE::Snapshot;
+using ROCKSDB_NAMESPACE::SstFileWriter;
+using ROCKSDB_NAMESPACE::Status;
+using ROCKSDB_NAMESPACE::Transaction;
+using ROCKSDB_NAMESPACE::TransactionDB;
+using ROCKSDB_NAMESPACE::TransactionDBOptions;
+using ROCKSDB_NAMESPACE::TransactionLogIterator;
+using ROCKSDB_NAMESPACE::TransactionOptions;
+using ROCKSDB_NAMESPACE::WALRecoveryMode;
+using ROCKSDB_NAMESPACE::WritableFile;
+using ROCKSDB_NAMESPACE::WriteBatch;
+using ROCKSDB_NAMESPACE::WriteBatchWithIndex;
+using ROCKSDB_NAMESPACE::WriteOptions;

 using std::shared_ptr;
 using std::vector;
@@ -452,7 +452,7 @@ struct rocksdb_slicetransform_t : public SliceTransform {
 };

 struct rocksdb_universal_compaction_options_t {
-  rocksdb::CompactionOptionsUniversal *rep;
+  ROCKSDB_NAMESPACE::CompactionOptionsUniversal* rep;
 };

 static bool SaveError(char** errptr, const Status& s) {
@@ -494,8 +494,9 @@ rocksdb_t* rocksdb_open_with_ttl(
    const char* name,
    int ttl,
    char** errptr) {
-  rocksdb::DBWithTTL* db;
-  if (SaveError(errptr, rocksdb::DBWithTTL::Open(options->rep, std::string(name), &db, ttl))) {
+  ROCKSDB_NAMESPACE::DBWithTTL* db;
+  if (SaveError(errptr, ROCKSDB_NAMESPACE::DBWithTTL::Open(
+                            options->rep, std::string(name), &db, ttl))) {
    return nullptr;
  }
  rocksdb_t* result = new rocksdb_t;
@@ -664,17 +665,15 @@ void rocksdb_close(rocksdb_t* db) {
 }

 void rocksdb_options_set_uint64add_merge_operator(rocksdb_options_t* opt) {
-  opt->rep.merge_operator = rocksdb::MergeOperators::CreateUInt64AddOperator();
+  opt->rep.merge_operator =
+      ROCKSDB_NAMESPACE::MergeOperators::CreateUInt64AddOperator();
 }

 rocksdb_t* rocksdb_open_column_families(
-    const rocksdb_options_t* db_options,
-    const char* name,
-    int num_column_families,
-    const char* const* column_family_names,
+    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,
-    char** errptr) {
+    rocksdb_column_family_handle_t** column_family_handles, char** errptr) {
  std::vector<ColumnFamilyDescriptor> column_families;
  for (int i = 0; i < num_column_families; i++) {
    column_families.push_back(ColumnFamilyDescriptor(
@@ -700,14 +699,11 @@ rocksdb_t* rocksdb_open_column_families(
 }

 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,
+    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,
-    unsigned char error_if_log_file_exist,
-    char** errptr) {
+    unsigned char error_if_log_file_exist, char** errptr) {
  std::vector<ColumnFamilyDescriptor> column_families;
  for (int i = 0; i < num_column_families; i++) {
    column_families.push_back(ColumnFamilyDescriptor(
@@ -2084,7 +2080,7 @@ void rocksdb_options_set_block_based_table_factory(
    rocksdb_block_based_table_options_t* table_options) {
  if (table_options) {
    opt->rep.table_factory.reset(
-        rocksdb::NewBlockBasedTableFactory(table_options->rep));
+        ROCKSDB_NAMESPACE::NewBlockBasedTableFactory(table_options->rep));
  }
 }

@@ -2128,7 +2124,7 @@ void rocksdb_options_set_cuckoo_table_factory(
    rocksdb_cuckoo_table_options_t* table_options) {
  if (table_options) {
    opt->rep.table_factory.reset(
-        rocksdb::NewCuckooTableFactory(table_options->rep));
+        ROCKSDB_NAMESPACE::NewCuckooTableFactory(table_options->rep));
  }
 }

@@ -2318,7 +2314,7 @@ void rocksdb_options_set_max_bytes_for_level_multiplier_additional(
 }

 void rocksdb_options_enable_statistics(rocksdb_options_t* opt) {
-  opt->rep.statistics = rocksdb::CreateDBStatistics();
+  opt->rep.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
 }

 void rocksdb_options_set_skip_stats_update_on_db_open(rocksdb_options_t* opt,
@@ -2326,6 +2322,11 @@ void rocksdb_options_set_skip_stats_update_on_db_open(rocksdb_options_t* opt,
  opt->rep.skip_stats_update_on_db_open = val;
 }

+void rocksdb_options_set_skip_checking_sst_file_sizes_on_db_open(
+    rocksdb_options_t* opt, unsigned char val) {
+  opt->rep.skip_checking_sst_file_sizes_on_db_open = val;
+}
+
 void rocksdb_options_set_num_levels(rocksdb_options_t* opt, int n) {
  opt->rep.num_levels = n;
 }
@@ -2469,16 +2470,20 @@ void rocksdb_options_set_access_hint_on_compaction_start(
    rocksdb_options_t* opt, int v) {
  switch(v) {
    case 0:
-      opt->rep.access_hint_on_compaction_start = rocksdb::Options::NONE;
+      opt->rep.access_hint_on_compaction_start =
+          ROCKSDB_NAMESPACE::Options::NONE;
      break;
    case 1:
-      opt->rep.access_hint_on_compaction_start = rocksdb::Options::NORMAL;
+      opt->rep.access_hint_on_compaction_start =
+          ROCKSDB_NAMESPACE::Options::NORMAL;
      break;
    case 2:
-      opt->rep.access_hint_on_compaction_start = rocksdb::Options::SEQUENTIAL;
+      opt->rep.access_hint_on_compaction_start =
+          ROCKSDB_NAMESPACE::Options::SEQUENTIAL;
      break;
    case 3:
-      opt->rep.access_hint_on_compaction_start = rocksdb::Options::WILLNEED;
+      opt->rep.access_hint_on_compaction_start =
+          ROCKSDB_NAMESPACE::Options::WILLNEED;
      break;
  }
 }
@@ -2644,7 +2649,7 @@ void rocksdb_options_prepare_for_bulk_load(rocksdb_options_t* opt) {
 }

 void rocksdb_options_set_memtable_vector_rep(rocksdb_options_t *opt) {
-  opt->rep.memtable_factory.reset(new rocksdb::VectorRepFactory);
+  opt->rep.memtable_factory.reset(new ROCKSDB_NAMESPACE::VectorRepFactory);
 }

 void rocksdb_options_set_memtable_prefix_bloom_size_ratio(
@@ -2660,26 +2665,29 @@ void rocksdb_options_set_memtable_huge_page_size(rocksdb_options_t* opt,
 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) {
-  rocksdb::MemTableRepFactory* factory = rocksdb::NewHashSkipListRepFactory(
-      bucket_count, skiplist_height, skiplist_branching_factor);
+  ROCKSDB_NAMESPACE::MemTableRepFactory* factory =
+      ROCKSDB_NAMESPACE::NewHashSkipListRepFactory(
+          bucket_count, skiplist_height, skiplist_branching_factor);
  opt->rep.memtable_factory.reset(factory);
 }

 void rocksdb_options_set_hash_link_list_rep(
    rocksdb_options_t *opt, size_t bucket_count) {
-  opt->rep.memtable_factory.reset(rocksdb::NewHashLinkListRepFactory(bucket_count));
+  opt->rep.memtable_factory.reset(
+      ROCKSDB_NAMESPACE::NewHashLinkListRepFactory(bucket_count));
 }

 void rocksdb_options_set_plain_table_factory(
    rocksdb_options_t *opt, uint32_t user_key_len, int bloom_bits_per_key,
    double hash_table_ratio, size_t index_sparseness) {
-  rocksdb::PlainTableOptions options;
+  ROCKSDB_NAMESPACE::PlainTableOptions options;
  options.user_key_len = user_key_len;
  options.bloom_bits_per_key = bloom_bits_per_key;
  options.hash_table_ratio = hash_table_ratio;
  options.index_sparseness = index_sparseness;

-  rocksdb::TableFactory* factory = rocksdb::NewPlainTableFactory(options);
+  ROCKSDB_NAMESPACE::TableFactory* factory =
+      ROCKSDB_NAMESPACE::NewPlainTableFactory(options);
  opt->rep.table_factory.reset(factory);
 }

@@ -2709,7 +2717,8 @@ void rocksdb_options_set_report_bg_io_stats(
 }

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

 void rocksdb_options_set_universal_compaction_options(rocksdb_options_t *opt, rocksdb_universal_compaction_options_t *uco) {
@@ -2723,7 +2732,7 @@ void rocksdb_options_set_fifo_compaction_options(
 }

 char *rocksdb_options_statistics_get_string(rocksdb_options_t *opt) {
-  rocksdb::Statistics *statistics = opt->rep.statistics.get();
+  ROCKSDB_NAMESPACE::Statistics* statistics = opt->rep.statistics.get();
  if (statistics) {
    return strdup(statistics->ToString().c_str());
  }
@@ -2736,10 +2745,9 @@ void rocksdb_options_set_ratelimiter(rocksdb_options_t *opt, rocksdb_ratelimiter
  }
 }

-void rocksdb_options_set_atomic_flush(
-  rocksdb_options_t *opt,
-  unsigned char atomic_flush) {
-    opt->rep.atomic_flush = atomic_flush;
+void rocksdb_options_set_atomic_flush(rocksdb_options_t* opt,
+                                      unsigned char atomic_flush) {
+  opt->rep.atomic_flush = atomic_flush;
 }

 rocksdb_ratelimiter_t* rocksdb_ratelimiter_create(
@@ -2757,6 +2765,12 @@ void rocksdb_ratelimiter_destroy(rocksdb_ratelimiter_t *limiter) {
  delete limiter;
 }

+void rocksdb_options_set_row_cache(rocksdb_options_t* opt, rocksdb_cache_t* cache) {
+  if(cache) {
+    opt->rep.row_cache = cache->rep;
+  }
+}
+
 void rocksdb_set_perf_level(int v) {
  PerfLevel level = static_cast<PerfLevel>(v);
  SetPerfLevel(level);
@@ -2764,7 +2778,7 @@ void rocksdb_set_perf_level(int v) {

 rocksdb_perfcontext_t* rocksdb_perfcontext_create() {
  rocksdb_perfcontext_t* context = new rocksdb_perfcontext_t;
-  context->rep = rocksdb::get_perf_context();
+  context->rep = ROCKSDB_NAMESPACE::get_perf_context();
  return context;
 }

@@ -3066,11 +3080,12 @@ rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_format(int bits_per_ke
      return rep_->KeyMayMatch(key, filter);
    }
    // No need to override GetFilterBitsBuilder if this one is overridden
-    rocksdb::FilterBitsBuilder* GetBuilderWithContext(
-        const rocksdb::FilterBuildingContext& context) const override {
+    ROCKSDB_NAMESPACE::FilterBitsBuilder* GetBuilderWithContext(
+        const ROCKSDB_NAMESPACE::FilterBuildingContext& context)
+        const override {
      return rep_->GetBuilderWithContext(context);
    }
-    rocksdb::FilterBitsReader* GetFilterBitsReader(
+    ROCKSDB_NAMESPACE::FilterBitsReader* GetFilterBitsReader(
        const Slice& contents) const override {
      return rep_->GetFilterBitsReader(contents);
    }
@@ -3172,7 +3187,7 @@ void rocksdb_readoptions_set_iterate_lower_bound(

 void rocksdb_readoptions_set_read_tier(
    rocksdb_readoptions_t* opt, int v) {
-  opt->rep.read_tier = static_cast<rocksdb::ReadTier>(v);
+  opt->rep.read_tier = static_cast<ROCKSDB_NAMESPACE::ReadTier>(v);
 }

 void rocksdb_readoptions_set_tailing(
@@ -3344,7 +3359,7 @@ rocksdb_env_t* rocksdb_create_default_env() {

 rocksdb_env_t* rocksdb_create_mem_env() {
  rocksdb_env_t* result = new rocksdb_env_t;
-  result->rep = rocksdb::NewMemEnv(Env::Default());
+  result->rep = ROCKSDB_NAMESPACE::NewMemEnv(Env::Default());
  result->is_default = false;
  return result;
 }
@@ -3559,7 +3574,7 @@ struct Wrapper : public rocksdb_slicetransform_t {

 rocksdb_slicetransform_t* rocksdb_slicetransform_create_fixed_prefix(size_t prefixLen) {
  Wrapper* wrapper = new Wrapper;
-  wrapper->rep_ = rocksdb::NewFixedPrefixTransform(prefixLen);
+  wrapper->rep_ = ROCKSDB_NAMESPACE::NewFixedPrefixTransform(prefixLen);
  wrapper->state_ = nullptr;
  wrapper->destructor_ = &Wrapper::DoNothing;
  return wrapper;
@@ -3567,7 +3582,7 @@ rocksdb_slicetransform_t* rocksdb_slicetransform_create_fixed_prefix(size_t pref

 rocksdb_slicetransform_t* rocksdb_slicetransform_create_noop() {
  Wrapper* wrapper = new Wrapper;
-  wrapper->rep_ = rocksdb::NewNoopTransform();
+  wrapper->rep_ = ROCKSDB_NAMESPACE::NewNoopTransform();
  wrapper->state_ = nullptr;
  wrapper->destructor_ = &Wrapper::DoNothing;
  return wrapper;
@@ -3575,7 +3590,7 @@ rocksdb_slicetransform_t* rocksdb_slicetransform_create_noop() {

 rocksdb_universal_compaction_options_t* rocksdb_universal_compaction_options_create() {
  rocksdb_universal_compaction_options_t* result = new rocksdb_universal_compaction_options_t;
-  result->rep = new rocksdb::CompactionOptionsUniversal;
+  result->rep = new ROCKSDB_NAMESPACE::CompactionOptionsUniversal;
  return result;
 }

@@ -3606,7 +3621,8 @@ void rocksdb_universal_compaction_options_set_compression_size_percent(

 void rocksdb_universal_compaction_options_set_stop_style(
  rocksdb_universal_compaction_options_t* uco, int style) {
-  uco->rep->stop_style = static_cast<rocksdb::CompactionStopStyle>(style);
+  uco->rep->stop_style =
+      static_cast<ROCKSDB_NAMESPACE::CompactionStopStyle>(style);
 }

 void rocksdb_universal_compaction_options_destroy(
@@ -3636,7 +3652,7 @@ void rocksdb_options_set_min_level_to_compress(rocksdb_options_t* opt, int level
    assert(level <= opt->rep.num_levels);
    opt->rep.compression_per_level.resize(opt->rep.num_levels);
    for (int i = 0; i < level; i++) {
-      opt->rep.compression_per_level[i] = rocksdb::kNoCompression;
+      opt->rep.compression_per_level[i] = ROCKSDB_NAMESPACE::kNoCompression;
    }
    for (int i = level; i < opt->rep.num_levels; i++) {
      opt->rep.compression_per_level[i] = opt->rep.compression;
@@ -4338,7 +4354,8 @@ const char* rocksdb_pinnableslice_value(const rocksdb_pinnableslice_t* v,
  return v->rep.data();
 }

-// container to keep databases and caches in order to use rocksdb::MemoryUtil
+// container to keep databases and caches in order to use
+// ROCKSDB_NAMESPACE::MemoryUtil
 struct rocksdb_memory_consumers_t {
  std::vector<rocksdb_t*> dbs;
  std::unordered_set<rocksdb_cache_t*> caches;
@@ -4366,7 +4383,7 @@ void rocksdb_memory_consumers_destroy(rocksdb_memory_consumers_t* consumers) {
  delete consumers;
 }

-// contains memory usage statistics provided by rocksdb::MemoryUtil
+// contains memory usage statistics provided by ROCKSDB_NAMESPACE::MemoryUtil
 struct rocksdb_memory_usage_t {
  uint64_t mem_table_total;
  uint64_t mem_table_unflushed;
@@ -4388,7 +4405,7 @@ rocksdb_memory_usage_t* rocksdb_approximate_memory_usage_create(
    cache_set.insert(const_cast<const Cache*>(cache->rep.get()));
  }

-  std::map<rocksdb::MemoryUtil::UsageType, uint64_t> usage_by_type;
+  std::map<ROCKSDB_NAMESPACE::MemoryUtil::UsageType, uint64_t> usage_by_type;

  auto status = MemoryUtil::GetApproximateMemoryUsageByType(dbs, cache_set,
                                                            &usage_by_type);
@@ -4424,6 +4441,16 @@ uint64_t rocksdb_approximate_memory_usage_get_cache_total(
  return memory_usage->cache_total;
 }

+void rocksdb_options_set_dump_malloc_stats(rocksdb_options_t* opt,
+                                           unsigned char val) {
+  opt->rep.dump_malloc_stats = val;
+}
+
+void rocksdb_options_set_memtable_whole_key_filtering(rocksdb_options_t* opt,
+                                                      unsigned char val) {
+  opt->rep.memtable_whole_key_filtering = val;
+}
+
 // deletes container with memory usage estimates
 void rocksdb_approximate_memory_usage_destroy(rocksdb_memory_usage_t* usage) {
  delete usage;
@@ -835,23 +835,6 @@ int main(int argc, char** argv) {
    rocksdb_writebatch_wi_iterate(wbi, &pos, CheckPut, CheckDel);
    CheckCondition(pos == 3);
    rocksdb_writebatch_wi_clear(wbi);
-    rocksdb_writebatch_wi_put(wbi, "bar", 3, "b", 1);
-    rocksdb_writebatch_wi_put(wbi, "bay", 3, "d", 1);
-    rocksdb_writebatch_wi_delete_range(wbi, "bar", 3, "bay", 3);
-    rocksdb_write_writebatch_wi(db, woptions, wbi, &err);
-    CheckNoError(err);
-    CheckGet(db, roptions, "bar", NULL);
-    CheckGet(db, roptions, "bay", "d");
-    rocksdb_writebatch_wi_clear(wbi);
-    const char* start_list[1] = {"bay"};
-    const size_t start_sizes[1] = {3};
-    const char* end_list[1] = {"baz"};
-    const size_t end_sizes[1] = {3};
-    rocksdb_writebatch_wi_delete_rangev(wbi, 1, start_list, start_sizes, end_list,
-                                     end_sizes);
-    rocksdb_write_writebatch_wi(db, woptions, wbi, &err);
-    CheckNoError(err);
-    CheckGet(db, roptions, "bay", NULL);
    rocksdb_writebatch_wi_destroy(wbi);
  }

@@ -36,7 +36,7 @@
 #include "util/autovector.h"
 #include "util/compression.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 ColumnFamilyHandleImpl::ColumnFamilyHandleImpl(
    ColumnFamilyData* column_family_data, DBImpl* db, InstrumentedMutex* mutex)
@@ -471,6 +471,17 @@ void SuperVersionUnrefHandle(void* ptr) {
 }
 }  // anonymous namespace

+std::vector<std::string> ColumnFamilyData::GetDbPaths() const {
+  std::vector<std::string> paths;
+  paths.reserve(ioptions_.cf_paths.size());
+  for (const DbPath& db_path : ioptions_.cf_paths) {
+    paths.emplace_back(db_path.path);
+  }
+  return paths;
+}
+
+const uint32_t ColumnFamilyData::kDummyColumnFamilyDataId = port::kMaxUint32;
+
 ColumnFamilyData::ColumnFamilyData(
    uint32_t id, const std::string& name, Version* _dummy_versions,
    Cache* _table_cache, WriteBufferManager* write_buffer_manager,
@@ -507,7 +518,23 @@ ColumnFamilyData::ColumnFamilyData(
      queued_for_compaction_(false),
      prev_compaction_needed_bytes_(0),
      allow_2pc_(db_options.allow_2pc),
-      last_memtable_id_(0) {
+      last_memtable_id_(0),
+      db_paths_registered_(false) {
+  if (id_ != kDummyColumnFamilyDataId) {
+    // TODO(cc): RegisterDbPaths can be expensive, considering moving it
+    // outside of this constructor which might be called with db mutex held.
+    // TODO(cc): considering using ioptions_.fs, currently some tests rely on
+    // EnvWrapper, that's the main reason why we use env here.
+    Status s = ioptions_.env->RegisterDbPaths(GetDbPaths());
+    if (s.ok()) {
+      db_paths_registered_ = true;
+    } else {
+      ROCKS_LOG_ERROR(
+          ioptions_.info_log,
+          "Failed to register data paths of column family (id: %d, name: %s)",
+          id_, name_.c_str());
+    }
+  }
  Ref();

  // Convert user defined table properties collector factories to internal ones.
@@ -601,6 +628,18 @@ ColumnFamilyData::~ColumnFamilyData() {
  for (MemTable* m : to_delete) {
    delete m;
  }
+
+  if (db_paths_registered_) {
+    // TODO(cc): considering using ioptions_.fs, currently some tests rely on
+    // EnvWrapper, that's the main reason why we use env here.
+    Status s = ioptions_.env->UnregisterDbPaths(GetDbPaths());
+    if (!s.ok()) {
+      ROCKS_LOG_ERROR(
+          ioptions_.info_log,
+          "Failed to unregister data paths of column family (id: %d, name: %s)",
+          id_, name_.c_str());
+    }
+  }
 }

 bool ColumnFamilyData::UnrefAndTryDelete() {
@@ -1320,23 +1359,32 @@ Env::WriteLifeTimeHint ColumnFamilyData::CalculateSSTWriteHint(int level) {
                            static_cast<int>(Env::WLTH_MEDIUM));
 }

-Status ColumnFamilyData::AddDirectories() {
+Status ColumnFamilyData::AddDirectories(
+    std::map<std::string, std::shared_ptr<FSDirectory>>* created_dirs) {
  Status s;
+  assert(created_dirs != nullptr);
  assert(data_dirs_.empty());
  for (auto& p : ioptions_.cf_paths) {
-    std::unique_ptr<Directory> path_directory;
-    s = DBImpl::CreateAndNewDirectory(ioptions_.env, p.path, &path_directory);
-    if (!s.ok()) {
-      return s;
+    auto existing_dir = created_dirs->find(p.path);
+
+    if (existing_dir == created_dirs->end()) {
+      std::unique_ptr<FSDirectory> path_directory;
+      s = DBImpl::CreateAndNewDirectory(ioptions_.fs, p.path, &path_directory);
+      if (!s.ok()) {
+        return s;
+      }
+      assert(path_directory != nullptr);
+      data_dirs_.emplace_back(path_directory.release());
+      (*created_dirs)[p.path] = data_dirs_.back();
+    } else {
+      data_dirs_.emplace_back(existing_dir->second);
    }
-    assert(path_directory != nullptr);
-    data_dirs_.emplace_back(path_directory.release());
  }
  assert(data_dirs_.size() == ioptions_.cf_paths.size());
  return s;
 }

-Directory* ColumnFamilyData::GetDataDir(size_t path_id) const {
+FSDirectory* ColumnFamilyData::GetDataDir(size_t path_id) const {
  if (data_dirs_.empty()) {
    return nullptr;
  }
@@ -1349,20 +1397,21 @@ ColumnFamilySet::ColumnFamilySet(const std::string& dbname,
                                 const ImmutableDBOptions* db_options,
                                 const FileOptions& file_options,
                                 Cache* table_cache,
-                                 WriteBufferManager* write_buffer_manager,
-                                 WriteController* write_controller,
+                                 WriteBufferManager* _write_buffer_manager,
+                                 WriteController* _write_controller,
                                 BlockCacheTracer* const block_cache_tracer)
    : max_column_family_(0),
      dummy_cfd_(new ColumnFamilyData(
-          0, "", nullptr, nullptr, nullptr, ColumnFamilyOptions(), *db_options,
-          file_options, nullptr, block_cache_tracer)),
+          ColumnFamilyData::kDummyColumnFamilyDataId, "", nullptr, nullptr,
+          nullptr, ColumnFamilyOptions(), *db_options, file_options, nullptr,
+          block_cache_tracer)),
      default_cfd_cache_(nullptr),
      db_name_(dbname),
      db_options_(db_options),
      file_options_(file_options),
      table_cache_(table_cache),
-      write_buffer_manager_(write_buffer_manager),
-      write_controller_(write_controller),
+      write_buffer_manager_(_write_buffer_manager),
+      write_controller_(_write_controller),
      block_cache_tracer_(block_cache_tracer) {
  // initialize linked list
  dummy_cfd_->prev_ = dummy_cfd_;
@@ -1511,4 +1560,4 @@ const Comparator* GetColumnFamilyUserComparator(
  return nullptr;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -27,7 +27,7 @@
 #include "trace_replay/block_cache_tracer.h"
 #include "util/thread_local.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class Version;
 class VersionSet;
@@ -497,14 +497,18 @@ class ColumnFamilyData {

  Env::WriteLifeTimeHint CalculateSSTWriteHint(int level);

-  Status AddDirectories();
+  // created_dirs remembers directory created, so that we don't need to call
+  // the same data creation operation again.
+  Status AddDirectories(
+      std::map<std::string, std::shared_ptr<FSDirectory>>* created_dirs);

-  Directory* GetDataDir(size_t path_id) const;
+  FSDirectory* GetDataDir(size_t path_id) const;

  ThreadLocalPtr* TEST_GetLocalSV() { return local_sv_.get(); }

 private:
  friend class ColumnFamilySet;
+  static const uint32_t kDummyColumnFamilyDataId;
  ColumnFamilyData(uint32_t id, const std::string& name,
                   Version* dummy_versions, Cache* table_cache,
                   WriteBufferManager* write_buffer_manager,
@@ -514,6 +518,8 @@ class ColumnFamilyData {
                   ColumnFamilySet* column_family_set,
                   BlockCacheTracer* const block_cache_tracer);

+  std::vector<std::string> GetDbPaths() const;
+
  uint32_t id_;
  const std::string name_;
  Version* dummy_versions_;  // Head of circular doubly-linked list of versions.
@@ -589,7 +595,9 @@ class ColumnFamilyData {
  std::atomic<uint64_t> last_memtable_id_;

  // Directories corresponding to cf_paths.
-  std::vector<std::unique_ptr<Directory>> data_dirs_;
+  std::vector<std::shared_ptr<FSDirectory>> data_dirs_;
+
+  bool db_paths_registered_;
 };

 // ColumnFamilySet has interesting thread-safety requirements
@@ -639,8 +647,8 @@ class ColumnFamilySet {
  ColumnFamilySet(const std::string& dbname,
                  const ImmutableDBOptions* db_options,
                  const FileOptions& file_options, Cache* table_cache,
-                  WriteBufferManager* write_buffer_manager,
-                  WriteController* write_controller,
+                  WriteBufferManager* _write_buffer_manager,
+                  WriteController* _write_controller,
                  BlockCacheTracer* const block_cache_tracer);
  ~ColumnFamilySet();

@@ -670,6 +678,10 @@ class ColumnFamilySet {

  Cache* get_table_cache() { return table_cache_; }

+  WriteBufferManager* write_buffer_manager() { return write_buffer_manager_; }
+
+  WriteController* write_controller() { return write_controller_; }
+
 private:
  friend class ColumnFamilyData;
  // helper function that gets called from cfd destructor
@@ -751,4 +763,4 @@ extern uint32_t GetColumnFamilyID(ColumnFamilyHandle* column_family);
 extern const Comparator* GetColumnFamilyUserComparator(
    ColumnFamilyHandle* column_family);

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -8,16 +8,16 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.

 #include <algorithm>
-#include <vector>
 #include <string>
 #include <thread>
+#include <vector>

 #include "db/db_impl/db_impl.h"
 #include "db/db_test_util.h"
-#include "memtable/hash_skiplist_rep.h"
 #include "options/options_parser.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
+#include "rocksdb/convenience.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
 #include "rocksdb/iterator.h"
@@ -30,7 +30,7 @@
 #include "util/string_util.h"
 #include "utilities/merge_operators.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 static const int kValueSize = 1000;

@@ -91,7 +91,7 @@ class ColumnFamilyTestBase : public testing::Test {
      column_families.push_back(cfdescriptor);
    }
    Close();
-    rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
    Destroy(column_families);
    delete env_;
  }
@@ -287,7 +287,9 @@ class ColumnFamilyTestBase : public testing::Test {
      // Verify the CF options of the returned CF handle.
      ColumnFamilyDescriptor desc;
      ASSERT_OK(handles_[cfi]->GetDescriptor(&desc));
-      RocksDBOptionsParser::VerifyCFOptions(desc.options, current_cf_opt);
+      RocksDBOptionsParser::VerifyCFOptions(ConfigOptions(), desc.options,
+                                            current_cf_opt);
+
 #endif  // !ROCKSDB_LITE
      cfi++;
    }
@@ -595,14 +597,15 @@ TEST_P(ColumnFamilyTest, DontReuseColumnFamilyID) {
 TEST_P(ColumnFamilyTest, CreateCFRaceWithGetAggProperty) {
  Open();

-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"DBImpl::WriteOptionsFile:1",
        "ColumnFamilyTest.CreateCFRaceWithGetAggProperty:1"},
       {"ColumnFamilyTest.CreateCFRaceWithGetAggProperty:2",
        "DBImpl::WriteOptionsFile:2"}});
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

-  rocksdb::port::Thread thread([&] { CreateColumnFamilies({"one"}); });
+  ROCKSDB_NAMESPACE::port::Thread thread(
+      [&] { CreateColumnFamilies({"one"}); });

  TEST_SYNC_POINT("ColumnFamilyTest.CreateCFRaceWithGetAggProperty:1");
  uint64_t pv;
@@ -611,7 +614,7 @@ TEST_P(ColumnFamilyTest, CreateCFRaceWithGetAggProperty) {

  thread.join();

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
 }
 #endif  // !ROCKSDB_LITE

@@ -820,7 +823,7 @@ TEST_P(ColumnFamilyTest, BulkAddDrop) {
 }

 TEST_P(ColumnFamilyTest, DropTest) {
-  // first iteration - dont reopen DB before dropping
+  // first iteration - don't reopen DB before dropping
  // second iteration - reopen DB before dropping
  for (int iter = 0; iter < 2; ++iter) {
    Open({"default"});
@@ -1253,13 +1256,14 @@ TEST_P(ColumnFamilyTest, DifferentWriteBufferSizes) {
 // #endif  // !ROCKSDB_LITE

 class TestComparator : public Comparator {
-  int Compare(const rocksdb::Slice& /*a*/,
-              const rocksdb::Slice& /*b*/) const override {
+  int Compare(const ROCKSDB_NAMESPACE::Slice& /*a*/,
+              const ROCKSDB_NAMESPACE::Slice& /*b*/) const override {
    return 0;
  }
  const char* Name() const override { return "Test"; }
-  void FindShortestSeparator(std::string* /*start*/,
-                             const rocksdb::Slice& /*limit*/) const override {}
+  void FindShortestSeparator(
+      std::string* /*start*/,
+      const ROCKSDB_NAMESPACE::Slice& /*limit*/) const override {}
  void FindShortSuccessor(std::string* /*key*/) const override {}
 };

@@ -1427,11 +1431,11 @@ TEST_P(ColumnFamilyTest, MultipleManualCompactions) {
    AssertFilesPerLevel(ToString(i + 1), 1);
  }
  bool cf_1_1 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::MultiManual:4", "ColumnFamilyTest::MultiManual:1"},
       {"ColumnFamilyTest::MultiManual:2", "ColumnFamilyTest::MultiManual:5"},
       {"ColumnFamilyTest::MultiManual:2", "ColumnFamilyTest::MultiManual:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          TEST_SYNC_POINT("ColumnFamilyTest::MultiManual:4");
@@ -1440,7 +1444,7 @@ TEST_P(ColumnFamilyTest, MultipleManualCompactions) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
  std::vector<port::Thread> threads;
  threads.emplace_back([&] {
    CompactRangeOptions compact_options;
@@ -1483,8 +1487,8 @@ TEST_P(ColumnFamilyTest, MultipleManualCompactions) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
  Close();
 }

@@ -1523,11 +1527,11 @@ TEST_P(ColumnFamilyTest, AutomaticAndManualCompactions) {
      dbfull()->TEST_write_controler().GetCompactionPressureToken();

  bool cf_1_1 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::AutoManual:4", "ColumnFamilyTest::AutoManual:1"},
       {"ColumnFamilyTest::AutoManual:2", "ColumnFamilyTest::AutoManual:5"},
       {"ColumnFamilyTest::AutoManual:2", "ColumnFamilyTest::AutoManual:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          cf_1_1 = false;
@@ -1535,7 +1539,7 @@ TEST_P(ColumnFamilyTest, AutomaticAndManualCompactions) {
          TEST_SYNC_POINT("ColumnFamilyTest::AutoManual:3");
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
  // SETUP column family "one" -- universal style
  for (int i = 0; i < one.level0_file_num_compaction_trigger; ++i) {
    PutRandomData(1, 10, 12000, true);
@@ -1553,7 +1557,7 @@ TEST_P(ColumnFamilyTest, AutomaticAndManualCompactions) {
    WaitForFlush(2);
    AssertFilesPerLevel(ToString(i + 1), 2);
  }
-  rocksdb::port::Thread threads([&] {
+  ROCKSDB_NAMESPACE::port::Thread threads([&] {
    CompactRangeOptions compact_options;
    compact_options.exclusive_manual_compaction = false;
    ASSERT_OK(
@@ -1580,8 +1584,8 @@ TEST_P(ColumnFamilyTest, AutomaticAndManualCompactions) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }

 TEST_P(ColumnFamilyTest, ManualAndAutomaticCompactions) {
@@ -1627,11 +1631,11 @@ TEST_P(ColumnFamilyTest, ManualAndAutomaticCompactions) {
  }
  bool cf_1_1 = true;
  bool cf_1_2 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::ManualAuto:4", "ColumnFamilyTest::ManualAuto:1"},
       {"ColumnFamilyTest::ManualAuto:5", "ColumnFamilyTest::ManualAuto:2"},
       {"ColumnFamilyTest::ManualAuto:2", "ColumnFamilyTest::ManualAuto:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          TEST_SYNC_POINT("ColumnFamilyTest::ManualAuto:4");
@@ -1643,8 +1647,8 @@ TEST_P(ColumnFamilyTest, ManualAndAutomaticCompactions) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-  rocksdb::port::Thread threads([&] {
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::port::Thread threads([&] {
    CompactRangeOptions compact_options;
    compact_options.exclusive_manual_compaction = false;
    ASSERT_OK(
@@ -1679,8 +1683,8 @@ TEST_P(ColumnFamilyTest, ManualAndAutomaticCompactions) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }

 TEST_P(ColumnFamilyTest, SameCFManualManualCompactions) {
@@ -1721,13 +1725,13 @@ TEST_P(ColumnFamilyTest, SameCFManualManualCompactions) {
  }
  bool cf_1_1 = true;
  bool cf_1_2 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::ManualManual:4", "ColumnFamilyTest::ManualManual:2"},
       {"ColumnFamilyTest::ManualManual:4", "ColumnFamilyTest::ManualManual:5"},
       {"ColumnFamilyTest::ManualManual:1", "ColumnFamilyTest::ManualManual:2"},
       {"ColumnFamilyTest::ManualManual:1",
        "ColumnFamilyTest::ManualManual:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          TEST_SYNC_POINT("ColumnFamilyTest::ManualManual:4");
@@ -1739,8 +1743,8 @@ TEST_P(ColumnFamilyTest, SameCFManualManualCompactions) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-  rocksdb::port::Thread threads([&] {
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::port::Thread threads([&] {
    CompactRangeOptions compact_options;
    compact_options.exclusive_manual_compaction = true;
    ASSERT_OK(
@@ -1760,7 +1764,7 @@ TEST_P(ColumnFamilyTest, SameCFManualManualCompactions) {
                        1);
  }

-  rocksdb::port::Thread threads1([&] {
+  ROCKSDB_NAMESPACE::port::Thread threads1([&] {
    CompactRangeOptions compact_options;
    compact_options.exclusive_manual_compaction = false;
    ASSERT_OK(
@@ -1781,8 +1785,8 @@ TEST_P(ColumnFamilyTest, SameCFManualManualCompactions) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }

 TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactions) {
@@ -1823,12 +1827,12 @@ TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactions) {
  }
  bool cf_1_1 = true;
  bool cf_1_2 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::ManualAuto:4", "ColumnFamilyTest::ManualAuto:2"},
       {"ColumnFamilyTest::ManualAuto:4", "ColumnFamilyTest::ManualAuto:5"},
       {"ColumnFamilyTest::ManualAuto:1", "ColumnFamilyTest::ManualAuto:2"},
       {"ColumnFamilyTest::ManualAuto:1", "ColumnFamilyTest::ManualAuto:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          TEST_SYNC_POINT("ColumnFamilyTest::ManualAuto:4");
@@ -1840,8 +1844,8 @@ TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactions) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-  rocksdb::port::Thread threads([&] {
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::port::Thread threads([&] {
    CompactRangeOptions compact_options;
    compact_options.exclusive_manual_compaction = false;
    ASSERT_OK(
@@ -1874,8 +1878,8 @@ TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactions) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }

 TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactionsLevel) {
@@ -1916,14 +1920,14 @@ TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactionsLevel) {
  }
  bool cf_1_1 = true;
  bool cf_1_2 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::ManualAuto:4", "ColumnFamilyTest::ManualAuto:2"},
       {"ColumnFamilyTest::ManualAuto:4", "ColumnFamilyTest::ManualAuto:5"},
       {"ColumnFamilyTest::ManualAuto:3", "ColumnFamilyTest::ManualAuto:2"},
       {"LevelCompactionPicker::PickCompactionBySize:0",
        "ColumnFamilyTest::ManualAuto:3"},
       {"ColumnFamilyTest::ManualAuto:1", "ColumnFamilyTest::ManualAuto:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          TEST_SYNC_POINT("ColumnFamilyTest::ManualAuto:4");
@@ -1935,8 +1939,8 @@ TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactionsLevel) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-  rocksdb::port::Thread threads([&] {
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::port::Thread threads([&] {
    CompactRangeOptions compact_options;
    compact_options.exclusive_manual_compaction = false;
    ASSERT_OK(
@@ -1967,8 +1971,8 @@ TEST_P(ColumnFamilyTest, SameCFManualAutomaticCompactionsLevel) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }

 // In this test, we generate enough files to trigger automatic compactions.
@@ -2008,12 +2012,12 @@ TEST_P(ColumnFamilyTest, SameCFAutomaticManualCompactions) {

  bool cf_1_1 = true;
  bool cf_1_2 = true;
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"ColumnFamilyTest::AutoManual:4", "ColumnFamilyTest::AutoManual:2"},
       {"ColumnFamilyTest::AutoManual:4", "ColumnFamilyTest::AutoManual:5"},
       {"CompactionPicker::CompactRange:Conflict",
        "ColumnFamilyTest::AutoManual:3"}});
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial:AfterRun", [&](void* /*arg*/) {
        if (cf_1_1) {
          TEST_SYNC_POINT("ColumnFamilyTest::AutoManual:4");
@@ -2025,7 +2029,7 @@ TEST_P(ColumnFamilyTest, SameCFAutomaticManualCompactions) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  // SETUP column family "one" -- universal style
  for (int i = 0; i < one.level0_file_num_compaction_trigger; ++i) {
@@ -2059,8 +2063,8 @@ TEST_P(ColumnFamilyTest, SameCFAutomaticManualCompactions) {
    ASSERT_NE("NOT_FOUND", Get(1, *key_iter));
    key_iter++;
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }
 #endif  // !ROCKSDB_LITE

@@ -2415,7 +2419,7 @@ TEST_P(ColumnFamilyTest, FlushAndDropRaceCondition) {
  options.write_buffer_size = 100000;  // small write buffer size
  Reopen({options, options});

-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"VersionSet::LogAndApply::ColumnFamilyDrop:0",
        "FlushJob::WriteLevel0Table"},
       {"VersionSet::LogAndApply::ColumnFamilyDrop:1",
@@ -2423,7 +2427,7 @@ TEST_P(ColumnFamilyTest, FlushAndDropRaceCondition) {
       {"FlushJob::InstallResults",
        "VersionSet::LogAndApply::ColumnFamilyDrop:2"}});

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
  test::SleepingBackgroundTask sleeping_task;

  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task,
@@ -2502,21 +2506,21 @@ TEST_P(ColumnFamilyTest, CreateAndDropRace) {

  auto main_thread_id = std::this_thread::get_id();

-  rocksdb::SyncPoint::GetInstance()->SetCallBack("PersistRocksDBOptions:start",
-                                                 [&](void* /*arg*/) {
-    auto current_thread_id = std::this_thread::get_id();
-    // If it's the main thread hitting this sync-point, then it
-    // will be blocked until some other thread update the test_stage.
-    if (main_thread_id == current_thread_id) {
-      test_stage = kMainThreadStartPersistingOptionsFile;
-      while (test_stage < kChildThreadFinishDroppingColumnFamily &&
-             !ordered_by_writethread) {
-        Env::Default()->SleepForMicroseconds(100);
-      }
-    }
-  });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "PersistRocksDBOptions:start", [&](void* /*arg*/) {
+        auto current_thread_id = std::this_thread::get_id();
+        // If it's the main thread hitting this sync-point, then it
+        // will be blocked until some other thread update the test_stage.
+        if (main_thread_id == current_thread_id) {
+          test_stage = kMainThreadStartPersistingOptionsFile;
+          while (test_stage < kChildThreadFinishDroppingColumnFamily &&
+                 !ordered_by_writethread) {
+            Env::Default()->SleepForMicroseconds(100);
+          }
+        }
+      });

-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "WriteThread::EnterUnbatched:Wait", [&](void* /*arg*/) {
        // This means a thread doing DropColumnFamily() is waiting for
        // other thread to finish persisting options.
@@ -2528,12 +2532,12 @@ TEST_P(ColumnFamilyTest, CreateAndDropRace) {
  Open({"default", "one", "two", "three"},
       {cf_opts[0], cf_opts[1], cf_opts[2], cf_opts[3]});

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  // Start a thread that will drop the first column family
  // and its comparator
-  rocksdb::port::Thread drop_cf_thread(DropSingleColumnFamily, this, 1,
-                                       &comparators);
+  ROCKSDB_NAMESPACE::port::Thread drop_cf_thread(DropSingleColumnFamily, this,
+                                                 1, &comparators);

  DropColumnFamilies({2});

@@ -2546,8 +2550,8 @@ TEST_P(ColumnFamilyTest, CreateAndDropRace) {
    }
  }

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
 }
 #endif  // !ROCKSDB_LITE

@@ -2982,10 +2986,10 @@ TEST_P(ColumnFamilyTest, FlushCloseWALFiles) {
  ASSERT_OK(Put(0, "fodor", "mirko"));
  ASSERT_OK(Put(1, "fodor", "mirko"));

-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"DBImpl::BGWorkFlush:done", "FlushCloseWALFiles:0"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  // Block flush jobs from running
  test::SleepingBackgroundTask sleeping_task;
@@ -3001,7 +3005,7 @@ TEST_P(ColumnFamilyTest, FlushCloseWALFiles) {
  sleeping_task.WakeUp();
  sleeping_task.WaitUntilDone();
  TEST_SYNC_POINT("FlushCloseWALFiles:0");
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_EQ(1, env.num_open_wal_file_.load());

  Reopen();
@@ -3079,14 +3083,14 @@ TEST_P(ColumnFamilyTest, IteratorCloseWALFile2) {
  ASSERT_OK(Put(0, "fodor", "mirko"));
  ASSERT_OK(Put(1, "fodor", "mirko"));

-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"ColumnFamilyTest::IteratorCloseWALFile2:0",
       "DBImpl::BGWorkPurge:start"},
      {"ColumnFamilyTest::IteratorCloseWALFile2:2",
       "DBImpl::BackgroundCallFlush:start"},
      {"DBImpl::BGWorkPurge:end", "ColumnFamilyTest::IteratorCloseWALFile2:1"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  WriteOptions wo;
  wo.sync = true;
@@ -3104,7 +3108,7 @@ TEST_P(ColumnFamilyTest, IteratorCloseWALFile2) {
  TEST_SYNC_POINT("ColumnFamilyTest::IteratorCloseWALFile2:2");
  WaitForFlush(1);
  ASSERT_EQ(1, env.num_open_wal_file_.load());
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();

  Reopen();
  ASSERT_EQ("mirko", Get(0, "fodor"));
@@ -3148,14 +3152,14 @@ TEST_P(ColumnFamilyTest, ForwardIteratorCloseWALFile) {
  ASSERT_OK(Put(0, "fodor", "mirko"));
  ASSERT_OK(Put(1, "fodor", "mirko"));

-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"ColumnFamilyTest::IteratorCloseWALFile2:0",
       "DBImpl::BGWorkPurge:start"},
      {"ColumnFamilyTest::IteratorCloseWALFile2:2",
       "DBImpl::BackgroundCallFlush:start"},
      {"DBImpl::BGWorkPurge:end", "ColumnFamilyTest::IteratorCloseWALFile2:1"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  WriteOptions wo;
  wo.sync = true;
@@ -3179,7 +3183,7 @@ TEST_P(ColumnFamilyTest, ForwardIteratorCloseWALFile) {
  ASSERT_EQ(1, env.delete_count_.load());

  delete it;
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();

  Reopen();
  ASSERT_EQ("mirko", Get(0, "fodor"));
@@ -3199,15 +3203,15 @@ TEST_P(ColumnFamilyTest, LogSyncConflictFlush) {
  Put(0, "", "");
  Put(1, "foo", "bar");

-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
      {{"DBImpl::SyncWAL:BeforeMarkLogsSynced:1",
        "ColumnFamilyTest::LogSyncConflictFlush:1"},
       {"ColumnFamilyTest::LogSyncConflictFlush:2",
        "DBImpl::SyncWAL:BeforeMarkLogsSynced:2"}});

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

-  rocksdb::port::Thread thread([&] { db_->SyncWAL(); });
+  ROCKSDB_NAMESPACE::port::Thread thread([&] { db_->SyncWAL(); });

  TEST_SYNC_POINT("ColumnFamilyTest::LogSyncConflictFlush:1");
  Flush(1);
@@ -3218,7 +3222,7 @@ TEST_P(ColumnFamilyTest, LogSyncConflictFlush) {

  thread.join();

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  Close();
 }
 #endif
@@ -3367,7 +3371,7 @@ TEST_P(ColumnFamilyTest, MultipleCFPathsTest) {
  }
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #ifdef ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS
 extern "C" {
@@ -3378,7 +3382,7 @@ void RegisterCustomObjects(int /*argc*/, char** /*argv*/) {}
 #endif  // !ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  RegisterCustomObjects(argc, argv);
  return RUN_ALL_TESTS();
@@ -18,7 +18,7 @@
 #include "test_util/testharness.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CompactFilesTest : public testing::Test {
 public:
@@ -81,11 +81,11 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
  assert(s.ok());
  assert(db);

-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"CompactFilesImpl:0", "BackgroundCallCompaction:0"},
      {"BackgroundCallCompaction:1", "CompactFilesImpl:1"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  // create couple files
  // Background compaction starts and waits in BackgroundCallCompaction:0
@@ -95,7 +95,7 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
    db->Flush(FlushOptions());
  }

-  rocksdb::ColumnFamilyMetaData meta;
+  ROCKSDB_NAMESPACE::ColumnFamilyMetaData meta;
  db->GetColumnFamilyMetaData(&meta);
  std::string file1;
  for (auto& file : meta.levels[0].files) {
@@ -108,12 +108,12 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
      // The background compaction then notices that there is an L0 compaction
      // already in progress and doesn't do an L0 compaction
      // Once the background compaction finishes, the compact files finishes
-      ASSERT_OK(
-          db->CompactFiles(rocksdb::CompactionOptions(), {file1, file2}, 0));
+      ASSERT_OK(db->CompactFiles(ROCKSDB_NAMESPACE::CompactionOptions(),
+                                 {file1, file2}, 0));
      break;
    }
  }
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  delete db;
 }

@@ -224,14 +224,14 @@ TEST_F(CompactFilesTest, CapturingPendingFiles) {
  auto l0_files = collector->GetFlushedFiles();
  EXPECT_EQ(5, l0_files.size());

-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"CompactFilesImpl:2", "CompactFilesTest.CapturingPendingFiles:0"},
      {"CompactFilesTest.CapturingPendingFiles:1", "CompactFilesImpl:3"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  // Start compacting files.
-  rocksdb::port::Thread compaction_thread(
+  ROCKSDB_NAMESPACE::port::Thread compaction_thread(
      [&] { EXPECT_OK(db->CompactFiles(CompactionOptions(), l0_files, 1)); });

  // In the meantime flush another file.
@@ -242,7 +242,7 @@ TEST_F(CompactFilesTest, CapturingPendingFiles) {

  compaction_thread.join();

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();

  delete db;

@@ -296,12 +296,12 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
  db->Flush(FlushOptions());

  // Compact all L0 files using CompactFiles
-  rocksdb::ColumnFamilyMetaData meta;
+  ROCKSDB_NAMESPACE::ColumnFamilyMetaData meta;
  db->GetColumnFamilyMetaData(&meta);
  for (auto& file : meta.levels[0].files) {
    std::string fname = file.db_path + "/" + file.name;
    ASSERT_OK(
-        db->CompactFiles(rocksdb::CompactionOptions(), {fname}, 0));
+        db->CompactFiles(ROCKSDB_NAMESPACE::CompactionOptions(), {fname}, 0));
  }


@@ -347,7 +347,7 @@ TEST_F(CompactFilesTest, SentinelCompressionType) {
    compaction_opts.compression = CompressionType::kDisableCompressionOption;
    ASSERT_OK(db->CompactFiles(compaction_opts, l0_files, 1));

-    rocksdb::TablePropertiesCollection all_tables_props;
+    ROCKSDB_NAMESPACE::TablePropertiesCollection all_tables_props;
    ASSERT_OK(db->GetPropertiesOfAllTables(&all_tables_props));
    for (const auto& name_and_table_props : all_tables_props) {
      ASSERT_EQ(CompressionTypeToString(CompressionType::kZlibCompression),
@@ -402,7 +402,7 @@ TEST_F(CompactFilesTest, GetCompactionJobInfo) {
  delete db;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -9,7 +9,7 @@
 #include "db/version_set.h"
 #include "table/get_context.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 extern void MarkKeyMayExist(void* arg);
 extern bool SaveValue(void* arg, const ParsedInternalKey& parsed_key,
@@ -37,7 +37,7 @@ Status CompactedDBImpl::Get(const ReadOptions& options, ColumnFamilyHandle*,
                            const Slice& key, PinnableSlice* value) {
  GetContext get_context(user_comparator_, nullptr, nullptr, nullptr,
                         GetContext::kNotFound, key, value, nullptr, nullptr,
-                         true, nullptr, nullptr);
+                         nullptr, true, nullptr, nullptr);
  LookupKey lkey(key, kMaxSequenceNumber);
  files_.files[FindFile(key)].fd.table_reader->Get(options, lkey.internal_key(),
                                                   &get_context, nullptr);
@@ -70,7 +70,7 @@ std::vector<Status> CompactedDBImpl::MultiGet(const ReadOptions& options,
      std::string& value = (*values)[idx];
      GetContext get_context(user_comparator_, nullptr, nullptr, nullptr,
                             GetContext::kNotFound, keys[idx], &pinnable_val,
-                             nullptr, nullptr, true, nullptr, nullptr);
+                             nullptr, nullptr, nullptr, true, nullptr, nullptr);
      LookupKey lkey(keys[idx], kMaxSequenceNumber);
      r->Get(options, lkey.internal_key(), &get_context, nullptr);
      value.assign(pinnable_val.data(), pinnable_val.size());
@@ -156,5 +156,5 @@ Status CompactedDBImpl::Open(const Options& options,
  return s;
 }

-}   // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -9,7 +9,7 @@
 #include <vector>
 #include "db/db_impl/db_impl.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CompactedDBImpl : public DBImpl {
 public:
@@ -109,5 +109,5 @@ class CompactedDBImpl : public DBImpl {
  const Comparator* user_comparator_;
  LevelFilesBrief files_;
 };
-}
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // ROCKSDB_LITE
@@ -16,14 +16,14 @@
 #include "test_util/sync_point.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 const uint64_t kRangeTombstoneSentinel =
    PackSequenceAndType(kMaxSequenceNumber, kTypeRangeDeletion);

 int sstableKeyCompare(const Comparator* user_cmp, const InternalKey& a,
                      const InternalKey& b) {
-  auto c = user_cmp->Compare(a.user_key(), b.user_key());
+  auto c = user_cmp->CompareWithoutTimestamp(a.user_key(), b.user_key());
  if (c != 0) {
    return c;
  }
@@ -561,4 +561,4 @@ int Compaction::GetInputBaseLevel() const {
  return input_vstorage_->base_level();
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -13,7 +13,7 @@
 #include "options/cf_options.h"
 #include "util/autovector.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 // The file contains class Compaction, as well as some helper functions
 // and data structures used by the class.

@@ -381,4 +381,4 @@ class Compaction {
 // Return sum of sizes of all files in `files`.
 extern uint64_t TotalFileSize(const std::vector<FileMetaData*>& files);

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -5,6 +5,8 @@

 #pragma once

+#include "rocksdb/rocksdb_namespace.h"
+
 struct CompactionIterationStats {
  // Compaction statistics

@@ -28,7 +28,7 @@
  ((seq) <= earliest_snapshot_ && \
   (snapshot_checker_ == nullptr || LIKELY(IsInEarliestSnapshot(seq))))

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 CompactionIterator::CompactionIterator(
    InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
@@ -771,4 +771,4 @@ bool CompactionIterator::IsInEarliestSnapshot(SequenceNumber sequence) {
  return in_snapshot == SnapshotCheckerResult::kInSnapshot;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -19,7 +19,7 @@
 #include "options/cf_options.h"
 #include "rocksdb/compaction_filter.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CompactionIterator {
 public:
@@ -237,4 +237,4 @@ class CompactionIterator {
           manual_compaction_paused_->load(std::memory_order_relaxed);
  }
 };
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -14,7 +14,7 @@
 #include "util/string_util.h"
 #include "utilities/merge_operators.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // Expects no merging attempts.
 class NoMergingMergeOp : public MergeOperator {
@@ -494,7 +494,7 @@ TEST_P(CompactionIteratorTest, ShuttingDownInFilter) {
  compaction_proxy_->key_not_exists_beyond_output_level = true;

  std::atomic<bool> seek_done{false};
-  rocksdb::port::Thread compaction_thread([&] {
+  ROCKSDB_NAMESPACE::port::Thread compaction_thread([&] {
    c_iter_->SeekToFirst();
    EXPECT_FALSE(c_iter_->Valid());
    EXPECT_TRUE(c_iter_->status().IsShutdownInProgress());
@@ -531,7 +531,7 @@ TEST_P(CompactionIteratorTest, ShuttingDownInMerge) {
  compaction_proxy_->key_not_exists_beyond_output_level = true;

  std::atomic<bool> seek_done{false};
-  rocksdb::port::Thread compaction_thread([&] {
+  ROCKSDB_NAMESPACE::port::Thread compaction_thread([&] {
    c_iter_->SeekToFirst();
    ASSERT_FALSE(c_iter_->Valid());
    ASSERT_TRUE(c_iter_->status().IsShutdownInProgress());
@@ -968,7 +968,7 @@ TEST_F(CompactionIteratorWithSnapshotCheckerTest, CompactionFilter_FullMerge) {
      compaction_filter.get());
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -59,7 +59,7 @@
 #include "util/stop_watch.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 const char* GetCompactionReasonString(CompactionReason compaction_reason) {
  switch (compaction_reason) {
@@ -298,7 +298,7 @@ CompactionJob::CompactionJob(
    const FileOptions& file_options, VersionSet* versions,
    const std::atomic<bool>* shutting_down,
    const SequenceNumber preserve_deletes_seqnum, LogBuffer* log_buffer,
-    Directory* db_directory, Directory* output_directory, Statistics* stats,
+    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,
@@ -613,8 +613,13 @@ Status CompactionJob::Run() {
    }
  }

+  IOStatus io_s;
  if (status.ok() && output_directory_) {
-    status = output_directory_->Fsync();
+    io_s = output_directory_->Fsync(IOOptions(), nullptr);
+  }
+  if (!io_s.ok()) {
+    io_status_ = io_s;
+    status = io_s;
  }

  if (status.ok()) {
@@ -650,7 +655,8 @@ Status CompactionJob::Run() {
            TableReaderCaller::kCompactionRefill, /*arena=*/nullptr,
            /*skip_filters=*/false, compact_->compaction->output_level(),
            /*smallest_compaction_key=*/nullptr,
-            /*largest_compaction_key=*/nullptr);
+            /*largest_compaction_key=*/nullptr,
+            /*allow_unprepared_value=*/false);
        auto s = iter->status();

        if (s.ok() && paranoid_file_checks_) {
@@ -713,9 +719,13 @@ 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);
  }
+  if (!versions_->io_status().ok()) {
+    io_status_ = versions_->io_status();
+  }
  VersionStorageInfo::LevelSummaryStorage tmp;
  auto vstorage = cfd->current()->storage_info();
  const auto& stats = compaction_stats_;
@@ -928,7 +938,8 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
    assert(sub_compact->builder != nullptr);
    assert(sub_compact->current_output() != nullptr);
    sub_compact->builder->Add(key, value);
-    sub_compact->current_output_file_size = sub_compact->builder->FileSize();
+    sub_compact->current_output_file_size =
+        sub_compact->builder->EstimatedFileSize();
    const ParsedInternalKey& ikey = c_iter->ikey();
    sub_compact->current_output()->meta.UpdateBoundaries(
        key, value, ikey.sequence, ikey.type);
@@ -1294,6 +1305,10 @@ 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_;
+  }
  const uint64_t current_bytes = sub_compact->builder->FileSize();
  if (s.ok()) {
    meta->fd.file_size = current_bytes;
@@ -1302,12 +1317,23 @@ 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);
-    s = sub_compact->outfile->Sync(db_options_.use_fsync);
+    io_s = sub_compact->outfile->Sync(db_options_.use_fsync);
  }
-  if (s.ok()) {
-    s = sub_compact->outfile->Close();
+  if (io_s.ok()) {
+    io_s = sub_compact->outfile->Close();
+  }
+  if (io_s.ok()) {
+    // Add the checksum information to file metadata.
+    meta->file_checksum = sub_compact->outfile->GetFileChecksum();
+    meta->file_checksum_func_name =
+        sub_compact->outfile->GetFileChecksumFuncName();
+  }
+  if (!io_s.ok()) {
+    io_status_ = io_s;
+    s = io_s;
  }
  sub_compact->outfile.reset();

@@ -1508,7 +1534,8 @@ Status CompactionJob::OpenCompactionOutputFile(
      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));
+                             env_, 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
@@ -1691,4 +1718,4 @@ void CompactionJob::LogCompaction() {
  }
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -44,7 +44,7 @@
 #include "util/stop_watch.h"
 #include "util/thread_local.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class Arena;
 class ErrorHandler;
@@ -67,8 +67,8 @@ class CompactionJob {
                const FileOptions& file_options, VersionSet* versions,
                const std::atomic<bool>* shutting_down,
                const SequenceNumber preserve_deletes_seqnum,
-                LogBuffer* log_buffer, Directory* db_directory,
-                Directory* output_directory, Statistics* stats,
+                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,
@@ -100,6 +100,9 @@ class CompactionJob {
  // Add compaction input/output to the current version
  Status Install(const MutableCFOptions& mutable_cf_options);

+  // Return the IO status
+  IOStatus io_status() const { return io_status_; }
+
 private:
  struct SubcompactionState;

@@ -161,8 +164,8 @@ class CompactionJob {
  const std::atomic<bool>* manual_compaction_paused_;
  const SequenceNumber preserve_deletes_seqnum_;
  LogBuffer* log_buffer_;
-  Directory* db_directory_;
-  Directory* output_directory_;
+  FSDirectory* db_directory_;
+  FSDirectory* output_directory_;
  Statistics* stats_;
  InstrumentedMutex* db_mutex_;
  ErrorHandler* db_error_handler_;
@@ -193,6 +196,7 @@ class CompactionJob {
  std::vector<uint64_t> sizes_;
  Env::WriteLifeTimeHint write_hint_;
  Env::Priority thread_pri_;
+  IOStatus io_status_;
 };

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -61,7 +61,7 @@

 #if !defined(IOS_CROSS_COMPILE)
 #ifndef ROCKSDB_LITE
-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 static std::string RandomString(Random* rnd, int len, double ratio) {
  std::string r;
@@ -110,9 +110,9 @@ class CompactionJobStatsTest : public testing::Test,
  }

  ~CompactionJobStatsTest() override {
-    rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-    rocksdb::SyncPoint::GetInstance()->LoadDependency({});
-    rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({});
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
    Close();
    Options options;
    options.db_paths.emplace_back(dbname_, 0);
@@ -801,7 +801,7 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {

    stats_checker->set_verify_next_comp_io_stats(true);
    std::atomic<bool> first_prepare_write(true);
-    rocksdb::SyncPoint::GetInstance()->SetCallBack(
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
        "WritableFileWriter::Append:BeforePrepareWrite", [&](void* /*arg*/) {
          if (first_prepare_write.load()) {
            options.env->SleepForMicroseconds(3);
@@ -810,7 +810,7 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {
        });

    std::atomic<bool> first_flush(true);
-    rocksdb::SyncPoint::GetInstance()->SetCallBack(
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
        "WritableFileWriter::Flush:BeforeAppend", [&](void* /*arg*/) {
          if (first_flush.load()) {
            options.env->SleepForMicroseconds(3);
@@ -819,7 +819,7 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {
        });

    std::atomic<bool> first_sync(true);
-    rocksdb::SyncPoint::GetInstance()->SetCallBack(
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
        "WritableFileWriter::SyncInternal:0", [&](void* /*arg*/) {
          if (first_sync.load()) {
            options.env->SleepForMicroseconds(3);
@@ -828,14 +828,14 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {
        });

    std::atomic<bool> first_range_sync(true);
-    rocksdb::SyncPoint::GetInstance()->SetCallBack(
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
        "WritableFileWriter::RangeSync:0", [&](void* /*arg*/) {
          if (first_range_sync.load()) {
            options.env->SleepForMicroseconds(3);
            first_range_sync.store(false);
          }
        });
-    rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

    Compact(1, smallest_key, largest_key);

@@ -844,7 +844,7 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {
    ASSERT_TRUE(!first_flush.load());
    ASSERT_TRUE(!first_sync.load());
    ASSERT_TRUE(!first_range_sync.load());
-    rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+    ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  }
  ASSERT_EQ(stats_checker->NumberOfUnverifiedStats(), 0U);
 }
@@ -1019,10 +1019,10 @@ TEST_P(CompactionJobStatsTest, UniversalCompactionTest) {

 INSTANTIATE_TEST_CASE_P(CompactionJobStatsTest, CompactionJobStatsTest,
                        ::testing::Values(1, 4));
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -12,7 +12,7 @@
 #include <string>
 #include <tuple>

-#include "db/blob_index.h"
+#include "db/blob/blob_index.h"
 #include "db/column_family.h"
 #include "db/compaction/compaction_job.h"
 #include "db/db_impl/db_impl.h"
@@ -29,7 +29,7 @@
 #include "util/string_util.h"
 #include "utilities/merge_operators.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 namespace {

@@ -187,7 +187,8 @@ class CompactionJobTest : public testing::Test {
    VersionEdit edit;
    edit.AddFile(level, file_number, 0, 10, smallest_key, largest_key,
                 smallest_seqno, largest_seqno, false, oldest_blob_file_number,
-                 kUnknownOldestAncesterTime, kUnknownFileCreationTime);
+                 kUnknownOldestAncesterTime, kUnknownFileCreationTime,
+                 kUnknownFileChecksum, kUnknownFileChecksumFuncName);

    mutex_.Lock();
    versions_->LogAndApply(versions_->GetColumnFamilySet()->GetDefault(),
@@ -280,7 +281,7 @@ class CompactionJobTest : public testing::Test {
    }
    ASSERT_OK(s);
    // Make "CURRENT" file that points to the new manifest file.
-    s = SetCurrentFile(env_, dbname_, 1, nullptr);
+    s = SetCurrentFile(fs_.get(), dbname_, 1, nullptr);

    std::vector<ColumnFamilyDescriptor> column_families;
    cf_options_.table_factory = mock_table_factory_;
@@ -313,11 +314,11 @@ class CompactionJobTest : public testing::Test {
      num_input_files += level_files.size();
    }

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

    LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, db_options_.info_log.get());
@@ -1062,7 +1063,7 @@ TEST_F(CompactionJobTest, OldestBlobFileNumber) {
                /* expected_oldest_blob_file_number */ 19);
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -23,7 +23,7 @@
 #include "util/random.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 namespace {
 uint64_t TotalCompensatedFileSize(const std::vector<FileMetaData*>& files) {
@@ -110,9 +110,9 @@ CompressionType GetCompressionType(const ImmutableCFOptions& ioptions,

  // If bottommost_compression is set and we are compacting to the
  // bottommost level then we should use it.
-  if (ioptions.bottommost_compression != kDisableCompressionOption &&
+  if (mutable_cf_options.bottommost_compression != kDisableCompressionOption &&
      level >= (vstorage->num_non_empty_levels() - 1)) {
-    return ioptions.bottommost_compression;
+    return mutable_cf_options.bottommost_compression;
  }
  // If the user has specified a different compression level for each level,
  // then pick the compression for that level.
@@ -132,22 +132,22 @@ CompressionType GetCompressionType(const ImmutableCFOptions& ioptions,
  }
 }

-CompressionOptions GetCompressionOptions(const ImmutableCFOptions& ioptions,
+CompressionOptions GetCompressionOptions(const MutableCFOptions& cf_options,
                                         const VersionStorageInfo* vstorage,
                                         int level,
                                         const bool enable_compression) {
  if (!enable_compression) {
-    return ioptions.compression_opts;
+    return cf_options.compression_opts;
  }
  // If bottommost_compression is set and we are compacting to the
  // bottommost level then we should use the specified compression options
  // for the bottmomost_compression.
-  if (ioptions.bottommost_compression != kDisableCompressionOption &&
+  if (cf_options.bottommost_compression != kDisableCompressionOption &&
      level >= (vstorage->num_non_empty_levels() - 1) &&
-      ioptions.bottommost_compression_opts.enabled) {
-    return ioptions.bottommost_compression_opts;
+      cf_options.bottommost_compression_opts.enabled) {
+    return cf_options.bottommost_compression_opts;
  }
-  return ioptions.compression_opts;
+  return cf_options.compression_opts;
 }

 CompactionPicker::CompactionPicker(const ImmutableCFOptions& ioptions,
@@ -359,7 +359,7 @@ Compaction* CompactionPicker::CompactFiles(
      vstorage, ioptions_, mutable_cf_options, input_files, output_level,
      compact_options.output_file_size_limit,
      mutable_cf_options.max_compaction_bytes, output_path_id, compression_type,
-      GetCompressionOptions(ioptions_, vstorage, output_level),
+      GetCompressionOptions(mutable_cf_options, vstorage, output_level),
      compact_options.max_subcompactions,
      /* grandparents */ {}, true);
  RegisterCompaction(c);
@@ -634,7 +634,7 @@ Compaction* CompactionPicker::CompactRange(
        compact_range_options.target_path_id,
        GetCompressionType(ioptions_, vstorage, mutable_cf_options,
                           output_level, 1),
-        GetCompressionOptions(ioptions_, vstorage, output_level),
+        GetCompressionOptions(mutable_cf_options, vstorage, output_level),
        compact_range_options.max_subcompactions, /* grandparents */ {},
        /* is manual */ true);
    RegisterCompaction(c);
@@ -787,7 +787,7 @@ Compaction* CompactionPicker::CompactRange(
      compact_range_options.target_path_id,
      GetCompressionType(ioptions_, vstorage, mutable_cf_options, output_level,
                         vstorage->base_level()),
-      GetCompressionOptions(ioptions_, vstorage, output_level),
+      GetCompressionOptions(mutable_cf_options, vstorage, output_level),
      compact_range_options.max_subcompactions, std::move(grandparents),
      /* is manual compaction */ true);

@@ -1128,4 +1128,4 @@ bool CompactionPicker::GetOverlappingL0Files(
  return true;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -22,7 +22,7 @@
 #include "rocksdb/options.h"
 #include "rocksdb/status.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // The file contains an abstract class CompactionPicker, and its two
 // sub-classes LevelCompactionPicker and NullCompactionPicker, as
@@ -305,9 +305,9 @@ CompressionType GetCompressionType(const ImmutableCFOptions& ioptions,
                                   int level, int base_level,
                                   const bool enable_compression = true);

-CompressionOptions GetCompressionOptions(const ImmutableCFOptions& ioptions,
-                                         const VersionStorageInfo* vstorage,
-                                         int level,
-                                         const bool enable_compression = true);
+CompressionOptions GetCompressionOptions(
+    const MutableCFOptions& mutable_cf_options,
+    const VersionStorageInfo* vstorage, int level,
+    const bool enable_compression = true);

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -17,7 +17,7 @@
 #include "logging/log_buffer.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 namespace {
 uint64_t GetTotalFilesSize(const std::vector<FileMetaData*>& files) {
  uint64_t total_size = 0;
@@ -70,18 +70,18 @@ Compaction* FIFOCompactionPicker::PickTTLCompaction(
  // avoid underflow
  if (current_time > mutable_cf_options.ttl) {
    for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
-      auto f = *ritr;
-      if (f->fd.table_reader != nullptr &&
-          f->fd.table_reader->GetTableProperties() != nullptr) {
-        auto creation_time =
+      FileMetaData* f = *ritr;
+      assert(f);
+      if (f->fd.table_reader && f->fd.table_reader->GetTableProperties()) {
+        uint64_t creation_time =
            f->fd.table_reader->GetTableProperties()->creation_time;
        if (creation_time == 0 ||
            creation_time >= (current_time - mutable_cf_options.ttl)) {
          break;
        }
-        total_size -= f->compensated_file_size;
-        inputs[0].files.push_back(f);
      }
+      total_size -= f->compensated_file_size;
+      inputs[0].files.push_back(f);
    }
  }

@@ -96,17 +96,22 @@ Compaction* FIFOCompactionPicker::PickTTLCompaction(
  }

  for (const auto& f : inputs[0].files) {
+    uint64_t creation_time = 0;
+    assert(f);
+    if (f->fd.table_reader && f->fd.table_reader->GetTableProperties()) {
+      creation_time = f->fd.table_reader->GetTableProperties()->creation_time;
+    }
    ROCKS_LOG_BUFFER(log_buffer,
                     "[%s] FIFO compaction: picking file %" PRIu64
                     " with creation time %" PRIu64 " for deletion",
-                     cf_name.c_str(), f->fd.GetNumber(),
-                     f->fd.table_reader->GetTableProperties()->creation_time);
+                     cf_name.c_str(), f->fd.GetNumber(), creation_time);
  }

  Compaction* c = new Compaction(
      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
-      kNoCompression, ioptions_.compression_opts, /* max_subcompactions */ 0,
-      {}, /* is manual */ false, vstorage->CompactionScore(0),
+      kNoCompression, mutable_cf_options.compression_opts,
+      /* max_subcompactions */ 0, {}, /* is manual */ false,
+      vstorage->CompactionScore(0),
      /* is deletion compaction */ true, CompactionReason::kFIFOTtl);
  return c;
 }
@@ -146,7 +151,7 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(
            16 * 1024 * 1024 /* output file size limit */,
            0 /* max compaction bytes, not applicable */,
            0 /* output path ID */, mutable_cf_options.compression,
-            ioptions_.compression_opts, 0 /* max_subcompactions */, {},
+            mutable_cf_options.compression_opts, 0 /* max_subcompactions */, {},
            /* is manual */ false, vstorage->CompactionScore(0),
            /* is deletion compaction */ false,
            CompactionReason::kFIFOReduceNumFiles);
@@ -194,8 +199,9 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(

  Compaction* c = new Compaction(
      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
-      kNoCompression, ioptions_.compression_opts, /* max_subcompactions */ 0,
-      {}, /* is manual */ false, vstorage->CompactionScore(0),
+      kNoCompression, mutable_cf_options.compression_opts,
+      /* max_subcompactions */ 0, {}, /* is manual */ false,
+      vstorage->CompactionScore(0),
      /* is deletion compaction */ true, CompactionReason::kFIFOMaxSize);
  return c;
 }
@@ -238,5 +244,5 @@ Compaction* FIFOCompactionPicker::CompactRange(
  return c;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // !ROCKSDB_LITE
@@ -12,7 +12,7 @@

 #include "db/compaction/compaction_picker.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 class FIFOCompactionPicker : public CompactionPicker {
 public:
  FIFOCompactionPicker(const ImmutableCFOptions& ioptions,
@@ -49,5 +49,5 @@ class FIFOCompactionPicker : public CompactionPicker {
                                 VersionStorageInfo* version,
                                 LogBuffer* log_buffer);
 };
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // !ROCKSDB_LITE
@@ -15,7 +15,7 @@
 #include "logging/log_buffer.h"
 #include "test_util/sync_point.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 bool LevelCompactionPicker::NeedsCompaction(
    const VersionStorageInfo* vstorage) const {
@@ -93,9 +93,13 @@ class LevelCompactionBuilder {
  // otherwise, returns false.
  bool PickIntraL0Compaction();

-  void PickExpiredTtlFiles();
-
-  void PickFilesMarkedForPeriodicCompaction();
+  // Picks a file from level_files to compact.
+  // level_files is a vector of (level, file metadata) in ascending order of
+  // level. If compact_to_next_level is true, compact the file to the next
+  // level, otherwise, compact to the same level as the input file.
+  void PickFileToCompact(
+      const autovector<std::pair<int, FileMetaData*>>& level_files,
+      bool compact_to_next_level);

  const std::string& cf_name_;
  VersionStorageInfo* vstorage_;
@@ -124,72 +128,34 @@ class LevelCompactionBuilder {
  static const int kMinFilesForIntraL0Compaction = 4;
 };

-void LevelCompactionBuilder::PickExpiredTtlFiles() {
-  if (vstorage_->ExpiredTtlFiles().empty()) {
-    return;
-  }
-
-  auto continuation = [&](std::pair<int, FileMetaData*> level_file) {
+void LevelCompactionBuilder::PickFileToCompact(
+    const autovector<std::pair<int, FileMetaData*>>& level_files,
+    bool compact_to_next_level) {
+  for (auto& level_file : level_files) {
    // If it's being compacted it has nothing to do here.
    // If this assert() fails that means that some function marked some
    // files as being_compacted, but didn't call ComputeCompactionScore()
    assert(!level_file.second->being_compacted);
    start_level_ = level_file.first;
-    output_level_ =
-        (start_level_ == 0) ? vstorage_->base_level() : start_level_ + 1;
-
-    if ((start_level_ == vstorage_->num_non_empty_levels() - 1) ||
+    if ((compact_to_next_level &&
+         start_level_ == vstorage_->num_non_empty_levels() - 1) ||
        (start_level_ == 0 &&
         !compaction_picker_->level0_compactions_in_progress()->empty())) {
-      return false;
+      continue;
+    }
+    if (compact_to_next_level) {
+      output_level_ =
+          (start_level_ == 0) ? vstorage_->base_level() : start_level_ + 1;
+    } else {
+      output_level_ = start_level_;
    }
-
    start_level_inputs_.files = {level_file.second};
    start_level_inputs_.level = start_level_;
-    return compaction_picker_->ExpandInputsToCleanCut(cf_name_, vstorage_,
-                                                      &start_level_inputs_);
-  };
-
-  for (auto& level_file : vstorage_->ExpiredTtlFiles()) {
-    if (continuation(level_file)) {
-      // found the compaction!
+    if (compaction_picker_->ExpandInputsToCleanCut(cf_name_, vstorage_,
+                                                   &start_level_inputs_)) {
      return;
    }
  }
-
-  start_level_inputs_.files.clear();
-}
-
-void LevelCompactionBuilder::PickFilesMarkedForPeriodicCompaction() {
-  if (vstorage_->FilesMarkedForPeriodicCompaction().empty()) {
-    return;
-  }
-
-  auto continuation = [&](std::pair<int, FileMetaData*> level_file) {
-    // If it's being compacted it has nothing to do here.
-    // If this assert() fails that means that some function marked some
-    // files as being_compacted, but didn't call ComputeCompactionScore()
-    assert(!level_file.second->being_compacted);
-    output_level_ = start_level_ = level_file.first;
-
-    if (start_level_ == 0 &&
-        !compaction_picker_->level0_compactions_in_progress()->empty()) {
-      return false;
-    }
-
-    start_level_inputs_.files = {level_file.second};
-    start_level_inputs_.level = start_level_;
-    return compaction_picker_->ExpandInputsToCleanCut(cf_name_, vstorage_,
-                                                      &start_level_inputs_);
-  };
-
-  for (auto& level_file : vstorage_->FilesMarkedForPeriodicCompaction()) {
-    if (continuation(level_file)) {
-      // found the compaction!
-      return;
-    }
-  }
-
  start_level_inputs_.files.clear();
 }

@@ -238,64 +204,46 @@ void LevelCompactionBuilder::SetupInitialFiles() {
          }
        }
      }
+    } else {
+      // Compaction scores are sorted in descending order, no further scores
+      // will be >= 1.
+      break;
    }
  }
+  if (!start_level_inputs_.empty()) {
+    return;
+  }

  // if we didn't find a compaction, check if there are any files marked for
  // compaction
-  if (start_level_inputs_.empty()) {
-    parent_index_ = base_index_ = -1;
+  parent_index_ = base_index_ = -1;

-    compaction_picker_->PickFilesMarkedForCompaction(
-        cf_name_, vstorage_, &start_level_, &output_level_,
-        &start_level_inputs_);
-    if (!start_level_inputs_.empty()) {
-      is_manual_ = true;
-      compaction_reason_ = CompactionReason::kFilesMarkedForCompaction;
-      return;
-    }
+  compaction_picker_->PickFilesMarkedForCompaction(
+      cf_name_, vstorage_, &start_level_, &output_level_, &start_level_inputs_);
+  if (!start_level_inputs_.empty()) {
+    compaction_reason_ = CompactionReason::kFilesMarkedForCompaction;
+    return;
  }

  // Bottommost Files Compaction on deleting tombstones
-  if (start_level_inputs_.empty()) {
-    size_t i;
-    for (i = 0; i < vstorage_->BottommostFilesMarkedForCompaction().size();
-         ++i) {
-      auto& level_and_file = vstorage_->BottommostFilesMarkedForCompaction()[i];
-      assert(!level_and_file.second->being_compacted);
-      start_level_inputs_.level = output_level_ = start_level_ =
-          level_and_file.first;
-      start_level_inputs_.files = {level_and_file.second};
-      if (compaction_picker_->ExpandInputsToCleanCut(cf_name_, vstorage_,
-                                                     &start_level_inputs_)) {
-        break;
-      }
-    }
-    if (i == vstorage_->BottommostFilesMarkedForCompaction().size()) {
-      start_level_inputs_.clear();
-    } else {
-      assert(!start_level_inputs_.empty());
-      compaction_reason_ = CompactionReason::kBottommostFiles;
-      return;
-    }
+  PickFileToCompact(vstorage_->BottommostFilesMarkedForCompaction(), false);
+  if (!start_level_inputs_.empty()) {
+    compaction_reason_ = CompactionReason::kBottommostFiles;
+    return;
  }

  // TTL Compaction
-  if (start_level_inputs_.empty()) {
-    PickExpiredTtlFiles();
-    if (!start_level_inputs_.empty()) {
-      compaction_reason_ = CompactionReason::kTtl;
-      return;
-    }
+  PickFileToCompact(vstorage_->ExpiredTtlFiles(), true);
+  if (!start_level_inputs_.empty()) {
+    compaction_reason_ = CompactionReason::kTtl;
+    return;
  }

  // Periodic Compaction
-  if (start_level_inputs_.empty()) {
-    PickFilesMarkedForPeriodicCompaction();
-    if (!start_level_inputs_.empty()) {
-      compaction_reason_ = CompactionReason::kPeriodicCompaction;
-      return;
-    }
+  PickFileToCompact(vstorage_->FilesMarkedForPeriodicCompaction(), false);
+  if (!start_level_inputs_.empty()) {
+    compaction_reason_ = CompactionReason::kPeriodicCompaction;
+    return;
  }
 }

@@ -384,7 +332,7 @@ Compaction* LevelCompactionBuilder::GetCompaction() {
      GetPathId(ioptions_, mutable_cf_options_, output_level_),
      GetCompressionType(ioptions_, vstorage_, mutable_cf_options_,
                         output_level_, vstorage_->base_level()),
-      GetCompressionOptions(ioptions_, vstorage_, output_level_),
+      GetCompressionOptions(mutable_cf_options_, vstorage_, output_level_),
      /* max_subcompactions */ 0, std::move(grandparents_), is_manual_,
      start_level_score_, false /* deletion_compaction */, compaction_reason_);

@@ -555,4 +503,4 @@ Compaction* LevelCompactionPicker::PickCompaction(
                                 log_buffer, mutable_cf_options, ioptions_);
  return builder.PickCompaction();
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -11,7 +11,7 @@

 #include "db/compaction/compaction_picker.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 // Picking compactions for leveled compaction. See wiki page
 // https://github.com/facebook/rocksdb/wiki/Leveled-Compaction
 // for description of Leveled compaction.
@@ -29,4 +29,4 @@ class LevelCompactionPicker : public CompactionPicker {
      const VersionStorageInfo* vstorage) const override;
 };

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -17,7 +17,7 @@
 #include "test_util/testutil.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CountingLogger : public Logger {
 public:
@@ -95,7 +95,8 @@ class CompactionPickerTest : public testing::Test {
        InternalKey(smallest, smallest_seq, kTypeValue),
        InternalKey(largest, largest_seq, kTypeValue), smallest_seq,
        largest_seq, /* marked_for_compact */ false, kInvalidBlobFileNumber,
-        kUnknownOldestAncesterTime, kUnknownFileCreationTime);
+        kUnknownOldestAncesterTime, kUnknownFileCreationTime,
+        kUnknownFileChecksum, kUnknownFileChecksumFuncName);
    f->compensated_file_size =
        (compensated_file_size != 0) ? compensated_file_size : file_size;
    vstorage_->AddFile(level, f);
@@ -1732,7 +1733,7 @@ TEST_F(CompactionPickerTest, IntraL0ForEarliestSeqno) {
  ASSERT_EQ(0, compaction->output_level());
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -23,7 +23,7 @@
 #include "util/random.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 namespace {
 // A helper class that form universal compactions. The class is used by
 // UniversalCompactionPicker::PickCompaction().
@@ -389,7 +389,7 @@ Compaction* UniversalCompactionBuilder::PickCompaction() {
  VersionStorageInfo::LevelSummaryStorage tmp;
  ROCKS_LOG_BUFFER_MAX_SZ(
      log_buffer_, 3072,
-      "[%s] Universal: sorted runs files(%" ROCKSDB_PRIszt "): %s\n",
+      "[%s] Universal: sorted runs: %" ROCKSDB_PRIszt " files: %s\n",
      cf_name_.c_str(), sorted_runs_.size(), vstorage_->LevelSummary(&tmp));

  Compaction* c = nullptr;
@@ -751,7 +751,7 @@ Compaction* UniversalCompactionBuilder::PickCompactionToReduceSortedRuns(
      LLONG_MAX, path_id,
      GetCompressionType(ioptions_, vstorage_, mutable_cf_options_, start_level,
                         1, enable_compression),
-      GetCompressionOptions(ioptions_, vstorage_, start_level,
+      GetCompressionOptions(mutable_cf_options_, vstorage_, start_level,
                            enable_compression),
      /* max_subcompactions */ 0, /* grandparents */ {}, /* is manual */ false,
      score_, false /* deletion_compaction */, compaction_reason);
@@ -959,8 +959,8 @@ Compaction* UniversalCompactionBuilder::PickDeleteTriggeredCompaction() {
      /* max_grandparent_overlap_bytes */ LLONG_MAX, path_id,
      GetCompressionType(ioptions_, vstorage_, mutable_cf_options_,
                         output_level, 1),
-      GetCompressionOptions(ioptions_, vstorage_, output_level),
-      /* max_subcompactions */ 0, /* grandparents */ {}, /* is manual */ true,
+      GetCompressionOptions(mutable_cf_options_, vstorage_, output_level),
+      /* max_subcompactions */ 0, /* grandparents */ {}, /* is manual */ false,
      score_, false /* deletion_compaction */,
      CompactionReason::kFilesMarkedForCompaction);
 }
@@ -1029,7 +1029,7 @@ Compaction* UniversalCompactionBuilder::PickCompactionToOldest(
      LLONG_MAX, path_id,
      GetCompressionType(ioptions_, vstorage_, mutable_cf_options_, start_level,
                         1, true /* enable_compression */),
-      GetCompressionOptions(ioptions_, vstorage_, start_level,
+      GetCompressionOptions(mutable_cf_options_, vstorage_, start_level,
                            true /* enable_compression */),
      /* max_subcompactions */ 0, /* grandparents */ {}, /* is manual */ false,
      score_, false /* deletion_compaction */, compaction_reason);
@@ -1100,6 +1100,6 @@ Compaction* UniversalCompactionBuilder::PickPeriodicCompaction() {

  return c;
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // !ROCKSDB_LITE
@@ -12,7 +12,7 @@

 #include "db/compaction/compaction_picker.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 class UniversalCompactionPicker : public CompactionPicker {
 public:
  UniversalCompactionPicker(const ImmutableCFOptions& ioptions,
@@ -27,5 +27,5 @@ class UniversalCompactionPicker : public CompactionPicker {
  virtual bool NeedsCompaction(
      const VersionStorageInfo* vstorage) const override;
 };
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // !ROCKSDB_LITE
@@ -18,10 +18,10 @@

 using std::unique_ptr;

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 namespace {

-static const Comparator* comparator;
+static const Comparator* kTestComparator = nullptr;

 class KVIter : public Iterator {
 public:
@@ -74,7 +74,7 @@ void AssertItersEqual(Iterator* iter1, Iterator* iter2) {
 void DoRandomIteraratorTest(DB* db, std::vector<std::string> source_strings,
                            Random* rnd, int num_writes, int num_iter_ops,
                            int num_trigger_flush) {
-  stl_wrappers::KVMap map((stl_wrappers::LessOfComparator(comparator)));
+  stl_wrappers::KVMap map((stl_wrappers::LessOfComparator(kTestComparator)));

  for (int i = 0; i < num_writes; i++) {
    if (num_trigger_flush > 0 && i != 0 && i % num_trigger_flush == 0) {
@@ -263,19 +263,19 @@ class ComparatorDBTest

 public:
  ComparatorDBTest() : env_(Env::Default()), db_(nullptr) {
-    comparator = BytewiseComparator();
+    kTestComparator = BytewiseComparator();
    dbname_ = test::PerThreadDBPath("comparator_db_test");
    BlockBasedTableOptions toptions;
    toptions.format_version = GetParam();
    last_options_.table_factory.reset(
-        rocksdb::NewBlockBasedTableFactory(toptions));
+        ROCKSDB_NAMESPACE::NewBlockBasedTableFactory(toptions));
    EXPECT_OK(DestroyDB(dbname_, last_options_));
  }

  ~ComparatorDBTest() override {
    delete db_;
    EXPECT_OK(DestroyDB(dbname_, last_options_));
-    comparator = BytewiseComparator();
+    kTestComparator = BytewiseComparator();
  }

  DB* GetDB() { return db_; }
@@ -286,7 +286,7 @@ class ComparatorDBTest
    } else {
      comparator_guard.reset();
    }
-    comparator = cmp;
+    kTestComparator = cmp;
    last_options_.comparator = cmp;
  }

@@ -334,7 +334,7 @@ TEST_P(ComparatorDBTest, SimpleSuffixReverseComparator) {

  for (int rnd_seed = 301; rnd_seed < 316; rnd_seed++) {
    Options* opt = GetOptions();
-    opt->comparator = comparator;
+    opt->comparator = kTestComparator;
    DestroyAndReopen();
    Random rnd(rnd_seed);

@@ -360,7 +360,7 @@ TEST_P(ComparatorDBTest, Uint64Comparator) {

  for (int rnd_seed = 301; rnd_seed < 316; rnd_seed++) {
    Options* opt = GetOptions();
-    opt->comparator = comparator;
+    opt->comparator = kTestComparator;
    DestroyAndReopen();
    Random rnd(rnd_seed);
    Random64 rnd64(rnd_seed);
@@ -384,7 +384,7 @@ TEST_P(ComparatorDBTest, DoubleComparator) {

  for (int rnd_seed = 301; rnd_seed < 316; rnd_seed++) {
    Options* opt = GetOptions();
-    opt->comparator = comparator;
+    opt->comparator = kTestComparator;
    DestroyAndReopen();
    Random rnd(rnd_seed);

@@ -409,7 +409,7 @@ TEST_P(ComparatorDBTest, HashComparator) {

  for (int rnd_seed = 301; rnd_seed < 316; rnd_seed++) {
    Options* opt = GetOptions();
-    opt->comparator = comparator;
+    opt->comparator = kTestComparator;
    DestroyAndReopen();
    Random rnd(rnd_seed);

@@ -428,7 +428,7 @@ TEST_P(ComparatorDBTest, TwoStrComparator) {

  for (int rnd_seed = 301; rnd_seed < 316; rnd_seed++) {
    Options* opt = GetOptions();
-    opt->comparator = comparator;
+    opt->comparator = kTestComparator;
    DestroyAndReopen();
    Random rnd(rnd_seed);

@@ -652,7 +652,7 @@ TEST_P(ComparatorDBTest, SeparatorSuccessorRandomizeTest) {
  }
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -11,10 +11,10 @@
 #include "db/db_impl/db_impl.h"
 #include "util/cast_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 void CancelAllBackgroundWork(DB* db, bool wait) {
-  (static_cast_with_check<DBImpl, DB>(db->GetRootDB()))
+  (static_cast_with_check<DBImpl>(db->GetRootDB()))
      ->CancelAllBackgroundWork(wait);
 }

@@ -28,7 +28,7 @@ Status DeleteFilesInRange(DB* db, ColumnFamilyHandle* column_family,
 Status DeleteFilesInRanges(DB* db, ColumnFamilyHandle* column_family,
                           const RangePtr* ranges, size_t n,
                           bool include_end) {
-  return (static_cast_with_check<DBImpl, DB>(db->GetRootDB()))
+  return (static_cast_with_check<DBImpl>(db->GetRootDB()))
      ->DeleteFilesInRanges(column_family, ranges, n, include_end);
 }

@@ -72,6 +72,6 @@ Status VerifySstFileChecksum(const Options& options,
  return s;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // ROCKSDB_LITE
@@ -33,7 +33,7 @@
 #include "test_util/testutil.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 static const int kValueSize = 1000;

@@ -98,7 +98,7 @@ class CorruptionTest : public testing::Test {
  void RepairDB() {
    delete db_;
    db_ = nullptr;
-    ASSERT_OK(::rocksdb::RepairDB(dbname_, options_));
+    ASSERT_OK(::ROCKSDB_NAMESPACE::RepairDB(dbname_, options_));
  }

  void Build(int n, int flush_every = 0) {
@@ -190,7 +190,6 @@ class CorruptionTest : public testing::Test {
    ASSERT_TRUE(s.ok()) << s.ToString();
    Options options;
    EnvOptions env_options;
-    options.file_system.reset(new LegacyFileSystemWrapper(options.env));
    ASSERT_NOK(VerifySstFileChecksum(options, env_options, fname));
  }

@@ -393,12 +392,16 @@ TEST_F(CorruptionTest, TableFileIndexData) {

  // corrupt an index block of an entire file
  Corrupt(kTableFile, -2000, 500);
-  Reopen();
+  options.paranoid_checks = false;
+  Reopen(&options);
  dbi = reinterpret_cast<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);
  ASSERT_NOK(dbi->VerifyChecksum());
+
+  // In paranoid mode, the db cannot be opened due to the corrupted file.
+  ASSERT_TRUE(TryReopen().IsCorruption());
 }

 TEST_F(CorruptionTest, MissingDescriptor) {
@@ -554,13 +557,7 @@ TEST_F(CorruptionTest, RangeDeletionCorrupted) {

  ASSERT_OK(TryReopen());
  CorruptFile(filename, static_cast<int>(range_del_handle.offset()), 1);
-  // The test case does not fail on TryReopen because failure to preload table
-  // handlers is not considered critical.
-  ASSERT_OK(TryReopen());
-  std::string val;
-  // However, it does fail on any read involving that file since that file
-  // cannot be opened with a corrupt range deletion meta-block.
-  ASSERT_TRUE(db_->Get(ReadOptions(), "a", &val).IsCorruption());
+  ASSERT_TRUE(TryReopen().IsCorruption());
 }

 TEST_F(CorruptionTest, FileSystemStateCorrupted) {
@@ -585,18 +582,19 @@ TEST_F(CorruptionTest, FileSystemStateCorrupted) {
      env_.NewWritableFile(filename, &file, EnvOptions());
      file->Append(Slice("corrupted sst"));
      file.reset();
+      Status x = TryReopen(&options);
+      ASSERT_TRUE(x.IsCorruption());
    } else {  // delete the file
      env_.DeleteFile(filename);
+      Status x = TryReopen(&options);
+      ASSERT_TRUE(x.IsPathNotFound());
    }

-    Status x = TryReopen(&options);
-    ASSERT_TRUE(x.IsCorruption());
    DestroyDB(dbname_, options_);
-    Reopen(&options);
  }
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -15,7 +15,7 @@
 #include "test_util/testutil.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CuckooTableDBTest : public testing::Test {
 private:
@@ -298,17 +298,25 @@ TEST_F(CuckooTableDBTest, AdaptiveTable) {
  dbfull()->TEST_FlushMemTable();

  // Write some keys using plain table.
+  std::shared_ptr<TableFactory> block_based_factory(
+      NewBlockBasedTableFactory());
+  std::shared_ptr<TableFactory> plain_table_factory(
+      NewPlainTableFactory());
+  std::shared_ptr<TableFactory> cuckoo_table_factory(
+      NewCuckooTableFactory());
  options.create_if_missing = false;
-  options.table_factory.reset(NewPlainTableFactory());
+  options.table_factory.reset(NewAdaptiveTableFactory(
+    plain_table_factory, block_based_factory, plain_table_factory,
+    cuckoo_table_factory));
  Reopen(&options);
  ASSERT_OK(Put("key4", "v4"));
  ASSERT_OK(Put("key1", "v5"));
  dbfull()->TEST_FlushMemTable();

  // Write some keys using block based table.
-  std::shared_ptr<TableFactory> block_based_factory(
-      NewBlockBasedTableFactory());
-  options.table_factory.reset(NewAdaptiveTableFactory(block_based_factory));
+  options.table_factory.reset(NewAdaptiveTableFactory(
+    block_based_factory, block_based_factory, plain_table_factory,
+    cuckoo_table_factory));
  Reopen(&options);
  ASSERT_OK(Put("key5", "v6"));
  ASSERT_OK(Put("key2", "v7"));
@@ -320,14 +328,13 @@ TEST_F(CuckooTableDBTest, AdaptiveTable) {
  ASSERT_EQ("v4", Get("key4"));
  ASSERT_EQ("v6", Get("key5"));
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  if (rocksdb::port::kLittleEndian) {
+  if (ROCKSDB_NAMESPACE::port::kLittleEndian) {
    ::testing::InitGoogleTest(&argc, argv);
    return RUN_ALL_TESTS();
-  }
-  else {
+  } else {
    fprintf(stderr, "SKIPPED as Cuckoo table doesn't support Big Endian\n");
    return 0;
  }
@@ -12,7 +12,7 @@
 #include "port/stack_trace.h"
 #include "util/compression.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBBlockCacheTest : public DBTestBase {
 private:
@@ -46,7 +46,7 @@ class DBBlockCacheTest : public DBTestBase {
    options.create_if_missing = true;
    options.avoid_flush_during_recovery = false;
    // options.compression = kNoCompression;
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    options.table_factory.reset(new BlockBasedTableFactory(table_options));
    return options;
  }
@@ -292,7 +292,7 @@ TEST_F(DBBlockCacheTest, TestWithCompressedBlockCache) {
 TEST_F(DBBlockCacheTest, IndexAndFilterBlocksOfNewTableAddedToCache) {
  Options options = CurrentOptions();
  options.create_if_missing = true;
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
  BlockBasedTableOptions table_options;
  table_options.cache_index_and_filter_blocks = true;
  table_options.filter_policy.reset(NewBloomFilterPolicy(20));
@@ -378,7 +378,7 @@ TEST_F(DBBlockCacheTest, FillCacheAndIterateDB) {
 TEST_F(DBBlockCacheTest, IndexAndFilterBlocksStats) {
  Options options = CurrentOptions();
  options.create_if_missing = true;
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
  BlockBasedTableOptions table_options;
  table_options.cache_index_and_filter_blocks = true;
  LRUCacheOptions co;
@@ -464,7 +464,7 @@ TEST_F(DBBlockCacheTest, IndexAndFilterBlocksCachePriority) {
  for (auto priority : {Cache::Priority::LOW, Cache::Priority::HIGH}) {
    Options options = CurrentOptions();
    options.create_if_missing = true;
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    BlockBasedTableOptions table_options;
    table_options.cache_index_and_filter_blocks = true;
    table_options.block_cache.reset(new MockCache());
@@ -517,10 +517,143 @@ TEST_F(DBBlockCacheTest, IndexAndFilterBlocksCachePriority) {
  }
 }

+namespace {
+
+// An LRUCache wrapper that can falsely report "not found" on Lookup.
+// This allows us to manipulate BlockBasedTableReader into thinking
+// another thread inserted the data in between Lookup and Insert,
+// while mostly preserving the LRUCache interface/behavior.
+class LookupLiarCache : public CacheWrapper {
+  int nth_lookup_not_found_ = 0;
+
+ public:
+  explicit LookupLiarCache(std::shared_ptr<Cache> target)
+      : CacheWrapper(std::move(target)) {}
+
+  Handle* Lookup(const Slice& key, Statistics* stats) override {
+    if (nth_lookup_not_found_ == 1) {
+      nth_lookup_not_found_ = 0;
+      return nullptr;
+    }
+    if (nth_lookup_not_found_ > 1) {
+      --nth_lookup_not_found_;
+    }
+    return CacheWrapper::Lookup(key, stats);
+  }
+
+  // 1 == next lookup, 2 == after next, etc.
+  void SetNthLookupNotFound(int n) { nth_lookup_not_found_ = n; }
+};
+
+}  // anonymous namespace
+
+TEST_F(DBBlockCacheTest, AddRedundantStats) {
+  const size_t capacity = size_t{1} << 25;
+  const int num_shard_bits = 0;  // 1 shard
+  int iterations_tested = 0;
+  for (std::shared_ptr<Cache> base_cache :
+       {NewLRUCache(capacity, num_shard_bits),
+        NewClockCache(capacity, num_shard_bits)}) {
+    if (!base_cache) {
+      // Skip clock cache when not supported
+      continue;
+    }
+    ++iterations_tested;
+    Options options = CurrentOptions();
+    options.create_if_missing = true;
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
+
+    std::shared_ptr<LookupLiarCache> cache =
+        std::make_shared<LookupLiarCache>(base_cache);
+
+    BlockBasedTableOptions table_options;
+    table_options.cache_index_and_filter_blocks = true;
+    table_options.block_cache = cache;
+    table_options.filter_policy.reset(NewBloomFilterPolicy(50));
+    options.table_factory.reset(new BlockBasedTableFactory(table_options));
+    DestroyAndReopen(options);
+
+    // Create a new table.
+    ASSERT_OK(Put("foo", "value"));
+    ASSERT_OK(Put("bar", "value"));
+    ASSERT_OK(Flush());
+    ASSERT_EQ(1, NumTableFilesAtLevel(0));
+
+    // Normal access filter+index+data.
+    ASSERT_EQ("value", Get("foo"));
+
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD));
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD));
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD));
+    // --------
+    ASSERT_EQ(3, TestGetTickerCount(options, BLOCK_CACHE_ADD));
+
+    ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD_REDUNDANT));
+    ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD_REDUNDANT));
+    ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD_REDUNDANT));
+    // --------
+    ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_ADD_REDUNDANT));
+
+    // Againt access filter+index+data, but force redundant load+insert on index
+    cache->SetNthLookupNotFound(2);
+    ASSERT_EQ("value", Get("bar"));
+
+    ASSERT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD));
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD));
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD));
+    // --------
+    ASSERT_EQ(4, TestGetTickerCount(options, BLOCK_CACHE_ADD));
+
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD_REDUNDANT));
+    ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD_REDUNDANT));
+    ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD_REDUNDANT));
+    // --------
+    ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_ADD_REDUNDANT));
+
+    // Access just filter (with high probability), and force redundant
+    // load+insert
+    cache->SetNthLookupNotFound(1);
+    ASSERT_EQ("NOT_FOUND", Get("this key was not added"));
+
+    EXPECT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD));
+    EXPECT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD));
+    EXPECT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD));
+    // --------
+    EXPECT_EQ(5, TestGetTickerCount(options, BLOCK_CACHE_ADD));
+
+    EXPECT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD_REDUNDANT));
+    EXPECT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD_REDUNDANT));
+    EXPECT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD_REDUNDANT));
+    // --------
+    EXPECT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_ADD_REDUNDANT));
+
+    // Access just data, forcing redundant load+insert
+    ReadOptions read_options;
+    std::unique_ptr<Iterator> iter{db_->NewIterator(read_options)};
+    cache->SetNthLookupNotFound(1);
+    iter->SeekToFirst();
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_EQ(iter->key(), "bar");
+
+    EXPECT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD));
+    EXPECT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD));
+    EXPECT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD));
+    // --------
+    EXPECT_EQ(6, TestGetTickerCount(options, BLOCK_CACHE_ADD));
+
+    EXPECT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_ADD_REDUNDANT));
+    EXPECT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_ADD_REDUNDANT));
+    EXPECT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_DATA_ADD_REDUNDANT));
+    // --------
+    EXPECT_EQ(3, TestGetTickerCount(options, BLOCK_CACHE_ADD_REDUNDANT));
+  }
+  EXPECT_GE(iterations_tested, 1);
+}
+
 TEST_F(DBBlockCacheTest, ParanoidFileChecks) {
  Options options = CurrentOptions();
  options.create_if_missing = true;
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
  options.level0_file_num_compaction_trigger = 2;
  options.paranoid_file_checks = true;
  BlockBasedTableOptions table_options;
@@ -576,7 +709,7 @@ TEST_F(DBBlockCacheTest, CompressedCache) {
  for (int iter = 0; iter < 4; iter++) {
    Options options = CurrentOptions();
    options.write_buffer_size = 64 * 1024;  // small write buffer
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();

    BlockBasedTableOptions table_options;
    switch (iter) {
@@ -705,7 +838,7 @@ TEST_F(DBBlockCacheTest, CacheCompressionDict) {
    options.compression_opts.max_dict_bytes = 4096;
    options.create_if_missing = true;
    options.num_levels = 2;
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    options.target_file_size_base = kNumEntriesPerFile * kNumBytesPerEntry;
    BlockBasedTableOptions table_options;
    table_options.cache_index_and_filter_blocks = true;
@@ -752,10 +885,10 @@ TEST_F(DBBlockCacheTest, CacheCompressionDict) {

 #endif  // ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -8,20 +8,15 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.

 #include "db/db_test_util.h"
+#include "options/options_helper.h"
 #include "port/stack_trace.h"
 #include "rocksdb/perf_context.h"
 #include "table/block_based/filter_policy_internal.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 namespace {
 using BFP = BloomFilterPolicy;
-
-namespace BFP2 {
-// Extends BFP::Mode with option to use Plain table
-using PseudoMode = int;
-static constexpr PseudoMode kPlainTable = -1;
-}  // namespace BFP2
 }  // namespace

 // DB tests related to bloom filter.
@@ -94,7 +89,7 @@ TEST_P(DBBloomFilterTestDefFormatVersion, KeyMayExist) {
      // indexes
      continue;
    }
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    CreateAndReopenWithCF({"pikachu"}, options);

    ASSERT_TRUE(!db_->KeyMayExist(ropts, handles_[1], "a", &value));
@@ -156,7 +151,7 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloomCustomPrefixExtractor) {
    Options options = last_options_;
    options.prefix_extractor =
        std::make_shared<SliceTransformLimitedDomainGeneric>();
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    get_perf_context()->EnablePerLevelPerfContext();
    BlockBasedTableOptions bbto;
    bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
@@ -222,7 +217,7 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloom) {
  for (bool partition_filters : {true, false}) {
    Options options = last_options_;
    options.prefix_extractor.reset(NewFixedPrefixTransform(8));
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    get_perf_context()->EnablePerLevelPerfContext();
    BlockBasedTableOptions bbto;
    bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
@@ -273,7 +268,7 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
  for (bool partition_filters : {true, false}) {
    Options options = last_options_;
    options.prefix_extractor.reset(NewFixedPrefixTransform(3));
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    get_perf_context()->EnablePerLevelPerfContext();

    BlockBasedTableOptions bbto;
@@ -537,7 +532,7 @@ INSTANTIATE_TEST_CASE_P(
 TEST_F(DBBloomFilterTest, BloomFilterRate) {
  while (ChangeFilterOptions()) {
    Options options = CurrentOptions();
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    get_perf_context()->EnablePerLevelPerfContext();
    CreateAndReopenWithCF({"pikachu"}, options);

@@ -569,7 +564,7 @@ TEST_F(DBBloomFilterTest, BloomFilterRate) {

 TEST_F(DBBloomFilterTest, BloomFilterCompatibility) {
  Options options = CurrentOptions();
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
  BlockBasedTableOptions table_options;
  table_options.filter_policy.reset(NewBloomFilterPolicy(10, true));
  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
@@ -613,7 +608,7 @@ TEST_F(DBBloomFilterTest, BloomFilterCompatibility) {
 TEST_F(DBBloomFilterTest, BloomFilterReverseCompatibility) {
  for (bool partition_filters : {true, false}) {
    Options options = CurrentOptions();
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    BlockBasedTableOptions table_options;
    if (partition_filters) {
      table_options.partition_filters = true;
@@ -660,17 +655,18 @@ class TestingWrappedBlockBasedFilterPolicy : public FilterPolicy {
    return "TestingWrappedBlockBasedFilterPolicy";
  }

-  void CreateFilter(const rocksdb::Slice* keys, int n,
+  void CreateFilter(const ROCKSDB_NAMESPACE::Slice* keys, int n,
                    std::string* dst) const override {
-    std::unique_ptr<rocksdb::Slice[]> user_keys(new rocksdb::Slice[n]);
+    std::unique_ptr<ROCKSDB_NAMESPACE::Slice[]> user_keys(
+        new ROCKSDB_NAMESPACE::Slice[n]);
    for (int i = 0; i < n; ++i) {
      user_keys[i] = convertKey(keys[i]);
    }
    return filter_->CreateFilter(user_keys.get(), n, dst);
  }

-  bool KeyMayMatch(const rocksdb::Slice& key,
-                   const rocksdb::Slice& filter) const override {
+  bool KeyMayMatch(const ROCKSDB_NAMESPACE::Slice& key,
+                   const ROCKSDB_NAMESPACE::Slice& filter) const override {
    counter_++;
    return filter_->KeyMayMatch(convertKey(key), filter);
  }
@@ -681,13 +677,16 @@ class TestingWrappedBlockBasedFilterPolicy : public FilterPolicy {
  const FilterPolicy* filter_;
  mutable uint32_t counter_;

-  rocksdb::Slice convertKey(const rocksdb::Slice& key) const { return key; }
+  ROCKSDB_NAMESPACE::Slice convertKey(
+      const ROCKSDB_NAMESPACE::Slice& key) const {
+    return key;
+  }
 };
 }  // namespace

 TEST_F(DBBloomFilterTest, WrappedBlockBasedFilterPolicy) {
  Options options = CurrentOptions();
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();

  BlockBasedTableOptions table_options;
  TestingWrappedBlockBasedFilterPolicy* policy =
@@ -803,7 +802,7 @@ class TestingContextCustomFilterPolicy
 TEST_F(DBBloomFilterTest, ContextCustomFilterPolicy) {
  for (bool fifo : {true, false}) {
    Options options = CurrentOptions();
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    options.compaction_style =
        fifo ? kCompactionStyleFIFO : kCompactionStyleLevel;

@@ -902,7 +901,7 @@ class SliceTransformLimitedDomain : public SliceTransform {
 TEST_F(DBBloomFilterTest, PrefixExtractorFullFilter) {
  BlockBasedTableOptions bbto;
  // Full Filter Block
-  bbto.filter_policy.reset(rocksdb::NewBloomFilterPolicy(10, false));
+  bbto.filter_policy.reset(ROCKSDB_NAMESPACE::NewBloomFilterPolicy(10, false));
  bbto.whole_key_filtering = false;

  Options options = CurrentOptions();
@@ -931,7 +930,7 @@ TEST_F(DBBloomFilterTest, PrefixExtractorFullFilter) {
 TEST_F(DBBloomFilterTest, PrefixExtractorBlockFilter) {
  BlockBasedTableOptions bbto;
  // Block Filter Block
-  bbto.filter_policy.reset(rocksdb::NewBloomFilterPolicy(10, true));
+  bbto.filter_policy.reset(ROCKSDB_NAMESPACE::NewBloomFilterPolicy(10, true));

  Options options = CurrentOptions();
  options.prefix_extractor = std::make_shared<SliceTransformLimitedDomain>();
@@ -970,7 +969,8 @@ TEST_F(DBBloomFilterTest, MemtableWholeKeyBloomFilter) {
  Options options = CurrentOptions();
  options.memtable_prefix_bloom_size_ratio =
      static_cast<double>(kMemtablePrefixFilterSize) / kMemtableSize;
-  options.prefix_extractor.reset(rocksdb::NewFixedPrefixTransform(kPrefixLen));
+  options.prefix_extractor.reset(
+      ROCKSDB_NAMESPACE::NewFixedPrefixTransform(kPrefixLen));
  options.write_buffer_size = kMemtableSize;
  options.memtable_whole_key_filtering = false;
  Reopen(options);
@@ -1030,7 +1030,215 @@ TEST_F(DBBloomFilterTest, MemtablePrefixBloomOutOfDomain) {
  ASSERT_EQ(kKey, iter->key());
 }

+class DBBloomFilterTestVaryPrefixAndFormatVer
+    : public DBTestBase,
+      public testing::WithParamInterface<std::tuple<bool, uint32_t>> {
+ protected:
+  bool use_prefix_;
+  uint32_t format_version_;
+
+ public:
+  DBBloomFilterTestVaryPrefixAndFormatVer()
+      : DBTestBase("/db_bloom_filter_tests") {}
+
+  ~DBBloomFilterTestVaryPrefixAndFormatVer() override {}
+
+  void SetUp() override {
+    use_prefix_ = std::get<0>(GetParam());
+    format_version_ = std::get<1>(GetParam());
+  }
+
+  static std::string UKey(uint32_t i) { return Key(static_cast<int>(i)); }
+};
+
+TEST_P(DBBloomFilterTestVaryPrefixAndFormatVer, PartitionedMultiGet) {
+  Options options = CurrentOptions();
+  if (use_prefix_) {
+    // Entire key from UKey()
+    options.prefix_extractor.reset(NewCappedPrefixTransform(9));
+  }
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
+  BlockBasedTableOptions bbto;
+  bbto.filter_policy.reset(NewBloomFilterPolicy(20));
+  bbto.partition_filters = true;
+  bbto.index_type = BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch;
+  bbto.whole_key_filtering = !use_prefix_;
+  bbto.metadata_block_size = 128;
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  DestroyAndReopen(options);
+  ReadOptions ropts;
+
+  constexpr uint32_t N = 10000;
+  // Add N/2 evens
+  for (uint32_t i = 0; i < N; i += 2) {
+    ASSERT_OK(Put(UKey(i), UKey(i)));
+  }
+  ASSERT_OK(Flush());
 #ifndef ROCKSDB_LITE
+  ASSERT_EQ(TotalTableFiles(), 1);
+#endif
+
+  constexpr uint32_t Q = 29;
+  // MultiGet In
+  std::array<std::string, Q> keys;
+  std::array<Slice, Q> key_slices;
+  std::array<ColumnFamilyHandle*, Q> column_families;
+  // MultiGet Out
+  std::array<Status, Q> statuses;
+  std::array<PinnableSlice, Q> values;
+
+  TestGetAndResetTickerCount(options, BLOCK_CACHE_FILTER_HIT);
+  TestGetAndResetTickerCount(options, BLOCK_CACHE_FILTER_MISS);
+  TestGetAndResetTickerCount(options, BLOOM_FILTER_PREFIX_USEFUL);
+  TestGetAndResetTickerCount(options, BLOOM_FILTER_USEFUL);
+  TestGetAndResetTickerCount(options, BLOOM_FILTER_PREFIX_CHECKED);
+  TestGetAndResetTickerCount(options, BLOOM_FILTER_FULL_POSITIVE);
+  TestGetAndResetTickerCount(options, BLOOM_FILTER_FULL_TRUE_POSITIVE);
+
+  // Check that initial clump of keys only loads one partition filter from
+  // block cache.
+  // And that spread out keys load many partition filters.
+  // In both cases, mix present vs. not present keys.
+  for (uint32_t stride : {uint32_t{1}, (N / Q) | 1}) {
+    for (uint32_t i = 0; i < Q; ++i) {
+      keys[i] = UKey(i * stride);
+      key_slices[i] = Slice(keys[i]);
+      column_families[i] = db_->DefaultColumnFamily();
+      statuses[i] = Status();
+      values[i] = PinnableSlice();
+    }
+
+    db_->MultiGet(ropts, Q, &column_families[0], &key_slices[0], &values[0],
+                  /*timestamps=*/nullptr, &statuses[0], true);
+
+    // Confirm correct status results
+    uint32_t number_not_found = 0;
+    for (uint32_t i = 0; i < Q; ++i) {
+      if ((i * stride % 2) == 0) {
+        ASSERT_OK(statuses[i]);
+      } else {
+        ASSERT_TRUE(statuses[i].IsNotFound());
+        ++number_not_found;
+      }
+    }
+
+    // Confirm correct Bloom stats (no FPs)
+    uint64_t filter_useful = TestGetAndResetTickerCount(
+        options,
+        use_prefix_ ? BLOOM_FILTER_PREFIX_USEFUL : BLOOM_FILTER_USEFUL);
+    uint64_t filter_checked =
+        TestGetAndResetTickerCount(options, use_prefix_
+                                                ? BLOOM_FILTER_PREFIX_CHECKED
+                                                : BLOOM_FILTER_FULL_POSITIVE) +
+        (use_prefix_ ? 0 : filter_useful);
+    EXPECT_EQ(filter_useful, number_not_found);
+    EXPECT_EQ(filter_checked, Q);
+    if (!use_prefix_) {
+      EXPECT_EQ(
+          TestGetAndResetTickerCount(options, BLOOM_FILTER_FULL_TRUE_POSITIVE),
+          Q - number_not_found);
+    }
+
+    // Confirm no duplicate loading same filter partition
+    uint64_t filter_accesses =
+        TestGetAndResetTickerCount(options, BLOCK_CACHE_FILTER_HIT) +
+        TestGetAndResetTickerCount(options, BLOCK_CACHE_FILTER_MISS);
+    if (stride == 1) {
+      EXPECT_EQ(filter_accesses, 1);
+    } else {
+      // for large stride
+      EXPECT_GE(filter_accesses, Q / 2 + 1);
+    }
+  }
+
+  // Check that a clump of keys (present and not) works when spanning
+  // two partitions
+  int found_spanning = 0;
+  for (uint32_t start = 0; start < N / 2;) {
+    for (uint32_t i = 0; i < Q; ++i) {
+      keys[i] = UKey(start + i);
+      key_slices[i] = Slice(keys[i]);
+      column_families[i] = db_->DefaultColumnFamily();
+      statuses[i] = Status();
+      values[i] = PinnableSlice();
+    }
+
+    db_->MultiGet(ropts, Q, &column_families[0], &key_slices[0], &values[0],
+                  /*timestamps=*/nullptr, &statuses[0], true);
+
+    // Confirm correct status results
+    uint32_t number_not_found = 0;
+    for (uint32_t i = 0; i < Q; ++i) {
+      if (((start + i) % 2) == 0) {
+        ASSERT_OK(statuses[i]);
+      } else {
+        ASSERT_TRUE(statuses[i].IsNotFound());
+        ++number_not_found;
+      }
+    }
+
+    // Confirm correct Bloom stats (might see some FPs)
+    uint64_t filter_useful = TestGetAndResetTickerCount(
+        options,
+        use_prefix_ ? BLOOM_FILTER_PREFIX_USEFUL : BLOOM_FILTER_USEFUL);
+    uint64_t filter_checked =
+        TestGetAndResetTickerCount(options, use_prefix_
+                                                ? BLOOM_FILTER_PREFIX_CHECKED
+                                                : BLOOM_FILTER_FULL_POSITIVE) +
+        (use_prefix_ ? 0 : filter_useful);
+    EXPECT_GE(filter_useful, number_not_found - 2);  // possible FP
+    EXPECT_EQ(filter_checked, Q);
+    if (!use_prefix_) {
+      EXPECT_EQ(
+          TestGetAndResetTickerCount(options, BLOOM_FILTER_FULL_TRUE_POSITIVE),
+          Q - number_not_found);
+    }
+
+    // Confirm no duplicate loading of same filter partition
+    uint64_t filter_accesses =
+        TestGetAndResetTickerCount(options, BLOCK_CACHE_FILTER_HIT) +
+        TestGetAndResetTickerCount(options, BLOCK_CACHE_FILTER_MISS);
+    if (filter_accesses == 2) {
+      // Spanned across partitions.
+      ++found_spanning;
+      if (found_spanning >= 2) {
+        break;
+      } else {
+        // Ensure that at least once we have at least one present and
+        // one non-present key on both sides of partition boundary.
+        start += 2;
+      }
+    } else {
+      EXPECT_EQ(filter_accesses, 1);
+      // See explanation at "start += 2"
+      start += Q - 4;
+    }
+  }
+  EXPECT_TRUE(found_spanning >= 2);
+}
+
+INSTANTIATE_TEST_CASE_P(DBBloomFilterTestVaryPrefixAndFormatVer,
+                        DBBloomFilterTestVaryPrefixAndFormatVer,
+                        ::testing::Values(
+                            // (use_prefix, format_version)
+                            std::make_tuple(false, 2),
+                            std::make_tuple(false, 3),
+                            std::make_tuple(false, 4),
+                            std::make_tuple(false, 5),
+                            std::make_tuple(true, 2),
+                            std::make_tuple(true, 3),
+                            std::make_tuple(true, 4),
+                            std::make_tuple(true, 5)));
+
+#ifndef ROCKSDB_LITE
+namespace {
+namespace BFP2 {
+// Extends BFP::Mode with option to use Plain table
+using PseudoMode = int;
+static constexpr PseudoMode kPlainTable = -1;
+}  // namespace BFP2
+}  // namespace
+
 class BloomStatsTestWithParam
    : public DBBloomFilterTest,
      public testing::WithParamInterface<std::tuple<BFP2::PseudoMode, bool>> {
@@ -1040,7 +1248,8 @@ class BloomStatsTestWithParam
    partition_filters_ = std::get<1>(GetParam());

    options_.create_if_missing = true;
-    options_.prefix_extractor.reset(rocksdb::NewFixedPrefixTransform(4));
+    options_.prefix_extractor.reset(
+        ROCKSDB_NAMESPACE::NewFixedPrefixTransform(4));
    options_.memtable_prefix_bloom_size_ratio =
        8.0 * 1024.0 / static_cast<double>(options_.write_buffer_size);
    if (bfp_impl_ == BFP2::kPlainTable) {
@@ -1321,7 +1530,7 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
  bbto.whole_key_filtering = true;
  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
  options.optimize_filters_for_hits = true;
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
  get_perf_context()->Reset();
  get_perf_context()->EnablePerLevelPerfContext();
  CreateAndReopenWithCF({"mypikachu"}, options);
@@ -1335,8 +1544,7 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
  for (int i = 0; i < numkeys; i += 2) {
    keys.push_back(i);
  }
-  std::random_shuffle(std::begin(keys), std::end(keys));
-
+  RandomShuffle(std::begin(keys), std::end(keys));
  int num_inserted = 0;
  for (int key : keys) {
    ASSERT_OK(Put(1, Key(key), "val"));
@@ -1447,13 +1655,13 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {

  int32_t trivial_move = 0;
  int32_t non_trivial_move = 0;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:TrivialMove",
      [&](void* /*arg*/) { trivial_move++; });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial",
      [&](void* /*arg*/) { non_trivial_move++; });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  CompactRangeOptions compact_options;
  compact_options.bottommost_level_compaction =
@@ -1893,10 +2101,10 @@ TEST_F(DBBloomFilterTest, DynamicBloomFilterOptions) {

 #endif  // ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -10,7 +10,7 @@
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 static int cfilter_count = 0;
 static int cfilter_skips = 0;
@@ -47,8 +47,7 @@ class DBTestCompactionFilterWithCompactParam

 #ifndef ROCKSDB_VALGRIND_RUN
 INSTANTIATE_TEST_CASE_P(
-    DBTestCompactionFilterWithCompactOption,
-    DBTestCompactionFilterWithCompactParam,
+    CompactionFilterWithOption, DBTestCompactionFilterWithCompactParam,
    ::testing::Values(DBTestBase::OptionConfig::kDefault,
                      DBTestBase::OptionConfig::kUniversalCompaction,
                      DBTestBase::OptionConfig::kUniversalCompactionMultiLevel,
@@ -56,7 +55,7 @@ INSTANTIATE_TEST_CASE_P(
                      DBTestBase::OptionConfig::kUniversalSubcompactions));
 #else
 // Run fewer cases in valgrind
-INSTANTIATE_TEST_CASE_P(DBTestCompactionFilterWithCompactOption,
+INSTANTIATE_TEST_CASE_P(CompactionFilterWithOption,
                        DBTestCompactionFilterWithCompactParam,
                        ::testing::Values(DBTestBase::OptionConfig::kDefault));
 #endif  // ROCKSDB_VALGRIND_RUN
@@ -864,10 +863,10 @@ TEST_F(DBTestCompactionFilter, IgnoreSnapshotsFalse) {
  delete options.compaction_filter;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -16,7 +16,7 @@
 #include "port/port.h"
 #include "port/stack_trace.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 class DBTestDynamicLevel : public DBTestBase {
 public:
  DBTestDynamicLevel() : DBTestBase("/db_dynamic_level_test") {}
@@ -50,7 +50,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase) {
      keys[i] = i;
    }
    if (ordered_insert == 0) {
-      std::random_shuffle(std::begin(keys), std::end(keys));
+      RandomShuffle(std::begin(keys), std::end(keys), rnd.Next());
    }
    for (int max_background_compactions = 1; max_background_compactions < 4;
         max_background_compactions += 2) {
@@ -221,10 +221,10 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {

  // Make sure that the compaction starts before the last bit of data is
  // flushed, so that the base level isn't raised to L1.
-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"CompactionJob::Run():Start", "DynamicLevelMaxBytesBase2:0"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "false"},
@@ -235,20 +235,20 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  dbfull()->TEST_WaitForCompact();
  ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
  ASSERT_EQ(2U, int_prop);
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();

  // Write more data until the base level changes to L1. There will be
  // a manual compaction going on at the same time.
-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
      {"CompactionJob::Run():Start", "DynamicLevelMaxBytesBase2:1"},
      {"DynamicLevelMaxBytesBase2:2", "CompactionJob::Run():End"},
      {"DynamicLevelMaxBytesBase2:compact_range_finish",
       "FlushJob::WriteLevel0Table"},
  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

-  rocksdb::port::Thread thread([this] {
+  ROCKSDB_NAMESPACE::port::Thread thread([this] {
    TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:compact_range_start");
    ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
    TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:compact_range_finish");
@@ -265,8 +265,8 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {

  thread.join();

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();

  ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
  ASSERT_EQ(1U, int_prop);
@@ -329,16 +329,16 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesCompactRange) {
  ASSERT_TRUE(db_->GetProperty("rocksdb.num-files-at-level2", &str_prop));
  ASSERT_EQ("0", str_prop);

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();

  std::set<int> output_levels;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "CompactionPicker::CompactRange:Return", [&](void* arg) {
        Compaction* compaction = reinterpret_cast<Compaction*>(arg);
        output_levels.insert(compaction->output_level());
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr);
  ASSERT_EQ(output_levels.size(), 2);
@@ -373,10 +373,10 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBaseInc) {
  DestroyAndReopen(options);

  int non_trivial = 0;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::BackgroundCompaction:NonTrivial",
      [&](void* /*arg*/) { non_trivial++; });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  Random rnd(301);
  const int total_keys = 3000;
@@ -388,7 +388,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBaseInc) {
  }
  Flush();
  dbfull()->TEST_WaitForCompact();
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();

  ASSERT_EQ(non_trivial, 0);

@@ -457,7 +457,7 @@ TEST_F(DBTestDynamicLevel, DISABLED_MigrateToDynamicLevelMaxBytesBase) {
  compaction_finished = false;
  // Issue manual compaction in one thread and still verify DB state
  // in main thread.
-  rocksdb::port::Thread t([&]() {
+  ROCKSDB_NAMESPACE::port::Thread t([&]() {
    CompactRangeOptions compact_options;
    compact_options.change_level = true;
    compact_options.target_level = options.num_levels - 1;
@@ -488,13 +488,13 @@ TEST_F(DBTestDynamicLevel, DISABLED_MigrateToDynamicLevelMaxBytesBase) {
  ASSERT_EQ(NumTableFilesAtLevel(1), 0);
  ASSERT_EQ(NumTableFilesAtLevel(2), 0);
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // !defined(ROCKSDB_LITE)

 int main(int argc, char** argv) {
 #if !defined(ROCKSDB_LITE)
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 #else
@@ -12,7 +12,7 @@
 #include <iostream>
 #include <string>

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBEncryptionTest : public DBTestBase {
 public:
@@ -113,10 +113,10 @@ TEST_F(DBEncryptionTest, ReadEmptyFile) {

 #endif // ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -21,7 +21,7 @@
 #include "test_util/sync_point.h"
 #include "util/mutexlock.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 Status DBImpl::DisableFileDeletions() {
  InstrumentedMutexLock l(&mutex_);
@@ -90,6 +90,9 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
      mutex_.Unlock();
      status = AtomicFlushMemTables(cfds, FlushOptions(),
                                    FlushReason::kGetLiveFiles);
+      if (status.IsColumnFamilyDropped()) {
+        status = Status::OK();
+      }
      mutex_.Lock();
    } else {
      for (auto cfd : *versions_->GetColumnFamilySet()) {
@@ -103,8 +106,10 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
        TEST_SYNC_POINT("DBImpl::GetLiveFiles:2");
        mutex_.Lock();
        cfd->UnrefAndTryDelete();
-        if (!status.ok()) {
+        if (!status.ok() && !status.IsColumnFamilyDropped()) {
          break;
+        } else if (status.IsColumnFamilyDropped()) {
+          status = Status::OK();
        }
      }
    }
@@ -118,27 +123,33 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
    }
  }

-  // Make a set of all of the live *.sst files
-  std::vector<FileDescriptor> live;
+  // Make a set of all of the live table and blob files
+  std::vector<uint64_t> live_table_files;
+  std::vector<uint64_t> live_blob_files;
  for (auto cfd : *versions_->GetColumnFamilySet()) {
    if (cfd->IsDropped()) {
      continue;
    }
-    cfd->current()->AddLiveFiles(&live);
+    cfd->current()->AddLiveFiles(&live_table_files, &live_blob_files);
  }

  ret.clear();
-  ret.reserve(live.size() + 3);  // *.sst + CURRENT + MANIFEST + OPTIONS
+  ret.reserve(live_table_files.size() + live_blob_files.size() +
+              3);  // for CURRENT + MANIFEST + OPTIONS

  // create names of the live files. The names are not absolute
  // paths, instead they are relative to dbname_;
-  for (const auto& live_file : live) {
-    ret.push_back(MakeTableFileName("", live_file.GetNumber()));
+  for (const auto& table_file_number : live_table_files) {
+    ret.emplace_back(MakeTableFileName("", table_file_number));
  }

-  ret.push_back(CurrentFileName(""));
-  ret.push_back(DescriptorFileName("", versions_->manifest_file_number()));
-  ret.push_back(OptionsFileName("", versions_->options_file_number()));
+  for (const auto& blob_file_number : live_blob_files) {
+    ret.emplace_back(BlobFileName("", blob_file_number));
+  }
+
+  ret.emplace_back(CurrentFileName(""));
+  ret.emplace_back(DescriptorFileName("", versions_->manifest_file_number()));
+  ret.emplace_back(OptionsFileName("", versions_->options_file_number()));

  // find length of manifest file while holding the mutex lock
  *manifest_file_size = versions_->manifest_file_size();
@@ -172,6 +183,6 @@ Status DBImpl::GetCurrentWalFile(std::unique_ptr<LogFile>* current_log_file) {

  return wal_manager_.GetLiveWalFile(current_logfile_number, current_log_file);
 }
-}
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // ROCKSDB_LITE
@@ -18,7 +18,7 @@
 #include "util/cast_util.h"
 #include "util/mutexlock.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBFlushTest : public DBTestBase {
 public:
@@ -741,15 +741,44 @@ TEST_P(DBAtomicFlushTest, CFDropRaceWithWaitForFlushMemTables) {
  SyncPoint::GetInstance()->DisableProcessing();
 }

+TEST_P(DBAtomicFlushTest, RollbackAfterFailToInstallResults) {
+  bool atomic_flush = GetParam();
+  if (!atomic_flush) {
+    return;
+  }
+  auto fault_injection_env = std::make_shared<FaultInjectionTestEnv>(env_);
+  Options options = CurrentOptions();
+  options.env = fault_injection_env.get();
+  options.create_if_missing = true;
+  options.atomic_flush = atomic_flush;
+  CreateAndReopenWithCF({"pikachu"}, options);
+  ASSERT_EQ(2, handles_.size());
+  for (size_t cf = 0; cf < handles_.size(); ++cf) {
+    ASSERT_OK(Put(static_cast<int>(cf), "a", "value"));
+  }
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionSet::ProcessManifestWrites:BeforeWriteLastVersionEdit:0",
+      [&](void* /*arg*/) { fault_injection_env->SetFilesystemActive(false); });
+  SyncPoint::GetInstance()->EnableProcessing();
+  FlushOptions flush_opts;
+  Status s = db_->Flush(flush_opts, handles_);
+  ASSERT_NOK(s);
+  fault_injection_env->SetFilesystemActive(true);
+  Close();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
 INSTANTIATE_TEST_CASE_P(DBFlushDirectIOTest, DBFlushDirectIOTest,
                        testing::Bool());

 INSTANTIATE_TEST_CASE_P(DBAtomicFlushTest, DBAtomicFlushTest, testing::Bool());

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
@@ -21,7 +21,6 @@
 #include <stdexcept>
 #include <string>
 #include <unordered_map>
-#include <unordered_set>
 #include <utility>
 #include <vector>

@@ -102,7 +101,7 @@
 #include "util/stop_watch.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 const std::string kDefaultColumnFamilyName("default");
 const std::string kPersistentStatsColumnFamilyName(
@@ -151,7 +150,7 @@ 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_.file_system),
+      fs_(initial_db_options_.env->GetFileSystem()),
      immutable_db_options_(initial_db_options_),
      mutable_db_options_(initial_db_options_),
      stats_(immutable_db_options_.statistics.get()),
@@ -312,6 +311,11 @@ Status DBImpl::ResumeImpl() {
    s = bg_error;
  }

+  // Make sure the IO Status stored in version set is set to OK.
+  if (s.ok()) {
+    versions_->SetIOStatusOK();
+  }
+
  // We cannot guarantee consistency of the WAL. So force flush Memtables of
  // all the column families
  if (s.ok()) {
@@ -644,7 +648,7 @@ void DBImpl::StartTimedTasks() {
    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::RepeatableThread(
+        thread_dump_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
            [this]() { DBImpl::DumpStats(); }, "dump_st", env_,
            static_cast<uint64_t>(stats_dump_period_sec) * kMicrosInSecond));
      }
@@ -652,7 +656,7 @@ void DBImpl::StartTimedTasks() {
    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::RepeatableThread(
+        thread_persist_stats_.reset(new ROCKSDB_NAMESPACE::RepeatableThread(
            [this]() { DBImpl::PersistStats(); }, "pst_st", env_,
            static_cast<uint64_t>(stats_persist_period_sec) * kMicrosInSecond));
      }
@@ -866,9 +870,7 @@ void DBImpl::DumpStats() {
 Status DBImpl::TablesRangeTombstoneSummary(ColumnFamilyHandle* column_family,
                                           int max_entries_to_print,
                                           std::string* out_str) {
-  auto* cfh =
-      static_cast_with_check<ColumnFamilyHandleImpl, ColumnFamilyHandle>(
-          column_family);
+  auto* cfh = static_cast_with_check<ColumnFamilyHandleImpl>(column_family);
  ColumnFamilyData* cfd = cfh->cfd();

  SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
@@ -890,9 +892,9 @@ void DBImpl::ScheduleBgLogWriterClose(JobContext* job_context) {
  }
 }

-Directory* DBImpl::GetDataDir(ColumnFamilyData* cfd, size_t path_id) const {
+FSDirectory* DBImpl::GetDataDir(ColumnFamilyData* cfd, size_t path_id) const {
  assert(cfd);
-  Directory* ret_dir = cfd->GetDataDir(path_id);
+  FSDirectory* ret_dir = cfd->GetDataDir(path_id);
  if (ret_dir == nullptr) {
    return directories_.GetDataDir(path_id);
  }
@@ -1006,12 +1008,12 @@ Status DBImpl::SetDBOptions(
    }
    if (s.ok()) {
      const BGJobLimits current_bg_job_limits =
-          GetBGJobLimits(immutable_db_options_.max_background_flushes,
+          GetBGJobLimits(mutable_db_options_.max_background_flushes,
                         mutable_db_options_.max_background_compactions,
                         mutable_db_options_.max_background_jobs,
                         /* parallelize_compactions */ true);
      const BGJobLimits new_bg_job_limits = GetBGJobLimits(
-          immutable_db_options_.max_background_flushes,
+          new_options.max_background_flushes,
          new_options.max_background_compactions,
          new_options.max_background_jobs, /* parallelize_compactions */ true);

@@ -1043,7 +1045,7 @@ Status DBImpl::SetDBOptions(
          mutex_.Lock();
        }
        if (new_options.stats_dump_period_sec > 0) {
-          thread_dump_stats_.reset(new rocksdb::RepeatableThread(
+          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));
@@ -1059,7 +1061,7 @@ Status DBImpl::SetDBOptions(
          mutex_.Lock();
        }
        if (new_options.stats_persist_period_sec > 0) {
-          thread_persist_stats_.reset(new rocksdb::RepeatableThread(
+          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));
@@ -1147,25 +1149,25 @@ int DBImpl::FindMinimumEmptyLevelFitting(

 Status DBImpl::FlushWAL(bool sync) {
  if (manual_wal_flush_) {
-    Status s;
+    IOStatus io_s;
    {
      // We need to lock log_write_mutex_ since logs_ might change concurrently
      InstrumentedMutexLock wl(&log_write_mutex_);
      log::Writer* cur_log_writer = logs_.back().writer;
-      s = cur_log_writer->WriteBuffer();
+      io_s = cur_log_writer->WriteBuffer();
    }
-    if (!s.ok()) {
+    if (!io_s.ok()) {
      ROCKS_LOG_ERROR(immutable_db_options_.info_log, "WAL flush error %s",
-                      s.ToString().c_str());
+                      io_s.ToString().c_str());
      // In case there is a fs error we should set it globally to prevent the
      // future writes
-      WriteStatusCheck(s);
+      IOStatusCheck(io_s);
      // whether sync or not, we should abort the rest of function upon error
-      return s;
+      return std::move(io_s);
    }
    if (!sync) {
      ROCKS_LOG_DEBUG(immutable_db_options_.info_log, "FlushWAL sync=false");
-      return s;
+      return std::move(io_s);
    }
  }
  if (!sync) {
@@ -1217,14 +1219,23 @@ Status DBImpl::SyncWAL() {
  TEST_SYNC_POINT("DBWALTest::SyncWALNotWaitWrite:1");
  RecordTick(stats_, WAL_FILE_SYNCED);
  Status status;
+  IOStatus io_s;
  for (log::Writer* log : logs_to_sync) {
-    status = log->file()->SyncWithoutFlush(immutable_db_options_.use_fsync);
-    if (!status.ok()) {
+    io_s = log->file()->SyncWithoutFlush(immutable_db_options_.use_fsync);
+    if (!io_s.ok()) {
+      status = io_s;
      break;
    }
  }
+  if (!io_s.ok()) {
+    ROCKS_LOG_ERROR(immutable_db_options_.info_log, "WAL Sync error %s",
+                    io_s.ToString().c_str());
+    // In case there is a fs error we should set it globally to prevent the
+    // future writes
+    IOStatusCheck(io_s);
+  }
  if (status.ok() && need_log_dir_sync) {
-    status = directories_.GetWalDir()->Fsync();
+    status = directories_.GetWalDir()->Fsync(IOOptions(), nullptr);
  }
  TEST_SYNC_POINT("DBWALTest::SyncWALNotWaitWrite:2");

@@ -1249,7 +1260,7 @@ Status DBImpl::LockWAL() {
    // future writes
    WriteStatusCheck(status);
  }
-  return status;
+  return std::move(status);
 }

 Status DBImpl::UnlockWAL() {
@@ -1300,7 +1311,7 @@ bool DBImpl::SetPreserveDeletesSequenceNumber(SequenceNumber seqnum) {

 InternalIterator* DBImpl::NewInternalIterator(
    Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
-    ColumnFamilyHandle* column_family) {
+    ColumnFamilyHandle* column_family, bool allow_unprepared_value) {
  ColumnFamilyData* cfd;
  if (column_family == nullptr) {
    cfd = default_cf_handle_->cfd();
@@ -1314,7 +1325,7 @@ InternalIterator* DBImpl::NewInternalIterator(
  mutex_.Unlock();
  ReadOptions roptions;
  return NewInternalIterator(roptions, cfd, super_version, arena, range_del_agg,
-                             sequence);
+                             sequence, allow_unprepared_value);
 }

 void DBImpl::SchedulePurge() {
@@ -1437,7 +1448,8 @@ InternalIterator* DBImpl::NewInternalIterator(const ReadOptions& read_options,
                                              SuperVersion* super_version,
                                              Arena* arena,
                                              RangeDelAggregator* range_del_agg,
-                                              SequenceNumber sequence) {
+                                              SequenceNumber sequence,
+                                              bool allow_unprepared_value) {
  InternalIterator* internal_iter;
  assert(arena != nullptr);
  assert(range_del_agg != nullptr);
@@ -1469,7 +1481,8 @@ InternalIterator* DBImpl::NewInternalIterator(const ReadOptions& read_options,
    // Collect iterators for files in L0 - Ln
    if (read_options.read_tier != kMemtableTier) {
      super_version->current->AddIterators(read_options, file_options_,
-                                           &merge_iter_builder, range_del_agg);
+                                           &merge_iter_builder, range_del_agg,
+                                           allow_unprepared_value);
    }
    internal_iter = merge_iter_builder.Finish();
    IterState* cleanup =
@@ -1496,16 +1509,38 @@ ColumnFamilyHandle* DBImpl::PersistentStatsColumnFamily() const {
 Status DBImpl::Get(const ReadOptions& read_options,
                   ColumnFamilyHandle* column_family, const Slice& key,
                   PinnableSlice* value) {
+  return Get(read_options, column_family, key, value, /*timestamp=*/nullptr);
+}
+
+Status DBImpl::Get(const ReadOptions& read_options,
+                   ColumnFamilyHandle* column_family, const Slice& key,
+                   PinnableSlice* value, std::string* timestamp) {
  GetImplOptions get_impl_options;
  get_impl_options.column_family = column_family;
  get_impl_options.value = value;
-  return GetImpl(read_options, key, get_impl_options);
+  get_impl_options.timestamp = timestamp;
+  Status s = GetImpl(read_options, key, get_impl_options);
+  return s;
 }

 Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
-                       GetImplOptions get_impl_options) {
+                       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();
+  assert(ucmp);
+  if (ucmp->timestamp_size() > 0) {
+    assert(read_options.timestamp);
+    assert(read_options.timestamp->size() == ucmp->timestamp_size());
+  } else {
+    assert(!read_options.timestamp);
+  }
+#endif  // NDEBUG
+
  PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
  StopWatch sw(env_, stats_, DB_GET);
  PERF_TIMER_GUARD(get_snapshot_time);
@@ -1583,10 +1618,14 @@ 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
    if (get_impl_options.get_value) {
-      if (sv->mem->Get(lkey, get_impl_options.value->GetSelf(), &s,
+      if (sv->mem->Get(lkey, get_impl_options.value->GetSelf(), timestamp, &s,
                       &merge_context, &max_covering_tombstone_seq,
                       read_options, get_impl_options.callback,
                       get_impl_options.is_blob_index)) {
@@ -1594,9 +1633,10 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
        get_impl_options.value->PinSelf();
        RecordTick(stats_, MEMTABLE_HIT);
      } else if ((s.ok() || s.IsMergeInProgress()) &&
-                 sv->imm->Get(lkey, get_impl_options.value->GetSelf(), &s,
-                              &merge_context, &max_covering_tombstone_seq,
-                              read_options, get_impl_options.callback,
+                 sv->imm->Get(lkey, get_impl_options.value->GetSelf(),
+                              timestamp, &s, &merge_context,
+                              &max_covering_tombstone_seq, read_options,
+                              get_impl_options.callback,
                              get_impl_options.is_blob_index)) {
        done = true;
        get_impl_options.value->PinSelf();
@@ -1605,9 +1645,9 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
    } else {
      // Get Merge Operands associated with key, Merge Operands should not be
      // merged and raw values should be returned to the user.
-      if (sv->mem->Get(lkey, nullptr, &s, &merge_context,
-                       &max_covering_tombstone_seq, read_options, nullptr,
-                       nullptr, false)) {
+      if (sv->mem->Get(lkey, /*value*/ nullptr, /*timestamp=*/nullptr, &s,
+                       &merge_context, &max_covering_tombstone_seq,
+                       read_options, nullptr, nullptr, false)) {
        done = true;
        RecordTick(stats_, MEMTABLE_HIT);
      } else if ((s.ok() || s.IsMergeInProgress()) &&
@@ -1626,8 +1666,8 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
  if (!done) {
    PERF_TIMER_GUARD(get_from_output_files_time);
    sv->current->Get(
-        read_options, lkey, get_impl_options.value, &s, &merge_context,
-        &max_covering_tombstone_seq,
+        read_options, lkey, get_impl_options.value, timestamp, &s,
+        &merge_context, &max_covering_tombstone_seq,
        get_impl_options.get_value ? get_impl_options.value_found : nullptr,
        nullptr, nullptr,
        get_impl_options.get_value ? get_impl_options.callback : nullptr,
@@ -1675,12 +1715,34 @@ std::vector<Status> DBImpl::MultiGet(
    const ReadOptions& read_options,
    const std::vector<ColumnFamilyHandle*>& column_family,
    const std::vector<Slice>& keys, std::vector<std::string>* values) {
+  return MultiGet(read_options, column_family, keys, values,
+                  /*timestamps=*/nullptr);
+}
+
+std::vector<Status> DBImpl::MultiGet(
+    const ReadOptions& read_options,
+    const std::vector<ColumnFamilyHandle*>& column_family,
+    const std::vector<Slice>& keys, std::vector<std::string>* values,
+    std::vector<std::string>* timestamps) {
  PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
  StopWatch sw(env_, stats_, DB_MULTIGET);
  PERF_TIMER_GUARD(get_snapshot_time);

+#ifndef NDEBUG
+  for (const auto* cfh : column_family) {
+    assert(cfh);
+    const Comparator* const ucmp = cfh->GetComparator();
+    assert(ucmp);
+    if (ucmp->timestamp_size() > 0) {
+      assert(read_options.timestamp);
+      assert(ucmp->timestamp_size() == read_options.timestamp->size());
+    } else {
+      assert(!read_options.timestamp);
+    }
+  }
+#endif  // NDEBUG
+
  SequenceNumber consistent_seqnum;
-  ;

  std::unordered_map<uint32_t, MultiGetColumnFamilyData> multiget_cf_data(
      column_family.size());
@@ -1711,6 +1773,9 @@ std::vector<Status> DBImpl::MultiGet(
  size_t num_keys = keys.size();
  std::vector<Status> stat_list(num_keys);
  values->resize(num_keys);
+  if (timestamps) {
+    timestamps->resize(num_keys);
+  }

  // Keep track of bytes that we read for statistics-recording later
  uint64_t bytes_read = 0;
@@ -1721,13 +1786,16 @@ std::vector<Status> DBImpl::MultiGet(
  // s is both in/out. When in, s could either be OK or MergeInProgress.
  // merge_operands will contain the sequence of merges in the latter case.
  size_t num_found = 0;
-  for (size_t i = 0; i < num_keys; ++i) {
+  size_t keys_read;
+  for (keys_read = 0; keys_read < num_keys; ++keys_read) {
    merge_context.Clear();
-    Status& s = stat_list[i];
-    std::string* value = &(*values)[i];
+    Status& s = stat_list[keys_read];
+    std::string* value = &(*values)[keys_read];
+    std::string* timestamp = timestamps ? &(*timestamps)[keys_read] : nullptr;

-    LookupKey lkey(keys[i], consistent_seqnum);
-    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family[i]);
+    LookupKey lkey(keys[keys_read], consistent_seqnum, read_options.timestamp);
+    auto cfh =
+        static_cast_with_check<ColumnFamilyHandleImpl>(column_family[keys_read]);
    SequenceNumber max_covering_tombstone_seq = 0;
    auto mgd_iter = multiget_cf_data.find(cfh->cfd()->GetID());
    assert(mgd_iter != multiget_cf_data.end());
@@ -1738,13 +1806,13 @@ std::vector<Status> DBImpl::MultiGet(
         has_unpersisted_data_.load(std::memory_order_relaxed));
    bool done = false;
    if (!skip_memtable) {
-      if (super_version->mem->Get(lkey, value, &s, &merge_context,
+      if (super_version->mem->Get(lkey, value, timestamp, &s, &merge_context,
                                  &max_covering_tombstone_seq, read_options)) {
        done = true;
        RecordTick(stats_, MEMTABLE_HIT);
-      } else if (super_version->imm->Get(lkey, value, &s, &merge_context,
-                                         &max_covering_tombstone_seq,
-                                         read_options)) {
+      } else if (super_version->imm->Get(
+                     lkey, value, timestamp, &s, &merge_context,
+                     &max_covering_tombstone_seq, read_options)) {
        done = true;
        RecordTick(stats_, MEMTABLE_HIT);
      }
@@ -1752,8 +1820,9 @@ std::vector<Status> DBImpl::MultiGet(
    if (!done) {
      PinnableSlice pinnable_val;
      PERF_TIMER_GUARD(get_from_output_files_time);
-      super_version->current->Get(read_options, lkey, &pinnable_val, &s,
-                                  &merge_context, &max_covering_tombstone_seq);
+      super_version->current->Get(read_options, lkey, &pinnable_val, timestamp,
+                                  &s, &merge_context,
+                                  &max_covering_tombstone_seq);
      value->assign(pinnable_val.data(), pinnable_val.size());
      RecordTick(stats_, MEMTABLE_MISS);
    }
@@ -1762,6 +1831,22 @@ std::vector<Status> DBImpl::MultiGet(
      bytes_read += value->size();
      num_found++;
    }
+
+    if (read_options.deadline.count() &&
+        env_->NowMicros() >
+            static_cast<uint64_t>(read_options.deadline.count())) {
+      break;
+    }
+  }
+
+  if (keys_read < num_keys) {
+    // The only reason to break out of the loop is when the deadline is
+    // exceeded
+    assert(env_->NowMicros() >
+        static_cast<uint64_t>(read_options.deadline.count()));
+    for (++keys_read; keys_read < num_keys; ++keys_read) {
+      stat_list[keys_read] = Status::TimedOut();
+    }
  }

  // Post processing (decrement reference counts and record statistics)
@@ -1915,14 +2000,39 @@ void DBImpl::MultiGet(const ReadOptions& read_options, const size_t num_keys,
                      ColumnFamilyHandle** column_families, const Slice* keys,
                      PinnableSlice* values, Status* statuses,
                      const bool sorted_input) {
+  return MultiGet(read_options, num_keys, column_families, keys, values,
+                  /*timestamps=*/nullptr, statuses, sorted_input);
+}
+
+void DBImpl::MultiGet(const ReadOptions& read_options, const size_t num_keys,
+                      ColumnFamilyHandle** column_families, const Slice* keys,
+                      PinnableSlice* values, std::string* timestamps,
+                      Status* statuses, const bool sorted_input) {
  if (num_keys == 0) {
    return;
  }
+
+#ifndef NDEBUG
+  for (size_t i = 0; i < num_keys; ++i) {
+    ColumnFamilyHandle* cfh = column_families[i];
+    assert(cfh);
+    const Comparator* const ucmp = cfh->GetComparator();
+    assert(ucmp);
+    if (ucmp->timestamp_size() > 0) {
+      assert(read_options.timestamp);
+      assert(read_options.timestamp->size() == ucmp->timestamp_size());
+    } else {
+      assert(!read_options.timestamp);
+    }
+  }
+#endif  // NDEBUG
+
  autovector<KeyContext, MultiGetContext::MAX_BATCH_SIZE> key_context;
  autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE> sorted_keys;
  sorted_keys.resize(num_keys);
  for (size_t i = 0; i < num_keys; ++i) {
    key_context.emplace_back(column_families[i], keys[i], &values[i],
+                             timestamps ? &timestamps[i] : nullptr,
                             &statuses[i]);
  }
  for (size_t i = 0; i < num_keys; ++i) {
@@ -1963,14 +2073,31 @@ void DBImpl::MultiGet(const ReadOptions& read_options, const size_t num_keys,
      read_options, nullptr, iter_deref_lambda, &multiget_cf_data,
      &consistent_seqnum);

-  for (auto cf_iter = multiget_cf_data.begin();
-       cf_iter != multiget_cf_data.end(); ++cf_iter) {
-    MultiGetImpl(read_options, cf_iter->start, cf_iter->num_keys, &sorted_keys,
-                 cf_iter->super_version, consistent_seqnum, nullptr, nullptr);
+  Status s;
+  auto cf_iter = multiget_cf_data.begin();
+  for (; cf_iter != multiget_cf_data.end(); ++cf_iter) {
+    s = MultiGetImpl(read_options, cf_iter->start, cf_iter->num_keys,
+                     &sorted_keys, cf_iter->super_version, consistent_seqnum,
+                     nullptr, nullptr);
+    if (!s.ok()) {
+      break;
+    }
+  }
+  if (!s.ok()) {
+    assert(s.IsTimedOut());
+    for (++cf_iter; cf_iter != multiget_cf_data.end(); ++cf_iter) {
+      for (size_t i = cf_iter->start; i < cf_iter->start + cf_iter->num_keys;
+           ++i) {
+        *sorted_keys[i]->s = Status::TimedOut();
+      }
+    }
+  }
+
+  for (const auto& iter : multiget_cf_data) {
    if (!unref_only) {
-      ReturnAndCleanupSuperVersion(cf_iter->cfd, cf_iter->super_version);
+      ReturnAndCleanupSuperVersion(iter.cfd, iter.super_version);
    } else {
-      cf_iter->cfd->GetSuperVersion()->Unref();
+      iter.cfd->GetSuperVersion()->Unref();
    }
  }
 }
@@ -1993,7 +2120,8 @@ struct CompareKeyContext {
    }

    // Both keys are from the same column family
-    int cmp = comparator->Compare(*(lhs->key), *(rhs->key));
+    int cmp = comparator->CompareWithoutTimestamp(
+        *(lhs->key), /*a_has_ts=*/false, *(rhs->key), /*b_has_ts=*/false);
    if (cmp < 0) {
      return true;
    }
@@ -2025,7 +2153,8 @@ void DBImpl::PrepareMultiGetKeys(
        }

        // Both keys are from the same column family
-        int cmp = comparator->Compare(*(lhs->key), *(rhs->key));
+        int cmp = comparator->CompareWithoutTimestamp(
+            *(lhs->key), /*a_has_ts=*/false, *(rhs->key), /*b_has_ts=*/false);
        assert(cmp <= 0);
      }
      index++;
@@ -2043,11 +2172,22 @@ void DBImpl::MultiGet(const ReadOptions& read_options,
                      ColumnFamilyHandle* column_family, const size_t num_keys,
                      const Slice* keys, PinnableSlice* values,
                      Status* statuses, const bool sorted_input) {
+  return MultiGet(read_options, column_family, num_keys, keys, values,
+                  /*timestamp=*/nullptr, statuses, sorted_input);
+}
+
+void DBImpl::MultiGet(const ReadOptions& read_options,
+                      ColumnFamilyHandle* column_family, const size_t num_keys,
+                      const Slice* keys, PinnableSlice* values,
+                      std::string* timestamps, Status* statuses,
+                      const bool sorted_input) {
  autovector<KeyContext, MultiGetContext::MAX_BATCH_SIZE> key_context;
  autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE> sorted_keys;
  sorted_keys.resize(num_keys);
  for (size_t i = 0; i < num_keys; ++i) {
-    key_context.emplace_back(column_family, keys[i], &values[i], &statuses[i]);
+    key_context.emplace_back(column_family, keys[i], &values[i],
+                             timestamps ? &timestamps[i] : nullptr,
+                             &statuses[i]);
  }
  for (size_t i = 0; i < num_keys; ++i) {
    sorted_keys[i] = &key_context[i];
@@ -2100,14 +2240,24 @@ void DBImpl::MultiGetWithCallback(
    consistent_seqnum = callback->max_visible_seq();
  }

-  MultiGetImpl(read_options, 0, num_keys, sorted_keys,
-               multiget_cf_data[0].super_version, consistent_seqnum, nullptr,
-               nullptr);
+  Status s = MultiGetImpl(read_options, 0, num_keys, sorted_keys,
+                          multiget_cf_data[0].super_version, consistent_seqnum,
+                          nullptr, nullptr);
+  assert(s.ok() || s.IsTimedOut());
  ReturnAndCleanupSuperVersion(multiget_cf_data[0].cfd,
                               multiget_cf_data[0].super_version);
 }

-void DBImpl::MultiGetImpl(
+// The actual implementation of batched MultiGet. Parameters -
+// start_key - Index in the sorted_keys vector to start processing from
+// num_keys - Number of keys to lookup, starting with sorted_keys[start_key]
+// sorted_keys - The entire batch of sorted keys for this CF
+//
+// The per key status is returned in the KeyContext structures pointed to by
+// sorted_keys. An overall Status is also returned, with the only possible
+// values being Status::OK() and Status::TimedOut(). The latter indicates
+// that the call exceeded read_options.deadline
+Status DBImpl::MultiGetImpl(
    const ReadOptions& read_options, size_t start_key, size_t num_keys,
    autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE>* sorted_keys,
    SuperVersion* super_version, SequenceNumber snapshot,
@@ -2120,12 +2270,20 @@ void DBImpl::MultiGetImpl(
  // s is both in/out. When in, s could either be OK or MergeInProgress.
  // merge_operands will contain the sequence of merges in the latter case.
  size_t keys_left = num_keys;
+  Status s;
  while (keys_left) {
+    if (read_options.deadline.count() &&
+        env_->NowMicros() >
+            static_cast<uint64_t>(read_options.deadline.count())) {
+      s = Status::TimedOut();
+      break;
+    }
+
    size_t batch_size = (keys_left > MultiGetContext::MAX_BATCH_SIZE)
                            ? MultiGetContext::MAX_BATCH_SIZE
                            : keys_left;
    MultiGetContext ctx(sorted_keys, start_key + num_keys - keys_left,
-                        batch_size, snapshot);
+                        batch_size, snapshot, read_options);
    MultiGetRange range = ctx.GetMultiGetRange();
    bool lookup_current = false;

@@ -2163,13 +2321,21 @@ void DBImpl::MultiGetImpl(
  PERF_TIMER_GUARD(get_post_process_time);
  size_t num_found = 0;
  uint64_t bytes_read = 0;
-  for (size_t i = start_key; i < start_key + num_keys; ++i) {
+  for (size_t i = start_key; i < start_key + num_keys - keys_left; ++i) {
    KeyContext* key = (*sorted_keys)[i];
    if (key->s->ok()) {
      bytes_read += key->value->size();
      num_found++;
    }
  }
+  if (keys_left) {
+    assert(s.IsTimedOut());
+    for (size_t i = start_key + num_keys - keys_left; i < start_key + num_keys;
+         ++i) {
+      KeyContext* key = (*sorted_keys)[i];
+      *key->s = Status::TimedOut();
+    }
+  }

  RecordTick(stats_, NUMBER_MULTIGET_CALLS);
  RecordTick(stats_, NUMBER_MULTIGET_KEYS_READ, num_keys);
@@ -2178,6 +2344,8 @@ void DBImpl::MultiGetImpl(
  RecordInHistogram(stats_, BYTES_PER_MULTIGET, bytes_read);
  PERF_COUNTER_ADD(multiget_read_bytes, bytes_read);
  PERF_TIMER_STOP(get_post_process_time);
+
+  return s;
 }

 Status DBImpl::CreateColumnFamily(const ColumnFamilyOptions& cf_options,
@@ -2301,7 +2469,8 @@ Status DBImpl::CreateColumnFamilyImpl(const ColumnFamilyOptions& cf_options,
      auto* cfd =
          versions_->GetColumnFamilySet()->GetColumnFamily(column_family_name);
      assert(cfd != nullptr);
-      s = cfd->AddDirectories();
+      std::map<std::string, std::shared_ptr<FSDirectory>> dummy_created_dirs;
+      s = cfd->AddDirectories(&dummy_created_dirs);
    }
    if (s.ok()) {
      single_column_family_mode_ = false;
@@ -2435,7 +2604,8 @@ Status DBImpl::DropColumnFamilyImpl(ColumnFamilyHandle* column_family) {

 bool DBImpl::KeyMayExist(const ReadOptions& read_options,
                         ColumnFamilyHandle* column_family, const Slice& key,
-                         std::string* value, bool* value_found) {
+                         std::string* value, std::string* timestamp,
+                         bool* value_found) {
  assert(value != nullptr);
  if (value_found != nullptr) {
    // falsify later if key-may-exist but can't fetch value
@@ -2448,6 +2618,7 @@ bool DBImpl::KeyMayExist(const ReadOptions& read_options,
  get_impl_options.column_family = column_family;
  get_impl_options.value = &pinnable_val;
  get_impl_options.value_found = value_found;
+  get_impl_options.timestamp = timestamp;
  auto s = GetImpl(roptions, key, get_impl_options);
  value->assign(pinnable_val.data(), pinnable_val.size());

@@ -2463,6 +2634,12 @@ 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(
@@ -2487,7 +2664,8 @@ Iterator* DBImpl::NewIterator(const ReadOptions& read_options,

 #else
    SuperVersion* sv = cfd->GetReferencedSuperVersion(this);
-    auto iter = new ForwardIterator(this, read_options, cfd, sv);
+    auto iter = new ForwardIterator(this, read_options, cfd, sv,
+                                    /* allow_unprepared_value */ true);
    result = NewDBIterator(
        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
        cfd->user_comparator(), iter, kMaxSequenceNumber,
@@ -2560,11 +2738,12 @@ ArenaWrappedDBIter* DBImpl::NewIteratorImpl(const ReadOptions& read_options,
      env_, read_options, *cfd->ioptions(), sv->mutable_cf_options, snapshot,
      sv->mutable_cf_options.max_sequential_skip_in_iterations,
      sv->version_number, read_callback, this, cfd, allow_blob,
-      ((read_options.snapshot != nullptr) ? false : allow_refresh));
+      read_options.snapshot != nullptr ? false : allow_refresh);

  InternalIterator* internal_iter =
      NewInternalIterator(read_options, cfd, sv, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), snapshot);
+                          db_iter->GetRangeDelAggregator(), snapshot,
+                          /* allow_unprepared_value */ true);
  db_iter->SetIterUnderDBIter(internal_iter);

  return db_iter;
@@ -2592,7 +2771,8 @@ Status DBImpl::NewIterators(
    for (auto cfh : column_families) {
      auto cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(cfh)->cfd();
      SuperVersion* sv = cfd->GetReferencedSuperVersion(this);
-      auto iter = new ForwardIterator(this, read_options, cfd, sv);
+      auto iter = new ForwardIterator(this, read_options, cfd, sv,
+                                      /* allow_unprepared_value */ true);
      iterators->push_back(NewDBIterator(
          env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
          cfd->user_comparator(), iter, kMaxSequenceNumber,
@@ -3317,27 +3497,67 @@ Status DBImpl::CheckConsistency() {
  TEST_SYNC_POINT("DBImpl::CheckConsistency:AfterGetLiveFilesMetaData");

  std::string corruption_messages;
-  for (const auto& md : metadata) {
-    // md.name has a leading "/".
-    std::string file_path = md.db_path + md.name;

-    uint64_t fsize = 0;
-    TEST_SYNC_POINT("DBImpl::CheckConsistency:BeforeGetFileSize");
-    Status s = env_->GetFileSize(file_path, &fsize);
-    if (!s.ok() &&
-        env_->GetFileSize(Rocks2LevelTableFileName(file_path), &fsize).ok()) {
-      s = Status::OK();
+  if (immutable_db_options_.skip_checking_sst_file_sizes_on_db_open) {
+    // Instead of calling GetFileSize() for each expected file, call
+    // GetChildren() for the DB directory and check that all expected files
+    // are listed, without checking their sizes.
+    // Since sst files might be in different directories, do it for each
+    // directory separately.
+    std::map<std::string, std::vector<std::string>> files_by_directory;
+    for (const auto& md : metadata) {
+      // md.name has a leading "/". Remove it.
+      std::string fname = md.name;
+      if (!fname.empty() && fname[0] == '/') {
+        fname = fname.substr(1);
+      }
+      files_by_directory[md.db_path].push_back(fname);
    }
-    if (!s.ok()) {
-      corruption_messages +=
-          "Can't access " + md.name + ": " + s.ToString() + "\n";
-    } else if (fsize != md.size) {
-      corruption_messages += "Sst file size mismatch: " + file_path +
-                             ". Size recorded in manifest " +
-                             ToString(md.size) + ", actual size " +
-                             ToString(fsize) + "\n";
+    for (const auto& dir_files : files_by_directory) {
+      std::string directory = dir_files.first;
+      std::vector<std::string> existing_files;
+      Status s = env_->GetChildren(directory, &existing_files);
+      if (!s.ok()) {
+        corruption_messages +=
+            "Can't list files in " + directory + ": " + s.ToString() + "\n";
+        continue;
+      }
+      std::sort(existing_files.begin(), existing_files.end());
+
+      for (const std::string& fname : dir_files.second) {
+        if (!std::binary_search(existing_files.begin(), existing_files.end(),
+                                fname) &&
+            !std::binary_search(existing_files.begin(), existing_files.end(),
+                                Rocks2LevelTableFileName(fname))) {
+          corruption_messages +=
+              "Missing sst file " + fname + " in " + directory + "\n";
+        }
+      }
+    }
+  } else {
+    for (const auto& md : metadata) {
+      // md.name has a leading "/".
+      std::string file_path = md.db_path + md.name;
+
+      uint64_t fsize = 0;
+      TEST_SYNC_POINT("DBImpl::CheckConsistency:BeforeGetFileSize");
+      Status s = env_->GetFileSize(file_path, &fsize);
+      if (!s.ok() &&
+          env_->GetFileSize(Rocks2LevelTableFileName(file_path), &fsize).ok()) {
+        s = Status::OK();
+      }
+      if (!s.ok()) {
+        corruption_messages +=
+            "Can't access " + md.name + ": " + s.ToString() + "\n";
+      } else if (fsize != md.size) {
+        corruption_messages += "Sst file size mismatch: " + file_path +
+                               ". Size recorded in manifest " +
+                               ToString(md.size) + ", actual size " +
+                               ToString(fsize) + "\n";
+      }
    }
  }
+
  if (corruption_messages.size() == 0) {
    return Status::OK();
  } else {
@@ -3353,31 +3573,12 @@ Status DBImpl::GetDbIdentity(std::string& identity) const {
 Status DBImpl::GetDbIdentityFromIdentityFile(std::string* identity) const {
  std::string idfilename = IdentityFileName(dbname_);
  const FileOptions soptions;
-  std::unique_ptr<SequentialFileReader> id_file_reader;
-  Status s;
-  {
-    std::unique_ptr<FSSequentialFile> idfile;
-    s = fs_->NewSequentialFile(idfilename, soptions, &idfile, nullptr);
-    if (!s.ok()) {
-      return s;
-    }
-    id_file_reader.reset(
-        new SequentialFileReader(std::move(idfile), idfilename));
+
+  Status s = ReadFileToString(fs_.get(), idfilename, identity);
+  if (!s.ok()) {
+    return s;
  }

-  uint64_t file_size;
-  s = fs_->GetFileSize(idfilename, IOOptions(), &file_size, nullptr);
-  if (!s.ok()) {
-    return s;
-  }
-  char* buffer =
-      reinterpret_cast<char*>(alloca(static_cast<size_t>(file_size)));
-  Slice id;
-  s = id_file_reader->Read(static_cast<size_t>(file_size), &id, buffer);
-  if (!s.ok()) {
-    return s;
-  }
-  identity->assign(id.ToString());
  // If last character is '\n' remove it from identity
  if (identity->size() > 0 && identity->back() == '\n') {
    identity->pop_back();
@@ -3440,12 +3641,8 @@ Status DBImpl::Close() {
 Status DB::ListColumnFamilies(const DBOptions& db_options,
                              const std::string& name,
                              std::vector<std::string>* column_families) {
-  FileSystem* fs = db_options.file_system.get();
-  LegacyFileSystemWrapper legacy_fs(db_options.env);
-  if (!fs) {
-    fs = &legacy_fs;
-  }
-  return VersionSet::ListColumnFamilies(column_families, name, fs);
+  const std::shared_ptr<FileSystem>& fs = db_options.env->GetFileSystem();
+  return VersionSet::ListColumnFamilies(column_families, name, fs.get());
 }

 Snapshot::~Snapshot() {}
@@ -3489,24 +3686,15 @@ Status DestroyDB(const std::string& dbname, const Options& options,
      }
    }

-    std::vector<std::string> paths;
-
-    for (const auto& path : options.db_paths) {
-      paths.emplace_back(path.path);
+    std::set<std::string> paths;
+    for (const DbPath& db_path : options.db_paths) {
+      paths.insert(db_path.path);
    }
-    for (const auto& cf : column_families) {
-      for (const auto& path : cf.options.cf_paths) {
-        paths.emplace_back(path.path);
+    for (const ColumnFamilyDescriptor& cf : column_families) {
+      for (const DbPath& cf_path : cf.options.cf_paths) {
+        paths.insert(cf_path.path);
      }
    }
-
-    // Remove duplicate paths.
-    // Note that we compare only the actual paths but not path ids.
-    // This reason is that same path can appear at different path_ids
-    // for different column families.
-    std::sort(paths.begin(), paths.end());
-    paths.erase(std::unique(paths.begin(), paths.end()), paths.end());
-
    for (const auto& path : paths) {
      if (env->GetChildren(path, &filenames).ok()) {
        for (const auto& fname : filenames) {
@@ -3569,6 +3757,11 @@ Status DestroyDB(const std::string& dbname, const Options& options,

    env->UnlockFile(lock);  // Ignore error since state is already gone
    env->DeleteFile(lockname);
+
+    // 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
  }
  return result;
@@ -3792,8 +3985,9 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  *found_record_for_key = false;

  // Check if there is a record for this key in the latest memtable
-  sv->mem->Get(lkey, nullptr, &s, &merge_context, &max_covering_tombstone_seq,
-               seq, read_options, nullptr /*read_callback*/, is_blob_index);
+  sv->mem->Get(lkey, nullptr, nullptr, &s, &merge_context,
+               &max_covering_tombstone_seq, seq, read_options,
+               nullptr /*read_callback*/, is_blob_index);

  if (!(s.ok() || s.IsNotFound() || s.IsMergeInProgress())) {
    // unexpected error reading memtable.
@@ -3818,8 +4012,9 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  }

  // Check if there is a record for this key in the immutable memtables
-  sv->imm->Get(lkey, nullptr, &s, &merge_context, &max_covering_tombstone_seq,
-               seq, read_options, nullptr /*read_callback*/, is_blob_index);
+  sv->imm->Get(lkey, nullptr, nullptr, &s, &merge_context,
+               &max_covering_tombstone_seq, seq, read_options,
+               nullptr /*read_callback*/, is_blob_index);

  if (!(s.ok() || s.IsNotFound() || s.IsMergeInProgress())) {
    // unexpected error reading memtable.
@@ -3844,7 +4039,7 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  }

  // Check if there is a record for this key in the immutable memtables
-  sv->imm->GetFromHistory(lkey, nullptr, &s, &merge_context,
+  sv->imm->GetFromHistory(lkey, nullptr, nullptr, &s, &merge_context,
                          &max_covering_tombstone_seq, seq, read_options,
                          is_blob_index);

@@ -3872,7 +4067,7 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  // SST files if cache_only=true?
  if (!cache_only) {
    // Check tables
-    sv->current->Get(read_options, lkey, nullptr, &s, &merge_context,
+    sv->current->Get(read_options, lkey, nullptr, nullptr, &s, &merge_context,
                     &max_covering_tombstone_seq, nullptr /* value_found */,
                     found_record_for_key, seq, nullptr /*read_callback*/,
                     is_blob_index);
@@ -4349,8 +4544,8 @@ Status DBImpl::VerifyChecksum(const ReadOptions& read_options) {
        const auto& fd = vstorage->LevelFilesBrief(i).files[j].fd;
        std::string fname = TableFileName(cfd->ioptions()->cf_paths,
                                          fd.GetNumber(), fd.GetPathId());
-        s = rocksdb::VerifySstFileChecksum(opts, file_options_, read_options,
-                                           fname);
+        s = ROCKSDB_NAMESPACE::VerifySstFileChecksum(opts, file_options_,
+                                                     read_options, fname);
      }
    }
    if (!s.ok()) {
@@ -4495,13 +4690,21 @@ Status DBImpl::GetCreationTimeOfOldestFile(uint64_t* creation_time) {
  if (mutable_db_options_.max_open_files == -1) {
    uint64_t oldest_time = port::kMaxUint64;
    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      uint64_t ctime;
-      cfd->current()->GetCreationTimeOfOldestFile(&ctime);
-      if (ctime < oldest_time) {
-        oldest_time = ctime;
-      }
-      if (oldest_time == 0) {
-        break;
+      if (!cfd->IsDropped()) {
+        uint64_t ctime;
+        {
+          SuperVersion* sv = GetAndRefSuperVersion(cfd);
+          Version* version = sv->current;
+          version->GetCreationTimeOfOldestFile(&ctime);
+          ReturnAndCleanupSuperVersion(cfd, sv);
+        }
+
+        if (ctime < oldest_time) {
+          oldest_time = ctime;
+        }
+        if (oldest_time == 0) {
+          break;
+        }
      }
    }
    *creation_time = oldest_time;
@@ -4512,4 +4715,4 @@ Status DBImpl::GetCreationTimeOfOldestFile(uint64_t* creation_time) {
 }
 #endif  // ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -62,7 +62,7 @@
 #include "util/stop_watch.h"
 #include "util/thread_local.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class Arena;
 class ArenaWrappedDBIter;
@@ -82,13 +82,13 @@ struct MemTableInfo;
 // Class to maintain directories for all database paths other than main one.
 class Directories {
 public:
-  Status SetDirectories(Env* env, const std::string& dbname,
-                        const std::string& wal_dir,
-                        const std::vector<DbPath>& data_paths);
+  IOStatus SetDirectories(FileSystem* fs, const std::string& dbname,
+                          const std::string& wal_dir,
+                          const std::vector<DbPath>& data_paths);

-  Directory* GetDataDir(size_t path_id) const {
+  FSDirectory* GetDataDir(size_t path_id) const {
    assert(path_id < data_dirs_.size());
-    Directory* ret_dir = data_dirs_[path_id].get();
+    FSDirectory* ret_dir = data_dirs_[path_id].get();
    if (ret_dir == nullptr) {
      // Should use db_dir_
      return db_dir_.get();
@@ -96,19 +96,19 @@ class Directories {
    return ret_dir;
  }

-  Directory* GetWalDir() {
+  FSDirectory* GetWalDir() {
    if (wal_dir_) {
      return wal_dir_.get();
    }
    return db_dir_.get();
  }

-  Directory* GetDbDir() { return db_dir_.get(); }
+  FSDirectory* GetDbDir() { return db_dir_.get(); }

 private:
-  std::unique_ptr<Directory> db_dir_;
-  std::vector<std::unique_ptr<Directory>> data_dirs_;
-  std::unique_ptr<Directory> wal_dir_;
+  std::unique_ptr<FSDirectory> db_dir_;
+  std::vector<std::unique_ptr<FSDirectory>> data_dirs_;
+  std::unique_ptr<FSDirectory> wal_dir_;
 };

 // While DB is the public interface of RocksDB, and DBImpl is the actual
@@ -163,6 +163,9 @@ class DBImpl : public DB {
  virtual Status Get(const ReadOptions& options,
                     ColumnFamilyHandle* column_family, const Slice& key,
                     PinnableSlice* value) override;
+  virtual Status Get(const ReadOptions& options,
+                     ColumnFamilyHandle* column_family, const Slice& key,
+                     PinnableSlice* value, std::string* timestamp) override;

  using DB::GetMergeOperands;
  Status GetMergeOperands(const ReadOptions& options,
@@ -185,6 +188,11 @@ class DBImpl : public DB {
      const std::vector<ColumnFamilyHandle*>& column_family,
      const std::vector<Slice>& keys,
      std::vector<std::string>* values) override;
+  virtual std::vector<Status> MultiGet(
+      const ReadOptions& options,
+      const std::vector<ColumnFamilyHandle*>& column_family,
+      const std::vector<Slice>& keys, std::vector<std::string>* values,
+      std::vector<std::string>* timestamps) override;

  // This MultiGet is a batched version, which may be faster than calling Get
  // multiple times, especially if the keys have some spatial locality that
@@ -198,11 +206,22 @@ class DBImpl : public DB {
                        const size_t num_keys, const Slice* keys,
                        PinnableSlice* values, Status* statuses,
                        const bool sorted_input = false) override;
+  virtual void MultiGet(const ReadOptions& options,
+                        ColumnFamilyHandle* column_family,
+                        const size_t num_keys, const Slice* keys,
+                        PinnableSlice* values, std::string* timestamps,
+                        Status* statuses,
+                        const bool sorted_input = false) override;

  virtual void MultiGet(const ReadOptions& options, const size_t num_keys,
                        ColumnFamilyHandle** column_families, const Slice* keys,
                        PinnableSlice* values, Status* statuses,
                        const bool sorted_input = false) override;
+  virtual void MultiGet(const ReadOptions& options, const size_t num_keys,
+                        ColumnFamilyHandle** column_families, const Slice* keys,
+                        PinnableSlice* values, std::string* timestamps,
+                        Status* statuses,
+                        const bool sorted_input = false) override;

  virtual void MultiGetWithCallback(
      const ReadOptions& options, ColumnFamilyHandle* column_family,
@@ -230,7 +249,7 @@ class DBImpl : public DB {
  using DB::KeyMayExist;
  virtual bool KeyMayExist(const ReadOptions& options,
                           ColumnFamilyHandle* column_family, const Slice& key,
-                           std::string* value,
+                           std::string* value, std::string* timestamp,
                           bool* value_found = nullptr) override;

  using DB::NewIterator;
@@ -433,6 +452,7 @@ class DBImpl : public DB {
  struct GetImplOptions {
    ColumnFamilyHandle* column_family = nullptr;
    PinnableSlice* value = nullptr;
+    std::string* timestamp = nullptr;
    bool* value_found = nullptr;
    ReadCallback* callback = nullptr;
    bool* is_blob_index = nullptr;
@@ -455,7 +475,7 @@ class DBImpl : public DB {
  // If get_impl_options.get_value = false get merge operands associated with
  // get_impl_options.key via get_impl_options.merge_operands
  Status GetImpl(const ReadOptions& options, const Slice& key,
-                 GetImplOptions get_impl_options);
+                 GetImplOptions& get_impl_options);

  ArenaWrappedDBIter* NewIteratorImpl(const ReadOptions& options,
                                      ColumnFamilyData* cfd,
@@ -561,9 +581,14 @@ class DBImpl : public DB {
  // Return an internal iterator over the current state of the database.
  // The keys of this iterator are internal keys (see format.h).
  // The returned iterator should be deleted when no longer needed.
+  // If allow_unprepared_value is true, the returned iterator may defer reading
+  // 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.
  InternalIterator* NewInternalIterator(
      Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
-      ColumnFamilyHandle* column_family = nullptr);
+      ColumnFamilyHandle* column_family = nullptr,
+      bool allow_unprepared_value = false);

  LogsWithPrepTracker* logs_with_prep_tracker() {
    return &logs_with_prep_tracker_;
@@ -689,7 +714,8 @@ class DBImpl : public DB {

  InternalIterator* NewInternalIterator(
      const ReadOptions&, ColumnFamilyData* cfd, SuperVersion* super_version,
-      Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence);
+      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
@@ -826,9 +852,9 @@ class DBImpl : public DB {
                     std::vector<ColumnFamilyHandle*>* handles, DB** dbptr,
                     const bool seq_per_batch, const bool batch_per_txn);

-
-  static Status CreateAndNewDirectory(Env* env, const std::string& dirname,
-                                      std::unique_ptr<Directory>* directory);
+  static IOStatus CreateAndNewDirectory(
+      FileSystem* fs, const std::string& dirname,
+      std::unique_ptr<FSDirectory>* directory);

  // find stats map from stats_history_ with smallest timestamp in
  // the range of [start_time, end_time)
@@ -943,6 +969,12 @@ class DBImpl : public DB {
  void TEST_WaitForPersistStatsRun(std::function<void()> callback) const;
  bool TEST_IsPersistentStatsEnabled() const;
  size_t TEST_EstimateInMemoryStatsHistorySize() const;
+
+  VersionSet* TEST_GetVersionSet() const { return versions_.get(); }
+
+  const std::unordered_set<uint64_t>& TEST_GetFilesGrabbedForPurge() const {
+    return files_grabbed_for_purge_;
+  }
 #endif  // NDEBUG

 protected:
@@ -1110,13 +1142,25 @@ class DBImpl : public DB {
  // Recover the descriptor from persistent storage.  May do a significant
  // amount of work to recover recently logged updates.  Any changes to
  // be made to the descriptor are added to *edit.
+  // recovered_seq is set to less than kMaxSequenceNumber if the log's tail is
+  // skipped.
  virtual Status Recover(
      const std::vector<ColumnFamilyDescriptor>& column_families,
      bool read_only = false, bool error_if_log_file_exist = false,
-      bool error_if_data_exists_in_logs = false);
+      bool error_if_data_exists_in_logs = false,
+      uint64_t* recovered_seq = nullptr);

  virtual bool OwnTablesAndLogs() const { return true; }

+  // REQUIRES: db mutex held when calling this function, but the db mutex can
+  // be released and re-acquired. Db mutex will be held when the function
+  // returns.
+  // After best-efforts recovery, there may be SST files in db/cf paths that are
+  // not referenced in the MANIFEST. We delete these SST files. In the
+  // meantime, we find out the largest file number present in the paths, and
+  // bump up the version set's next_file_number_ to be 1 + largest_file_number.
+  Status FinishBestEffortsRecovery();
+
 private:
  friend class DB;
  friend class ErrorHandler;
@@ -1330,7 +1374,7 @@ class DBImpl : public DB {
  void ReleaseFileNumberFromPendingOutputs(
      std::unique_ptr<std::list<uint64_t>::iterator>& v);

-  Status SyncClosedLogs(JobContext* job_context);
+  IOStatus SyncClosedLogs(JobContext* job_context);

  // Flush the in-memory write buffer to storage.  Switches to a new
  // log-file/memtable and writes a new descriptor iff successful. Then
@@ -1355,8 +1399,10 @@ class DBImpl : public DB {
      JobContext* job_context, LogBuffer* log_buffer, Env::Priority thread_pri);

  // REQUIRES: log_numbers are sorted in ascending order
+  // corrupted_log_found is set to true if we recover from a corrupted log file.
  Status RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
-                         SequenceNumber* next_sequence, bool read_only);
+                         SequenceNumber* next_sequence, bool read_only,
+                         bool* corrupted_log_found);

  // The following two methods are used to flush a memtable to
  // storage. The first one is used at database RecoveryTime (when the
@@ -1379,6 +1425,7 @@ class DBImpl : public DB {
  Status ThrottleLowPriWritesIfNeeded(const WriteOptions& write_options,
                                      WriteBatch* my_batch);

+  // REQUIRES: mutex locked and in write thread.
  Status ScheduleFlushes(WriteContext* context);

  void MaybeFlushStatsCF(autovector<ColumnFamilyData*>* cfds);
@@ -1450,10 +1497,10 @@ class DBImpl : public DB {
  // REQUIRES: mutex locked and in write thread.
  void AssignAtomicFlushSeq(const autovector<ColumnFamilyData*>& cfds);

-  // REQUIRES: mutex locked
+  // REQUIRES: mutex locked and in write thread.
  Status SwitchWAL(WriteContext* write_context);

-  // REQUIRES: mutex locked
+  // REQUIRES: mutex locked and in write thread.
  Status HandleWriteBufferFull(WriteContext* write_context);

  // REQUIRES: mutex locked
@@ -1464,21 +1511,25 @@ class DBImpl : public DB {
                         WriteBatch* tmp_batch, size_t* write_with_wal,
                         WriteBatch** to_be_cached_state);

-  Status WriteToWAL(const WriteBatch& merged_batch, log::Writer* log_writer,
-                    uint64_t* log_used, uint64_t* log_size);
+  IOStatus WriteToWAL(const WriteBatch& merged_batch, log::Writer* log_writer,
+                      uint64_t* log_used, uint64_t* log_size);

-  Status WriteToWAL(const WriteThread::WriteGroup& write_group,
-                    log::Writer* log_writer, uint64_t* log_used,
-                    bool need_log_sync, bool need_log_dir_sync,
-                    SequenceNumber sequence);
+  IOStatus WriteToWAL(const WriteThread::WriteGroup& write_group,
+                      log::Writer* log_writer, uint64_t* log_used,
+                      bool need_log_sync, bool need_log_dir_sync,
+                      SequenceNumber sequence);

-  Status ConcurrentWriteToWAL(const WriteThread::WriteGroup& write_group,
-                              uint64_t* log_used, SequenceNumber* last_sequence,
-                              size_t seq_inc);
+  IOStatus ConcurrentWriteToWAL(const WriteThread::WriteGroup& write_group,
+                                uint64_t* log_used,
+                                SequenceNumber* last_sequence, size_t seq_inc);

  // Used by WriteImpl to update bg_error_ if paranoid check is enabled.
  void WriteStatusCheck(const Status& status);

+  // Used by WriteImpl to update bg_error_ when IO error happens, e.g., write
+  // WAL, sync WAL fails, if paranoid check is enabled.
+  void IOStatusCheck(const IOStatus& status);
+
  // Used by WriteImpl to update bg_error_ in case of memtable insert error.
  void MemTableInsertStatusCheck(const Status& memtable_insert_status);

@@ -1593,7 +1644,7 @@ class DBImpl : public DB {

  uint64_t GetMaxTotalWalSize() const;

-  Directory* GetDataDir(ColumnFamilyData* cfd, size_t path_id) const;
+  FSDirectory* GetDataDir(ColumnFamilyData* cfd, size_t path_id) const;

  Status CloseHelper();

@@ -1722,7 +1773,7 @@ class DBImpl : public DB {
  // to have acquired the SuperVersion and pass in a snapshot sequence number
  // in order to construct the LookupKeys. The start_key and num_keys specify
  // the range of keys in the sorted_keys vector for a single column family.
-  void MultiGetImpl(
+  Status MultiGetImpl(
      const ReadOptions& read_options, size_t start_key, size_t num_keys,
      autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE>* sorted_keys,
      SuperVersion* sv, SequenceNumber snap_seqnum, ReadCallback* callback,
@@ -1890,7 +1941,7 @@ class DBImpl : public DB {
  std::unordered_map<uint64_t, PurgeFileInfo> purge_files_;

  // A vector to store the file numbers that have been assigned to certain
-  // JobContext. Current implementation tracks ssts only.
+  // JobContext. Current implementation tracks table and blob files only.
  std::unordered_set<uint64_t> files_grabbed_for_purge_;

  // A queue to store log writers to close
@@ -2010,13 +2061,13 @@ class DBImpl : public DB {

  // handle for scheduling stats dumping at fixed intervals
  // REQUIRES: mutex locked
-  std::unique_ptr<rocksdb::RepeatableThread> thread_dump_stats_;
+  std::unique_ptr<ROCKSDB_NAMESPACE::RepeatableThread> thread_dump_stats_;

  // handle for scheduling stats snapshoting at fixed intervals
  // REQUIRES: mutex locked
-  std::unique_ptr<rocksdb::RepeatableThread> thread_persist_stats_;
+  std::unique_ptr<ROCKSDB_NAMESPACE::RepeatableThread> thread_persist_stats_;

-  // When set, we use a separate queue for writes that dont write to memtable.
+  // When set, we use a separate queue for writes that don't write to memtable.
  // In 2PC these are the writes at Prepare phase.
  const bool two_write_queues_;
  const bool manual_wal_flush_;
@@ -2099,4 +2150,4 @@ static void ClipToRange(T* ptr, V minvalue, V maxvalue) {
  if (static_cast<V>(*ptr) < minvalue) *ptr = minvalue;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -22,7 +22,7 @@
 #include "util/cast_util.h"
 #include "util/concurrent_task_limiter_impl.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 bool DBImpl::EnoughRoomForCompaction(
    ColumnFamilyData* cfd, const std::vector<CompactionInputFiles>& inputs,
@@ -79,7 +79,7 @@ bool DBImpl::RequestCompactionToken(ColumnFamilyData* cfd, bool force,
  return false;
 }

-Status DBImpl::SyncClosedLogs(JobContext* job_context) {
+IOStatus DBImpl::SyncClosedLogs(JobContext* job_context) {
  TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Start");
  mutex_.AssertHeld();
  autovector<log::Writer*, 1> logs_to_sync;
@@ -96,7 +96,7 @@ Status DBImpl::SyncClosedLogs(JobContext* job_context) {
    logs_to_sync.push_back(log.writer);
  }

-  Status s;
+  IOStatus io_s;
  if (!logs_to_sync.empty()) {
    mutex_.Unlock();

@@ -104,34 +104,34 @@ Status DBImpl::SyncClosedLogs(JobContext* job_context) {
      ROCKS_LOG_INFO(immutable_db_options_.info_log,
                     "[JOB %d] Syncing log #%" PRIu64, job_context->job_id,
                     log->get_log_number());
-      s = log->file()->Sync(immutable_db_options_.use_fsync);
-      if (!s.ok()) {
+      io_s = log->file()->Sync(immutable_db_options_.use_fsync);
+      if (!io_s.ok()) {
        break;
      }

      if (immutable_db_options_.recycle_log_file_num > 0) {
-        s = log->Close();
-        if (!s.ok()) {
+        io_s = log->Close();
+        if (!io_s.ok()) {
          break;
        }
      }
    }
-    if (s.ok()) {
-      s = directories_.GetWalDir()->Fsync();
+    if (io_s.ok()) {
+      io_s = directories_.GetWalDir()->Fsync(IOOptions(), nullptr);
    }

    mutex_.Lock();

    // "number <= current_log_number - 1" is equivalent to
    // "number < current_log_number".
-    MarkLogsSynced(current_log_number - 1, true, s);
-    if (!s.ok()) {
-      error_handler_.SetBGError(s, BackgroundErrorReason::kFlush);
+    MarkLogsSynced(current_log_number - 1, true, io_s);
+    if (!io_s.ok()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
      TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Failed");
-      return s;
+      return io_s;
    }
  }
-  return s;
+  return io_s;
 }

 Status DBImpl::FlushMemTableToOutputFile(
@@ -155,7 +155,6 @@ Status DBImpl::FlushMemTableToOutputFile(
      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);
-
  FileMetaData file_meta;

  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:BeforePickMemtables");
@@ -168,6 +167,7 @@ Status DBImpl::FlushMemTableToOutputFile(
 #endif  // ROCKSDB_LITE

  Status s;
+  IOStatus io_s;
  if (logfile_number_ > 0 &&
      versions_->GetColumnFamilySet()->NumberOfColumnFamilies() > 1) {
    // If there are more than one column families, we need to make sure that
@@ -176,7 +176,8 @@ Status DBImpl::FlushMemTableToOutputFile(
    // flushed SST may contain data from write batches whose updates to
    // other column families are missing.
    // SyncClosedLogs() may unlock and re-lock the db_mutex.
-    s = SyncClosedLogs(job_context);
+    io_s = SyncClosedLogs(job_context);
+    s = io_s;
  } else {
    TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Skip");
  }
@@ -192,6 +193,7 @@ Status DBImpl::FlushMemTableToOutputFile(
  } else {
    flush_job.Cancel();
  }
+  io_s = flush_job.io_status();

  if (s.ok()) {
    InstallSuperVersionAndScheduleWork(cfd, superversion_context,
@@ -206,8 +208,13 @@ Status DBImpl::FlushMemTableToOutputFile(
  }

  if (!s.ok() && !s.IsShutdownInProgress() && !s.IsColumnFamilyDropped()) {
-    Status new_bg_error = s;
-    error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
+    if (!io_s.ok() && !io_s.IsShutdownInProgress() &&
+        !io_s.IsColumnFamilyDropped()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+    } else {
+      Status new_bg_error = s;
+      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
+    }
  }
  if (s.ok()) {
 #ifndef ROCKSDB_LITE
@@ -301,7 +308,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
  GetSnapshotContext(job_context, &snapshot_seqs,
                     &earliest_write_conflict_snapshot, &snapshot_checker);

-  autovector<Directory*> distinct_output_dirs;
+  autovector<FSDirectory*> distinct_output_dirs;
  autovector<std::string> distinct_output_dir_paths;
  std::vector<std::unique_ptr<FlushJob>> jobs;
  std::vector<MutableCFOptions> all_mutable_cf_options;
@@ -309,7 +316,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
  all_mutable_cf_options.reserve(num_cfs);
  for (int i = 0; i < num_cfs; ++i) {
    auto cfd = cfds[i];
-    Directory* data_dir = GetDataDir(cfd, 0U);
+    FSDirectory* data_dir = GetDataDir(cfd, 0U);
    const std::string& curr_path = cfd->ioptions()->cf_paths[0].path;

    // Add to distinct output directories if eligible. Use linear search. Since
@@ -344,6 +351,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(

  std::vector<FileMetaData> file_meta(num_cfs);
  Status s;
+  IOStatus io_s;
  assert(num_cfs == static_cast<int>(jobs.size()));

 #ifndef ROCKSDB_LITE
@@ -358,15 +366,18 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
  if (logfile_number_ > 0) {
    // TODO (yanqin) investigate whether we should sync the closed logs for
    // single column family case.
-    s = SyncClosedLogs(job_context);
+    io_s = SyncClosedLogs(job_context);
+    s = io_s;
  }

  // exec_status stores the execution status of flush_jobs as
  // <bool /* executed */, Status /* status code */>
  autovector<std::pair<bool, Status>> exec_status;
+  autovector<IOStatus> io_status;
  for (int i = 0; i != num_cfs; ++i) {
    // Initially all jobs are not executed, with status OK.
    exec_status.emplace_back(false, Status::OK());
+    io_status.emplace_back(IOStatus::OK());
  }

  if (s.ok()) {
@@ -375,6 +386,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
      exec_status[i].second =
          jobs[i]->Run(&logs_with_prep_tracker_, &file_meta[i]);
      exec_status[i].first = true;
+      io_status[i] = jobs[i]->io_status();
    }
    if (num_cfs > 1) {
      TEST_SYNC_POINT(
@@ -387,6 +399,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
    exec_status[0].second =
        jobs[0]->Run(&logs_with_prep_tracker_, &file_meta[0]);
    exec_status[0].first = true;
+    io_status[0] = jobs[0]->io_status();

    Status error_status;
    for (const auto& e : exec_status) {
@@ -405,21 +418,52 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
    s = error_status.ok() ? s : error_status;
  }

+  if (io_s.ok()) {
+    IOStatus io_error = IOStatus::OK();
+    for (int i = 0; i != static_cast<int>(io_status.size()); i++) {
+      if (!io_status[i].ok() && !io_status[i].IsShutdownInProgress() &&
+          !io_status[i].IsColumnFamilyDropped()) {
+        io_error = io_status[i];
+      }
+    }
+    io_s = io_error;
+    if (s.ok() && !io_s.ok()) {
+      s = io_s;
+    }
+  }
+
  if (s.IsColumnFamilyDropped()) {
    s = Status::OK();
  }

-  if (s.ok() || s.IsShutdownInProgress() || s.IsColumnFamilyDropped()) {
+  if (s.ok() || s.IsShutdownInProgress()) {
    // Sync on all distinct output directories.
    for (auto dir : distinct_output_dirs) {
      if (dir != nullptr) {
-        Status error_status = dir->Fsync();
+        Status error_status = dir->Fsync(IOOptions(), nullptr);
        if (!error_status.ok()) {
          s = error_status;
          break;
        }
      }
    }
+  } else {
+    // Need to undo atomic flush if something went wrong, i.e. s is not OK and
+    // it is not because of CF drop.
+    // Have to cancel the flush jobs that have NOT executed because we need to
+    // unref the versions.
+    for (int i = 0; i != num_cfs; ++i) {
+      if (!exec_status[i].first) {
+        jobs[i]->Cancel();
+      }
+    }
+    for (int i = 0; i != num_cfs; ++i) {
+      if (exec_status[i].first && exec_status[i].second.ok()) {
+        auto& mems = jobs[i]->GetMemTables();
+        cfds[i]->imm()->RollbackMemtableFlush(mems,
+                                              file_meta[i].fd.GetNumber());
+      }
+    }
  }

  if (s.ok()) {
@@ -529,22 +573,12 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
  // Need to undo atomic flush if something went wrong, i.e. s is not OK and
  // it is not because of CF drop.
  if (!s.ok() && !s.IsColumnFamilyDropped()) {
-    // Have to cancel the flush jobs that have NOT executed because we need to
-    // unref the versions.
-    for (int i = 0; i != num_cfs; ++i) {
-      if (!exec_status[i].first) {
-        jobs[i]->Cancel();
-      }
+    if (!io_s.ok() && io_s.IsColumnFamilyDropped()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+    } else {
+      Status new_bg_error = s;
+      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
    }
-    for (int i = 0; i != num_cfs; ++i) {
-      if (exec_status[i].first && exec_status[i].second.ok()) {
-        auto& mems = jobs[i]->GetMemTables();
-        cfds[i]->imm()->RollbackMemtableFlush(mems,
-                                              file_meta[i].fd.GetNumber());
-      }
-    }
-    Status new_bg_error = s;
-    error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
  }

  return s;
@@ -651,8 +685,6 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
    return Status::InvalidArgument("Invalid target path ID");
  }

-  bool exclusive = options.exclusive_manual_compaction;
-
  bool flush_needed = true;
  if (begin != nullptr && end != nullptr) {
    // TODO(ajkr): We could also optimize away the flush in certain cases where
@@ -685,25 +717,9 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
    }
  }

-  int max_level_with_files = 0;
-  // max_file_num_to_ignore can be used to filter out newly created SST files,
-  // useful for bottom level compaction in a manual compaction
-  uint64_t max_file_num_to_ignore = port::kMaxUint64;
-  uint64_t next_file_number = port::kMaxUint64;
-  {
-    InstrumentedMutexLock l(&mutex_);
-    Version* base = cfd->current();
-    for (int level = 1; level < base->storage_info()->num_non_empty_levels();
-         level++) {
-      if (base->storage_info()->OverlapInLevel(level, begin, end)) {
-        max_level_with_files = level;
-      }
-    }
-    next_file_number = versions_->current_next_file_number();
-  }
-
-  int final_output_level = 0;
-
+  constexpr int kInvalidLevel = -1;
+  int final_output_level = kInvalidLevel;
+  bool exclusive = options.exclusive_manual_compaction;
  if (cfd->ioptions()->compaction_style == kCompactionStyleUniversal &&
      cfd->NumberLevels() > 1) {
    // Always compact all files together.
@@ -714,58 +730,98 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
    }
    s = RunManualCompaction(cfd, ColumnFamilyData::kCompactAllLevels,
                            final_output_level, options, begin, end, exclusive,
-                            false, max_file_num_to_ignore);
+                            false, port::kMaxUint64);
  } else {
-    for (int level = 0; level <= max_level_with_files; level++) {
+    int first_overlapped_level = kInvalidLevel;
+    int max_overlapped_level = kInvalidLevel;
+    {
+      SuperVersion* super_version = cfd->GetReferencedSuperVersion(this);
+      Version* current_version = super_version->current;
+      ReadOptions ro;
+      ro.total_order_seek = true;
+      bool overlap;
+      for (int level = 0;
+           level < current_version->storage_info()->num_non_empty_levels();
+           level++) {
+        overlap = true;
+        if (begin != nullptr && end != nullptr) {
+          Status status = current_version->OverlapWithLevelIterator(
+              ro, file_options_, *begin, *end, level, &overlap);
+          if (!status.ok()) {
+            overlap = current_version->storage_info()->OverlapInLevel(
+                level, begin, end);
+          }
+        } else {
+          overlap = current_version->storage_info()->OverlapInLevel(level,
+                                                                    begin, end);
+        }
+        if (overlap) {
+          if (first_overlapped_level == kInvalidLevel) {
+            first_overlapped_level = level;
+          }
+          max_overlapped_level = level;
+        }
+      }
+      CleanupSuperVersion(super_version);
+    }
+    if (s.ok() && first_overlapped_level != kInvalidLevel) {
+      // max_file_num_to_ignore can be used to filter out newly created SST
+      // files, useful for bottom level compaction in a manual compaction
+      uint64_t max_file_num_to_ignore = port::kMaxUint64;
+      uint64_t next_file_number = versions_->current_next_file_number();
+      final_output_level = max_overlapped_level;
      int output_level;
-      // in case the compaction is universal or if we're compacting the
-      // bottom-most level, the output level will be the same as input one.
-      // level 0 can never be the bottommost level (i.e. if all files are in
-      // level 0, we will compact to level 1)
-      if (cfd->ioptions()->compaction_style == kCompactionStyleUniversal ||
-          cfd->ioptions()->compaction_style == kCompactionStyleFIFO) {
-        output_level = level;
-      } else if (level == max_level_with_files && level > 0) {
-        if (options.bottommost_level_compaction ==
-            BottommostLevelCompaction::kSkip) {
-          // Skip bottommost level compaction
-          continue;
-        } else if (options.bottommost_level_compaction ==
-                       BottommostLevelCompaction::kIfHaveCompactionFilter &&
-                   cfd->ioptions()->compaction_filter == nullptr &&
-                   cfd->ioptions()->compaction_filter_factory == nullptr) {
-          // Skip bottommost level compaction since we don't have a compaction
-          // filter
-          continue;
+      for (int level = first_overlapped_level; level <= max_overlapped_level;
+           level++) {
+        // in case the compaction is universal or if we're compacting the
+        // bottom-most level, the output level will be the same as input one.
+        // level 0 can never be the bottommost level (i.e. if all files are in
+        // level 0, we will compact to level 1)
+        if (cfd->ioptions()->compaction_style == kCompactionStyleUniversal ||
+            cfd->ioptions()->compaction_style == kCompactionStyleFIFO) {
+          output_level = level;
+        } else if (level == max_overlapped_level && level > 0) {
+          if (options.bottommost_level_compaction ==
+              BottommostLevelCompaction::kSkip) {
+            // Skip bottommost level compaction
+            continue;
+          } else if (options.bottommost_level_compaction ==
+                         BottommostLevelCompaction::kIfHaveCompactionFilter &&
+                     cfd->ioptions()->compaction_filter == nullptr &&
+                     cfd->ioptions()->compaction_filter_factory == nullptr) {
+            // Skip bottommost level compaction since we don't have a compaction
+            // filter
+            continue;
+          }
+          output_level = level;
+          // update max_file_num_to_ignore only for bottom level compaction
+          // because data in newly compacted files in middle levels may still
+          // need to be pushed down
+          max_file_num_to_ignore = next_file_number;
+        } else {
+          output_level = level + 1;
+          if (cfd->ioptions()->compaction_style == kCompactionStyleLevel &&
+              cfd->ioptions()->level_compaction_dynamic_level_bytes &&
+              level == 0) {
+            output_level = ColumnFamilyData::kCompactToBaseLevel;
+          }
        }
-        output_level = level;
-        // update max_file_num_to_ignore only for bottom level compaction
-        // because data in newly compacted files in middle levels may still need
-        // to be pushed down
-        max_file_num_to_ignore = next_file_number;
-      } else {
-        output_level = level + 1;
-        if (cfd->ioptions()->compaction_style == kCompactionStyleLevel &&
-            cfd->ioptions()->level_compaction_dynamic_level_bytes &&
-            level == 0) {
-          output_level = ColumnFamilyData::kCompactToBaseLevel;
+        s = RunManualCompaction(cfd, level, output_level, options, begin, end,
+                                exclusive, false, max_file_num_to_ignore);
+        if (!s.ok()) {
+          break;
        }
+        if (output_level == ColumnFamilyData::kCompactToBaseLevel) {
+          final_output_level = cfd->NumberLevels() - 1;
+        } else if (output_level > final_output_level) {
+          final_output_level = output_level;
+        }
+        TEST_SYNC_POINT("DBImpl::RunManualCompaction()::1");
+        TEST_SYNC_POINT("DBImpl::RunManualCompaction()::2");
      }
-      s = RunManualCompaction(cfd, level, output_level, options, begin, end,
-                              exclusive, false, max_file_num_to_ignore);
-      if (!s.ok()) {
-        break;
-      }
-      if (output_level == ColumnFamilyData::kCompactToBaseLevel) {
-        final_output_level = cfd->NumberLevels() - 1;
-      } else if (output_level > final_output_level) {
-        final_output_level = output_level;
-      }
-      TEST_SYNC_POINT("DBImpl::RunManualCompaction()::1");
-      TEST_SYNC_POINT("DBImpl::RunManualCompaction()::2");
    }
  }
-  if (!s.ok()) {
+  if (!s.ok() || final_output_level == kInvalidLevel) {
    LogFlush(immutable_db_options_.info_log);
    return s;
  }
@@ -1036,7 +1092,7 @@ Status DBImpl::CompactFilesImpl(
  }

  if (output_file_names != nullptr) {
-    for (const auto newf : c->edit()->GetNewFiles()) {
+    for (const auto& newf : c->edit()->GetNewFiles()) {
      (*output_file_names)
          .push_back(TableFileName(c->immutable_cf_options()->cf_paths,
                                   newf.second.fd.GetNumber(),
@@ -1136,7 +1192,7 @@ void DBImpl::NotifyOnCompactionBegin(ColumnFamilyData* cfd, Compaction* c,
        }
      }
    }
-    for (const auto newf : c->edit()->GetNewFiles()) {
+    for (const auto& newf : c->edit()->GetNewFiles()) {
      const FileMetaData& meta = newf.second;
      const FileDescriptor& desc = meta.fd;
      const uint64_t file_number = desc.GetNumber();
@@ -1257,7 +1313,8 @@ Status DBImpl::ReFitLevel(ColumnFamilyData* cfd, int level, int target_level) {
                   f->fd.GetFileSize(), f->smallest, f->largest,
                   f->fd.smallest_seqno, f->fd.largest_seqno,
                   f->marked_for_compaction, f->oldest_blob_file_number,
-                   f->oldest_ancester_time, f->file_creation_time);
+                   f->oldest_ancester_time, f->file_creation_time,
+                   f->file_checksum, f->file_checksum_func_name);
    }
    ROCKS_LOG_DEBUG(immutable_db_options_.info_log,
                    "[%s] Apply version edit:\n%s", cfd->GetName().c_str(),
@@ -1486,7 +1543,12 @@ Status DBImpl::RunManualCompaction(
      }
      manual.incomplete = false;
      bg_compaction_scheduled_++;
-      env_->Schedule(&DBImpl::BGWorkCompaction, ca, Env::Priority::LOW, this,
+      Env::Priority thread_pool_pri = Env::Priority::LOW;
+      if (compaction->bottommost_level() &&
+          env_->GetBackgroundThreads(Env::Priority::BOTTOM) > 0) {
+        thread_pool_pri = Env::Priority::BOTTOM;
+      }
+      env_->Schedule(&DBImpl::BGWorkCompaction, ca, thread_pool_pri, this,
                     &DBImpl::UnscheduleCompactionCallback);
      scheduled = true;
    }
@@ -1844,7 +1906,7 @@ Status DBImpl::WaitForFlushMemTables(
      }
    }
    if (1 == num_dropped && 1 == num) {
-      return Status::InvalidArgument("Cannot flush a dropped CF");
+      return Status::ColumnFamilyDropped();
    }
    // Column families involved in this flush request have either been dropped
    // or finished flush. Then it's time to finish waiting.
@@ -1968,7 +2030,7 @@ void DBImpl::MaybeScheduleFlushOrCompaction() {

 DBImpl::BGJobLimits DBImpl::GetBGJobLimits() const {
  mutex_.AssertHeld();
-  return GetBGJobLimits(immutable_db_options_.max_background_flushes,
+  return GetBGJobLimits(mutable_db_options_.max_background_flushes,
                        mutable_db_options_.max_background_compactions,
                        mutable_db_options_.max_background_jobs,
                        write_controller_.NeedSpeedupCompaction());
@@ -2081,8 +2143,7 @@ void DBImpl::BGWorkFlush(void* arg) {

  IOSTATS_SET_THREAD_POOL_ID(fta.thread_pri_);
  TEST_SYNC_POINT("DBImpl::BGWorkFlush");
-  static_cast_with_check<DBImpl, DB>(fta.db_)->BackgroundCallFlush(
-      fta.thread_pri_);
+  static_cast_with_check<DBImpl>(fta.db_)->BackgroundCallFlush(fta.thread_pri_);
  TEST_SYNC_POINT("DBImpl::BGWorkFlush:done");
 }

@@ -2093,7 +2154,7 @@ void DBImpl::BGWorkCompaction(void* arg) {
  TEST_SYNC_POINT("DBImpl::BGWorkCompaction");
  auto prepicked_compaction =
      static_cast<PrepickedCompaction*>(ca.prepicked_compaction);
-  static_cast_with_check<DBImpl, DB>(ca.db)->BackgroundCallCompaction(
+  static_cast_with_check<DBImpl>(ca.db)->BackgroundCallCompaction(
      prepicked_compaction, Env::Priority::LOW);
  delete prepicked_compaction;
 }
@@ -2487,8 +2548,8 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
                       "[%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().c_str() : "(begin)"),
-                       (m->end ? m->end->DebugString().c_str() : "(end)"));
+                       (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(
@@ -2507,11 +2568,11 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
            "[%s] Manual compaction from level-%d to level-%d from %s .. "
            "%s; will stop at %s\n",
            m->cfd->GetName().c_str(), m->input_level, c->output_level(),
-            (m->begin ? m->begin->DebugString().c_str() : "(begin)"),
-            (m->end ? m->end->DebugString().c_str() : "(end)"),
+            (m->begin ? m->begin->DebugString(true).c_str() : "(begin)"),
+            (m->end ? m->end->DebugString(true).c_str() : "(end)"),
            ((m->done || m->manual_end == nullptr)
                 ? "(end)"
-                 : m->manual_end->DebugString().c_str()));
+                 : m->manual_end->DebugString(true).c_str()));
      }
    }
  } else if (!is_prepicked && !compaction_queue_.empty()) {
@@ -2604,6 +2665,7 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
    }
  }

+  IOStatus io_s;
  if (!c) {
    // Nothing to do
    ROCKS_LOG_BUFFER(log_buffer, "Compaction nothing to do");
@@ -2625,9 +2687,11 @@ 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());
+    io_s = versions_->io_status();
    InstallSuperVersionAndScheduleWork(c->column_family_data(),
                                       &job_context->superversion_contexts[0],
                                       *c->mutable_cf_options());
@@ -2668,7 +2732,8 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
                           f->largest, f->fd.smallest_seqno,
                           f->fd.largest_seqno, f->marked_for_compaction,
                           f->oldest_blob_file_number, f->oldest_ancester_time,
-                           f->file_creation_time);
+                           f->file_creation_time, f->file_checksum,
+                           f->file_checksum_func_name);

        ROCKS_LOG_BUFFER(
            log_buffer,
@@ -2680,9 +2745,11 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
      }
    }

+    versions_->SetIOStatusOK();
    status = versions_->LogAndApply(c->column_family_data(),
                                    *c->mutable_cf_options(), c->edit(),
                                    &mutex_, directories_.GetDbDir());
+    io_s = versions_->io_status();
    // Use latest MutableCFOptions
    InstallSuperVersionAndScheduleWork(c->column_family_data(),
                                       &job_context->superversion_contexts[0],
@@ -2769,6 +2836,7 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
    mutex_.Lock();

    status = compaction_job.Install(*c->mutable_cf_options());
+    io_s = compaction_job.io_status();
    if (status.ok()) {
      InstallSuperVersionAndScheduleWork(c->column_family_data(),
                                         &job_context->superversion_contexts[0],
@@ -2778,6 +2846,11 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
    TEST_SYNC_POINT_CALLBACK("DBImpl::BackgroundCompaction:AfterCompaction",
                             c->column_family_data());
  }
+
+  if (status.ok() && !io_s.ok()) {
+    status = io_s;
+  }
+
  if (c != nullptr) {
    c->ReleaseCompactionFiles(status);
    *made_progress = true;
@@ -2803,7 +2876,11 @@ Status DBImpl::BackgroundCompaction(bool* made_progress,
  } else {
    ROCKS_LOG_WARN(immutable_db_options_.info_log, "Compaction error: %s",
                   status.ToString().c_str());
-    error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
+    if (!io_s.ok()) {
+      error_handler_.SetBGError(io_s, BackgroundErrorReason::kCompaction);
+    } else {
+      error_handler_.SetBGError(status, BackgroundErrorReason::kCompaction);
+    }
    if (c != nullptr && !is_manual && !error_handler_.IsBGWorkStopped()) {
      // Put this cfd back in the compaction queue so we can retry after some
      // time
@@ -3110,4 +3187,4 @@ void DBImpl::GetSnapshotContext(
  }
  *snapshot_seqs = snapshots_.GetAll(earliest_write_conflict_snapshot);
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -15,7 +15,7 @@
 #include "monitoring/thread_status_updater.h"
 #include "util/cast_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 uint64_t DBImpl::TEST_GetLevel0TotalSize() {
  InstrumentedMutexLock l(&mutex_);
  return default_cf_handle_->cfd()->current()->storage_info()->NumLevelBytes(0);
@@ -24,7 +24,9 @@ uint64_t DBImpl::TEST_GetLevel0TotalSize() {
 void DBImpl::TEST_SwitchWAL() {
  WriteContext write_context;
  InstrumentedMutexLock l(&mutex_);
+  void* writer = TEST_BeginWrite();
  SwitchWAL(&write_context);
+  TEST_EndWrite(writer);
 }

 bool DBImpl::TEST_WALBufferIsEmpty(bool lock) {
@@ -106,15 +108,18 @@ Status DBImpl::TEST_SwitchMemtable(ColumnFamilyData* cfd) {
    cfd = default_cf_handle_->cfd();
  }

+  Status s;
+  void* writer = TEST_BeginWrite();
  if (two_write_queues_) {
    WriteThread::Writer nonmem_w;
    nonmem_write_thread_.EnterUnbatched(&nonmem_w, &mutex_);
-    Status s = SwitchMemtable(cfd, &write_context);
+    s = SwitchMemtable(cfd, &write_context);
    nonmem_write_thread_.ExitUnbatched(&nonmem_w);
-    return s;
  } else {
-    return SwitchMemtable(cfd, &write_context);
+    s = SwitchMemtable(cfd, &write_context);
  }
+  TEST_EndWrite(writer);
+  return s;
 }

 Status DBImpl::TEST_FlushMemTable(bool wait, bool allow_write_stall,
@@ -285,5 +290,5 @@ bool DBImpl::TEST_IsPersistentStatsEnabled() const {
 size_t DBImpl::TEST_EstimateInMemoryStatsHistorySize() const {
  return EstimateInMemoryStatsHistorySize();
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
 #endif  // NDEBUG
@@ -17,7 +17,7 @@
 #include "db/version_set.h"
 #include "rocksdb/status.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 #ifndef ROCKSDB_LITE
 Status DBImpl::SuggestCompactRange(ColumnFamilyHandle* column_family,
@@ -129,7 +129,8 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
                   f->fd.GetFileSize(), f->smallest, f->largest,
                   f->fd.smallest_seqno, f->fd.largest_seqno,
                   f->marked_for_compaction, f->oldest_blob_file_number,
-                   f->oldest_ancester_time, f->file_creation_time);
+                   f->oldest_ancester_time, f->file_creation_time,
+                   f->file_checksum, f->file_checksum_func_name);
    }

    status = versions_->LogAndApply(cfd, *cfd->GetLatestMutableCFOptions(),
@@ -147,4 +148,4 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
 }
 #endif  // ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -14,10 +14,11 @@
 #include "db/event_helpers.h"
 #include "db/memtable_list.h"
 #include "file/file_util.h"
+#include "file/filename.h"
 #include "file/sst_file_manager_impl.h"
 #include "util/autovector.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 uint64_t DBImpl::MinLogNumberToKeep() {
  if (allow_2pc()) {
@@ -35,7 +36,7 @@ uint64_t DBImpl::MinObsoleteSstNumberToKeep() {
  return std::numeric_limits<uint64_t>::max();
 }

-// * Returns the list of live files in 'sst_live'
+// * 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
 // 'full_scan_candidate_files'.
@@ -76,26 +77,26 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
  // Since job_context->min_pending_output is set, until file scan finishes,
  // mutex_ cannot be released. Otherwise, we might see no min_pending_output
  // here but later find newer generated unfinalized files while scanning.
-  if (!pending_outputs_.empty()) {
-    job_context->min_pending_output = *pending_outputs_.begin();
-  } else {
-    // delete all of them
-    job_context->min_pending_output = std::numeric_limits<uint64_t>::max();
-  }
+  job_context->min_pending_output = MinObsoleteSstNumberToKeep();

  // Get obsolete files.  This function will also update the list of
  // pending files in VersionSet().
-  versions_->GetObsoleteFiles(&job_context->sst_delete_files,
-                              &job_context->manifest_delete_files,
-                              job_context->min_pending_output);
+  versions_->GetObsoleteFiles(
+      &job_context->sst_delete_files, &job_context->blob_delete_files,
+      &job_context->manifest_delete_files, job_context->min_pending_output);

-  // Mark the elements in job_context->sst_delete_files as grabbedForPurge
-  // so that other threads calling FindObsoleteFiles with full_scan=true
-  // will not add these files to candidate list for purge.
+  // Mark the elements in job_context->sst_delete_files and
+  // job_context->blob_delete_files as "grabbed for purge" so that other threads
+  // calling FindObsoleteFiles with full_scan=true will not add these files to
+  // candidate list for purge.
  for (const auto& sst_to_del : job_context->sst_delete_files) {
    MarkAsGrabbedForPurge(sst_to_del.metadata->fd.GetNumber());
  }

+  for (const auto& blob_file : job_context->blob_delete_files) {
+    MarkAsGrabbedForPurge(blob_file.GetBlobFileNumber());
+  }
+
  // store the current filenum, lognum, etc
  job_context->manifest_file_number = versions_->manifest_file_number();
  job_context->pending_manifest_file_number =
@@ -103,7 +104,7 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
  job_context->log_number = MinLogNumberToKeep();
  job_context->prev_log_number = versions_->prev_log_number();

-  versions_->AddLiveFiles(&job_context->sst_live);
+  versions_->AddLiveFiles(&job_context->sst_live, &job_context->blob_live);
  if (doing_the_full_scan) {
    InfoLogPrefix info_log_prefix(!immutable_db_options_.db_log_dir.empty(),
                                  dbname_);
@@ -256,8 +257,11 @@ bool CompareCandidateFile(const JobContext::CandidateFileInfo& first,
 void DBImpl::DeleteObsoleteFileImpl(int job_id, const std::string& fname,
                                    const std::string& path_to_sync,
                                    FileType type, uint64_t number) {
+  TEST_SYNC_POINT_CALLBACK("DBImpl::DeleteObsoleteFileImpl::BeforeDeletion",
+                           const_cast<std::string*>(&fname));
+
  Status file_deletion_status;
-  if (type == kTableFile || type == kLogFile) {
+  if (type == kTableFile || type == kBlobFile || type == kLogFile) {
    file_deletion_status =
        DeleteDBFile(&immutable_db_options_, fname, path_to_sync,
                     /*force_bg=*/false, /*force_fg=*/!wal_in_db_path_);
@@ -303,19 +307,19 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
  // FindObsoleteFiles() should've populated this so nonzero
  assert(state.manifest_file_number != 0);

-  // Now, convert live list to an unordered map, WITHOUT mutex held;
-  // set is slow.
-  std::unordered_map<uint64_t, const FileDescriptor*> sst_live_map;
-  for (const FileDescriptor& fd : state.sst_live) {
-    sst_live_map[fd.GetNumber()] = &fd;
-  }
+  // Now, convert lists to unordered sets, WITHOUT mutex held; set is slow.
+  std::unordered_set<uint64_t> sst_live_set(state.sst_live.begin(),
+                                            state.sst_live.end());
+  std::unordered_set<uint64_t> blob_live_set(state.blob_live.begin(),
+                                             state.blob_live.end());
  std::unordered_set<uint64_t> log_recycle_files_set(
      state.log_recycle_files.begin(), state.log_recycle_files.end());

  auto candidate_files = state.full_scan_candidate_files;
  candidate_files.reserve(
      candidate_files.size() + state.sst_delete_files.size() +
-      state.log_delete_files.size() + state.manifest_delete_files.size());
+      state.blob_delete_files.size() + state.log_delete_files.size() +
+      state.manifest_delete_files.size());
  // We may ignore the dbname when generating the file names.
  for (auto& file : state.sst_delete_files) {
    candidate_files.emplace_back(
@@ -326,6 +330,11 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
    file.DeleteMetadata();
  }

+  for (const auto& blob_file : state.blob_delete_files) {
+    candidate_files.emplace_back(BlobFileName(blob_file.GetBlobFileNumber()),
+                                 blob_file.GetPath());
+  }
+
  for (auto file_num : state.log_delete_files) {
    if (file_num > 0) {
      candidate_files.emplace_back(LogFileName(file_num),
@@ -412,12 +421,19 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
      case kTableFile:
        // If the second condition is not there, this makes
        // DontDeletePendingOutputs fail
-        keep = (sst_live_map.find(number) != sst_live_map.end()) ||
+        keep = (sst_live_set.find(number) != sst_live_set.end()) ||
               number >= state.min_pending_output;
        if (!keep) {
          files_to_del.insert(number);
        }
        break;
+      case kBlobFile:
+        keep = number >= state.min_pending_output ||
+               (blob_live_set.find(number) != blob_live_set.end());
+        if (!keep) {
+          files_to_del.insert(number);
+        }
+        break;
      case kTempFile:
        // Any temp files that are currently being written to must
        // be recorded in pending_outputs_, which is inserted into "live".
@@ -427,7 +443,8 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
        //
        // TODO(yhchiang): carefully modify the third condition to safely
        //                 remove the temp options files.
-        keep = (sst_live_map.find(number) != sst_live_map.end()) ||
+        keep = (sst_live_set.find(number) != sst_live_set.end()) ||
+               (blob_live_set.find(number) != blob_live_set.end()) ||
               (number == state.pending_manifest_file_number) ||
               (to_delete.find(kOptionsFileNamePrefix) != std::string::npos);
        break;
@@ -450,7 +467,6 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
      case kDBLockFile:
      case kIdentityFile:
      case kMetaDatabase:
-      case kBlobFile:
        keep = true;
        break;
    }
@@ -466,6 +482,9 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
      TableCache::Evict(table_cache_.get(), number);
      fname = MakeTableFileName(candidate_file.file_path, number);
      dir_to_sync = candidate_file.file_path;
+    } else if (type == kBlobFile) {
+      fname = BlobFileName(candidate_file.file_path, number);
+      dir_to_sync = candidate_file.file_path;
    } else {
      dir_to_sync =
          (type == kLogFile) ? immutable_db_options_.wal_dir : dbname_;
@@ -618,9 +637,9 @@ uint64_t PrecomputeMinLogNumberToKeep(
  // family being flushed (`cfd_to_flush`).
  uint64_t cf_min_log_number_to_keep = 0;
  for (auto& e : edit_list) {
-    if (e->has_log_number()) {
+    if (e->HasLogNumber()) {
      cf_min_log_number_to_keep =
-          std::max(cf_min_log_number_to_keep, e->log_number());
+          std::max(cf_min_log_number_to_keep, e->GetLogNumber());
    }
  }
  if (cf_min_log_number_to_keep == 0) {
@@ -664,4 +683,69 @@ uint64_t PrecomputeMinLogNumberToKeep(
  return min_log_number_to_keep;
 }

-}  // namespace rocksdb
+Status DBImpl::FinishBestEffortsRecovery() {
+  mutex_.AssertHeld();
+  std::vector<std::string> paths;
+  paths.push_back(dbname_);
+  for (const auto& db_path : immutable_db_options_.db_paths) {
+    paths.push_back(db_path.path);
+  }
+  for (const auto* cfd : *versions_->GetColumnFamilySet()) {
+    for (const auto& cf_path : cfd->ioptions()->cf_paths) {
+      paths.push_back(cf_path.path);
+    }
+  }
+  // Dedup paths
+  std::sort(paths.begin(), paths.end());
+  paths.erase(std::unique(paths.begin(), paths.end()), paths.end());
+
+  uint64_t next_file_number = versions_->current_next_file_number();
+  uint64_t largest_file_number = next_file_number;
+  std::set<std::string> files_to_delete;
+  for (const auto& path : paths) {
+    std::vector<std::string> files;
+    env_->GetChildren(path, &files);
+    for (const auto& fname : files) {
+      uint64_t number = 0;
+      FileType type;
+      if (!ParseFileName(fname, &number, &type)) {
+        continue;
+      }
+      const std::string normalized_fpath = NormalizePath(path + fname);
+      largest_file_number = std::max(largest_file_number, number);
+      if (type == kTableFile && number >= next_file_number &&
+          files_to_delete.find(normalized_fpath) == files_to_delete.end()) {
+        files_to_delete.insert(normalized_fpath);
+      }
+    }
+  }
+  if (largest_file_number > next_file_number) {
+    versions_->next_file_number_.store(largest_file_number + 1);
+  }
+
+  VersionEdit edit;
+  edit.SetNextFile(versions_->next_file_number_.load());
+  assert(versions_->GetColumnFamilySet());
+  ColumnFamilyData* default_cfd = versions_->GetColumnFamilySet()->GetDefault();
+  assert(default_cfd);
+  // Even if new_descriptor_log is false, we will still switch to a new
+  // MANIFEST and update CURRENT file, since this is in recovery.
+  Status s = versions_->LogAndApply(
+      default_cfd, *default_cfd->GetLatestMutableCFOptions(), &edit, &mutex_,
+      directories_.GetDbDir(), /*new_descriptor_log*/ false);
+  if (!s.ok()) {
+    return s;
+  }
+
+  mutex_.Unlock();
+  for (const auto& fname : files_to_delete) {
+    s = env_->DeleteFile(fname);
+    if (!s.ok()) {
+      break;
+    }
+  }
+  mutex_.Lock();
+  return s;
+}
+
+}  // namespace ROCKSDB_NAMESPACE
@@ -23,7 +23,7 @@
 #include "test_util/sync_point.h"
 #include "util/rate_limiter.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 Options SanitizeOptions(const std::string& dbname, const Options& src) {
  auto db_options = SanitizeOptions(dbname, DBOptions(src));
  ImmutableDBOptions immutable_db_options(db_options);
@@ -35,16 +35,8 @@ Options SanitizeOptions(const std::string& dbname, const Options& src) {
 DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
  DBOptions result(src);

-  if (result.file_system == nullptr) {
-    if (result.env == Env::Default()) {
-      result.file_system = FileSystem::Default();
-    } else {
-      result.file_system.reset(new LegacyFileSystemWrapper(result.env));
-    }
-  } else {
-    if (result.env == nullptr) {
-      result.env = Env::Default();
-    }
+  if (result.env == nullptr) {
+    result.env = Env::Default();
  }

  // result.max_open_files means an "infinite" open files.
@@ -100,12 +92,11 @@ 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::kAbsoluteConsistency)) {
-    // kPointInTimeRecovery is indistinguishable from
-    // kTolerateCorruptedTailRecords in recycle mode since we define
-    // the "end" of the log as the first corrupt record we encounter.
+    // 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.
-    result.wal_recovery_mode = WALRecoveryMode::kTolerateCorruptedTailRecords;
+    result.recycle_log_file_num = 0;
  }

  if (result.wal_dir.empty()) {
@@ -173,6 +164,13 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
    result.sst_file_manager = sst_file_manager;
  }
 #endif
+
+  if (!result.paranoid_checks) {
+    result.skip_checking_sst_file_sizes_on_db_open = true;
+    ROCKS_LOG_INFO(result.info_log,
+                   "file size check will be skipped during open.");
+  }
+
  return result;
 }

@@ -246,6 +244,12 @@ Status DBImpl::ValidateOptions(const DBOptions& db_options) {
        "atomic_flush is incompatible with enable_pipelined_write");
  }

+  // TODO remove this restriction
+  if (db_options.atomic_flush && db_options.best_efforts_recovery) {
+    return Status::InvalidArgument(
+        "atomic_flush is currently incompatible with best-efforts recovery");
+  }
+
  return Status::OK();
 }

@@ -288,15 +292,16 @@ Status DBImpl::NewDB() {
  }
  if (s.ok()) {
    // Make "CURRENT" file that points to the new manifest file.
-    s = SetCurrentFile(env_, dbname_, 1, directories_.GetDbDir());
+    s = SetCurrentFile(fs_.get(), dbname_, 1, directories_.GetDbDir());
  } else {
    fs_->DeleteFile(manifest, IOOptions(), nullptr);
  }
  return s;
 }

-Status DBImpl::CreateAndNewDirectory(Env* env, const std::string& dirname,
-                                     std::unique_ptr<Directory>* directory) {
+IOStatus DBImpl::CreateAndNewDirectory(
+    FileSystem* fs, const std::string& dirname,
+    std::unique_ptr<FSDirectory>* directory) {
  // We call CreateDirIfMissing() as the directory may already exist (if we
  // are reopening a DB), when this happens we don't want creating the
  // directory to cause an error. However, we need to check if creating the
@@ -304,24 +309,24 @@ Status DBImpl::CreateAndNewDirectory(Env* env, const std::string& dirname,
  // file not existing. One real-world example of this occurring is if
  // env->CreateDirIfMissing() doesn't create intermediate directories, e.g.
  // when dbname_ is "dir/db" but when "dir" doesn't exist.
-  Status s = env->CreateDirIfMissing(dirname);
-  if (!s.ok()) {
-    return s;
+  IOStatus io_s = fs->CreateDirIfMissing(dirname, IOOptions(), nullptr);
+  if (!io_s.ok()) {
+    return io_s;
  }
-  return env->NewDirectory(dirname, directory);
+  return fs->NewDirectory(dirname, IOOptions(), directory, nullptr);
 }

-Status Directories::SetDirectories(Env* env, const std::string& dbname,
-                                   const std::string& wal_dir,
-                                   const std::vector<DbPath>& data_paths) {
-  Status s = DBImpl::CreateAndNewDirectory(env, dbname, &db_dir_);
-  if (!s.ok()) {
-    return s;
+IOStatus Directories::SetDirectories(FileSystem* fs, const std::string& dbname,
+                                     const std::string& wal_dir,
+                                     const std::vector<DbPath>& data_paths) {
+  IOStatus io_s = DBImpl::CreateAndNewDirectory(fs, dbname, &db_dir_);
+  if (!io_s.ok()) {
+    return io_s;
  }
  if (!wal_dir.empty() && dbname != wal_dir) {
-    s = DBImpl::CreateAndNewDirectory(env, wal_dir, &wal_dir_);
-    if (!s.ok()) {
-      return s;
+    io_s = DBImpl::CreateAndNewDirectory(fs, wal_dir, &wal_dir_);
+    if (!io_s.ok()) {
+      return io_s;
    }
  }

@@ -331,27 +336,28 @@ Status Directories::SetDirectories(Env* env, const std::string& dbname,
    if (db_path == dbname) {
      data_dirs_.emplace_back(nullptr);
    } else {
-      std::unique_ptr<Directory> path_directory;
-      s = DBImpl::CreateAndNewDirectory(env, db_path, &path_directory);
-      if (!s.ok()) {
-        return s;
+      std::unique_ptr<FSDirectory> path_directory;
+      io_s = DBImpl::CreateAndNewDirectory(fs, db_path, &path_directory);
+      if (!io_s.ok()) {
+        return io_s;
      }
      data_dirs_.emplace_back(path_directory.release());
    }
  }
  assert(data_dirs_.size() == data_paths.size());
-  return Status::OK();
+  return IOStatus::OK();
 }

 Status DBImpl::Recover(
    const std::vector<ColumnFamilyDescriptor>& column_families, bool read_only,
-    bool error_if_log_file_exist, bool error_if_data_exists_in_logs) {
+    bool error_if_log_file_exist, bool error_if_data_exists_in_logs,
+    uint64_t* recovered_seq) {
  mutex_.AssertHeld();

  bool is_new_db = false;
  assert(db_lock_ == nullptr);
  if (!read_only) {
-    Status s = directories_.SetDirectories(env_, dbname_,
+    Status s = directories_.SetDirectories(fs_.get(), dbname_,
                                           immutable_db_options_.wal_dir,
                                           immutable_db_options_.db_paths);
    if (!s.ok()) {
@@ -367,9 +373,6 @@ Status DBImpl::Recover(
    s = env_->FileExists(current_fname);
    if (s.IsNotFound()) {
      if (immutable_db_options_.create_if_missing) {
-        // Has to be called only after Identity File creation is successful
-        // because DB ID is stored in Manifest if
-        // immutable_db_options_.write_dbid_to_manifest = true
        s = NewDB();
        is_new_db = true;
        if (!s.ok()) {
@@ -414,7 +417,17 @@ Status DBImpl::Recover(
    }
  }
  assert(db_id_.empty());
-  Status s = versions_->Recover(column_families, read_only, &db_id_);
+  Status s;
+  bool missing_table_file = false;
+  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);
+    if (s.ok()) {
+      s = FinishBestEffortsRecovery();
+    }
+  }
  if (!s.ok()) {
    return s;
  }
@@ -435,27 +448,28 @@ Status DBImpl::Recover(
      assert(s.IsIOError());
      return s;
    }
-    GetDbIdentityFromIdentityFile(&db_id_);
-    if (immutable_db_options_.write_dbid_to_manifest) {
+    s = GetDbIdentityFromIdentityFile(&db_id_);
+    if (immutable_db_options_.write_dbid_to_manifest && s.ok()) {
      VersionEdit edit;
      edit.SetDBId(db_id_);
      Options options;
      MutableCFOptions mutable_cf_options(options);
      versions_->db_id_ = db_id_;
-      versions_->LogAndApply(versions_->GetColumnFamilySet()->GetDefault(),
+      s = versions_->LogAndApply(versions_->GetColumnFamilySet()->GetDefault(),
                             mutable_cf_options, &edit, &mutex_, nullptr,
                             false);
    }
  } else {
-    SetIdentityFile(env_, dbname_, db_id_);
+    s = SetIdentityFile(env_, dbname_, db_id_);
  }

  if (immutable_db_options_.paranoid_checks && s.ok()) {
    s = CheckConsistency();
  }
  if (s.ok() && !read_only) {
+    std::map<std::string, std::shared_ptr<FSDirectory>> created_dirs;
    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      s = cfd->AddDirectories();
+      s = cfd->AddDirectories(&created_dirs);
      if (!s.ok()) {
        return s;
      }
@@ -466,16 +480,16 @@ Status DBImpl::Recover(
    s = InitPersistStatsColumnFamily();
  }

-  // Initial max_total_in_memory_state_ before recovery logs. Log recovery
-  // may check this value to decide whether to flush.
-  max_total_in_memory_state_ = 0;
-  for (auto cfd : *versions_->GetColumnFamilySet()) {
-    auto* mutable_cf_options = cfd->GetLatestMutableCFOptions();
-    max_total_in_memory_state_ += mutable_cf_options->write_buffer_size *
-                                  mutable_cf_options->max_write_buffer_number;
-  }
-
  if (s.ok()) {
+    // Initial max_total_in_memory_state_ before recovery logs. Log recovery
+    // may check this value to decide whether to flush.
+    max_total_in_memory_state_ = 0;
+    for (auto cfd : *versions_->GetColumnFamilySet()) {
+      auto* mutable_cf_options = cfd->GetLatestMutableCFOptions();
+      max_total_in_memory_state_ += mutable_cf_options->write_buffer_size *
+                                    mutable_cf_options->max_write_buffer_number;
+    }
+
    SequenceNumber next_sequence(kMaxSequenceNumber);
    default_cf_handle_ = new ColumnFamilyHandleImpl(
        versions_->GetColumnFamilySet()->GetDefault(), this, &mutex_);
@@ -493,7 +507,9 @@ Status DBImpl::Recover(
    // attention to it in case we are recovering a database
    // produced by an older version of rocksdb.
    std::vector<std::string> filenames;
-    s = env_->GetChildren(immutable_db_options_.wal_dir, &filenames);
+    if (!immutable_db_options_.best_efforts_recovery) {
+      s = env_->GetChildren(immutable_db_options_.wal_dir, &filenames);
+    }
    if (s.IsNotFound()) {
      return Status::InvalidArgument("wal_dir not found",
                                     immutable_db_options_.wal_dir);
@@ -541,7 +557,12 @@ Status DBImpl::Recover(
    if (!logs.empty()) {
      // Recover in the order in which the logs were generated
      std::sort(logs.begin(), logs.end());
-      s = RecoverLogFiles(logs, &next_sequence, read_only);
+      bool corrupted_log_found = false;
+      s = RecoverLogFiles(logs, &next_sequence, read_only,
+                          &corrupted_log_found);
+      if (corrupted_log_found && recovered_seq != nullptr) {
+        *recovered_seq = next_sequence;
+      }
      if (!s.ok()) {
        // Clear memtables if recovery failed
        for (auto cfd : *versions_->GetColumnFamilySet()) {
@@ -671,7 +692,8 @@ Status DBImpl::InitPersistStatsColumnFamily() {

 // REQUIRES: log_numbers are sorted in ascending order
 Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
-                               SequenceNumber* next_sequence, bool read_only) {
+                               SequenceNumber* next_sequence, bool read_only,
+                               bool* corrupted_log_found) {
  struct LogReporter : public log::Reader::Reporter {
    Env* env;
    Logger* info_log;
@@ -973,6 +995,9 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
        status = Status::OK();
        stop_replay_for_corruption = true;
        corrupted_log_number = log_number;
+        if (corrupted_log_found != nullptr) {
+          *corrupted_log_found = true;
+        }
        ROCKS_LOG_INFO(immutable_db_options_.info_log,
                       "Point in time recovered to log #%" PRIu64
                       " seq #%" PRIu64,
@@ -1001,6 +1026,9 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
  // the corrupted log number, which means CF contains data beyond the point of
  // corruption. This could during PIT recovery when the WAL is corrupted and
  // some (but not all) CFs are flushed
+  // Exclude the PIT case where no log is dropped after the corruption point.
+  // This is to cover the case for empty logs after corrupted log, in which we
+  // don't reset stop_replay_for_corruption.
  if (stop_replay_for_corruption == true &&
      (immutable_db_options_.wal_recovery_mode ==
           WALRecoveryMode::kPointInTimeRecovery ||
@@ -1211,6 +1239,7 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
      if (range_del_iter != nullptr) {
        range_del_iters.emplace_back(range_del_iter);
      }
+      IOStatus io_s;
      s = BuildTable(
          dbname_, env_, fs_.get(), *cfd->ioptions(), mutable_cf_options,
          file_options_for_compaction_, cfd->table_cache(), iter.get(),
@@ -1219,8 +1248,8 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
          snapshot_seqs, earliest_write_conflict_snapshot, snapshot_checker,
          GetCompressionFlush(*cfd->ioptions(), mutable_cf_options),
          mutable_cf_options.sample_for_compression,
-          cfd->ioptions()->compression_opts, paranoid_file_checks,
-          cfd->internal_stats(), TableFileCreationReason::kRecovery,
+          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);
      LogFlush(immutable_db_options_.info_log);
@@ -1242,7 +1271,8 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
                  meta.fd.GetFileSize(), meta.smallest, meta.largest,
                  meta.fd.smallest_seqno, meta.fd.largest_seqno,
                  meta.marked_for_compaction, meta.oldest_blob_file_number,
-                  meta.oldest_ancester_time, meta.file_creation_time);
+                  meta.oldest_ancester_time, meta.file_creation_time,
+                  meta.file_checksum, meta.file_checksum_func_name);
  }

  InternalStats::CompactionStats stats(CompactionReason::kFlush, 1);
@@ -1311,6 +1341,8 @@ Status DBImpl::CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
                   recycle_log_number);
    std::string old_log_fname =
        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);
  } else {
@@ -1380,13 +1412,9 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
      impl->error_handler_.EnableAutoRecovery();
    }
  }
-
-  if (!s.ok()) {
-    delete impl;
-    return s;
+  if (s.ok()) {
+    s = impl->CreateArchivalDirectory();
  }
-
-  s = impl->CreateArchivalDirectory();
  if (!s.ok()) {
    delete impl;
    return s;
@@ -1396,7 +1424,8 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,

  impl->mutex_.Lock();
  // Handles create_if_missing, error_if_exists
-  s = impl->Recover(column_families);
+  uint64_t recovered_seq(kMaxSequenceNumber);
+  s = impl->Recover(column_families, false, false, false, &recovered_seq);
  if (s.ok()) {
    uint64_t new_log_number = impl->versions_->NewFileNumber();
    log::Writer* new_log = nullptr;
@@ -1454,8 +1483,35 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
      if (impl->two_write_queues_) {
        impl->log_write_mutex_.Unlock();
      }
+
      impl->DeleteObsoleteFiles();
-      s = impl->directories_.GetDbDir()->Fsync();
+      s = impl->directories_.GetDbDir()->Fsync(IOOptions(), nullptr);
+    }
+    if (s.ok()) {
+      // In WritePrepared there could be gap in sequence numbers. This breaks
+      // the trick we use in kPointInTimeRecovery which assumes the first seq in
+      // the log right after the corrupted log is one larger than the last seq
+      // we read from the logs. To let this trick keep working, we add a dummy
+      // entry with the expected sequence to the first log right after recovery.
+      // In non-WritePrepared case also the new log after recovery could be
+      // empty, and thus missing the consecutive seq hint to distinguish
+      // middle-log corruption to corrupted-log-remained-after-recovery. This
+      // case also will be addressed by a dummy write.
+      if (recovered_seq != kMaxSequenceNumber) {
+        WriteBatch empty_batch;
+        WriteBatchInternal::SetSequence(&empty_batch, recovered_seq);
+        WriteOptions write_options;
+        uint64_t log_used, log_size;
+        log::Writer* log_writer = impl->logs_.back().writer;
+        s = impl->WriteToWAL(empty_batch, log_writer, &log_used, &log_size);
+        if (s.ok()) {
+          // Need to fsync, otherwise it might get lost after a power reset.
+          s = impl->FlushWAL(false);
+          if (s.ok()) {
+            s = log_writer->file()->Sync(impl->immutable_db_options_.use_fsync);
+          }
+        }
+      }
    }
  }
  if (s.ok() && impl->immutable_db_options_.persist_stats_to_disk) {
@@ -1512,6 +1568,27 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
  if (s.ok() && sfm) {
    // Notify SstFileManager about all sst files that already exist in
    // db_paths[0] and cf_paths[0] when the DB is opened.
+
+    // SstFileManagerImpl needs to know sizes of the files. For files whose size
+    // we already know (sst files that appear in manifest - typically that's the
+    // vast majority of all files), we'll pass the size to SstFileManager.
+    // For all other files SstFileManager will query the size from filesystem.
+
+    std::vector<LiveFileMetaData> metadata;
+
+    impl->mutex_.Lock();
+    impl->versions_->GetLiveFilesMetaData(&metadata);
+    impl->mutex_.Unlock();
+
+    std::unordered_map<std::string, uint64_t> known_file_sizes;
+    for (const auto& md : metadata) {
+      std::string name = md.name;
+      if (!name.empty() && name[0] == '/') {
+        name = name.substr(1);
+      }
+      known_file_sizes[name] = md.size;
+    }
+
    std::vector<std::string> paths;
    paths.emplace_back(impl->immutable_db_options_.db_paths[0].path);
    for (auto& cf : column_families) {
@@ -1531,7 +1608,14 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
        std::string file_path = path + "/" + file_name;
        if (ParseFileName(file_name, &file_number, &file_type) &&
            file_type == kTableFile) {
-          sfm->OnAddFile(file_path);
+          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);
+          } else {
+            sfm->OnAddFile(file_path);
+          }
        }
      }
    }
@@ -1572,4 +1656,4 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
  }
  return s;
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -12,7 +12,7 @@
 #include "db/merge_context.h"
 #include "monitoring/perf_context_imp.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 #ifndef ROCKSDB_LITE

@@ -48,14 +48,16 @@ Status DBImplReadOnly::Get(const ReadOptions& read_options,
  SequenceNumber max_covering_tombstone_seq = 0;
  LookupKey lkey(key, snapshot);
  PERF_TIMER_STOP(get_snapshot_time);
-  if (super_version->mem->Get(lkey, pinnable_val->GetSelf(), &s, &merge_context,
+  if (super_version->mem->Get(lkey, pinnable_val->GetSelf(),
+                              /*timestamp=*/nullptr, &s, &merge_context,
                              &max_covering_tombstone_seq, read_options)) {
    pinnable_val->PinSelf();
    RecordTick(stats_, MEMTABLE_HIT);
  } else {
    PERF_TIMER_GUARD(get_from_output_files_time);
-    super_version->current->Get(read_options, lkey, pinnable_val, &s,
-                                &merge_context, &max_covering_tombstone_seq);
+    super_version->current->Get(read_options, lkey, pinnable_val,
+                                /*timestamp=*/nullptr, &s, &merge_context,
+                                &max_covering_tombstone_seq);
    RecordTick(stats_, MEMTABLE_MISS);
  }
  RecordTick(stats_, NUMBER_KEYS_READ);
@@ -85,7 +87,8 @@ Iterator* DBImplReadOnly::NewIterator(const ReadOptions& read_options,
      super_version->version_number, read_callback);
  auto internal_iter =
      NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), read_seq);
+                          db_iter->GetRangeDelAggregator(), read_seq,
+                          /* allow_unprepared_value */ true);
  db_iter->SetIterUnderDBIter(internal_iter);
  return db_iter;
 }
@@ -116,7 +119,8 @@ Status DBImplReadOnly::NewIterators(
        sv->version_number, read_callback);
    auto* internal_iter =
        NewInternalIterator(read_options, cfd, sv, db_iter->GetArena(),
-                            db_iter->GetRangeDelAggregator(), read_seq);
+                            db_iter->GetRangeDelAggregator(), read_seq,
+                            /* allow_unprepared_value */ true);
    db_iter->SetIterUnderDBIter(internal_iter);
    iterators->push_back(db_iter);
  }
@@ -218,4 +222,4 @@ Status DB::OpenForReadOnly(
 }
 #endif  // !ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -11,7 +11,7 @@
 #include <vector>
 #include "db/db_impl/db_impl.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBImplReadOnly : public DBImpl {
 public:
@@ -132,6 +132,6 @@ class DBImplReadOnly : public DBImpl {
 private:
  friend class DB;
 };
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // !ROCKSDB_LITE
@@ -13,7 +13,7 @@
 #include "monitoring/perf_context_imp.h"
 #include "util/cast_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 #ifndef ROCKSDB_LITE
 DBImplSecondary::DBImplSecondary(const DBOptions& db_options,
@@ -29,7 +29,7 @@ DBImplSecondary::~DBImplSecondary() {}
 Status DBImplSecondary::Recover(
    const std::vector<ColumnFamilyDescriptor>& column_families,
    bool /*readonly*/, bool /*error_if_log_file_exist*/,
-    bool /*error_if_data_exists_in_logs*/) {
+    bool /*error_if_data_exists_in_logs*/, uint64_t*) {
  mutex_.AssertHeld();

  JobContext job_context(0);
@@ -340,15 +340,16 @@ Status DBImplSecondary::GetImpl(const ReadOptions& read_options,
  PERF_TIMER_STOP(get_snapshot_time);

  bool done = false;
-  if (super_version->mem->Get(lkey, pinnable_val->GetSelf(), &s, &merge_context,
+  if (super_version->mem->Get(lkey, pinnable_val->GetSelf(),
+                              /*timestamp=*/nullptr, &s, &merge_context,
                              &max_covering_tombstone_seq, read_options)) {
    done = true;
    pinnable_val->PinSelf();
    RecordTick(stats_, MEMTABLE_HIT);
  } else if ((s.ok() || s.IsMergeInProgress()) &&
             super_version->imm->Get(
-                 lkey, pinnable_val->GetSelf(), &s, &merge_context,
-                 &max_covering_tombstone_seq, read_options)) {
+                 lkey, pinnable_val->GetSelf(), /*timestamp=*/nullptr, &s,
+                 &merge_context, &max_covering_tombstone_seq, read_options)) {
    done = true;
    pinnable_val->PinSelf();
    RecordTick(stats_, MEMTABLE_HIT);
@@ -359,8 +360,9 @@ Status DBImplSecondary::GetImpl(const ReadOptions& read_options,
  }
  if (!done) {
    PERF_TIMER_GUARD(get_from_output_files_time);
-    super_version->current->Get(read_options, lkey, pinnable_val, &s,
-                                &merge_context, &max_covering_tombstone_seq);
+    super_version->current->Get(read_options, lkey, pinnable_val,
+                                /*timestamp=*/nullptr, &s, &merge_context,
+                                &max_covering_tombstone_seq);
    RecordTick(stats_, MEMTABLE_MISS);
  }
  {
@@ -415,7 +417,8 @@ ArenaWrappedDBIter* DBImplSecondary::NewIteratorImpl(
      super_version->version_number, read_callback);
  auto internal_iter =
      NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), snapshot);
+                          db_iter->GetRangeDelAggregator(), snapshot,
+                          /* allow_unprepared_value */ true);
  db_iter->SetIterUnderDBIter(internal_iter);
  return db_iter;
 }
@@ -468,6 +471,11 @@ Status DBImplSecondary::CheckConsistency() {
  // approach and just proceed.
  TEST_SYNC_POINT_CALLBACK(
      "DBImplSecondary::CheckConsistency:AfterFirstAttempt", &s);
+
+  if (immutable_db_options_.skip_checking_sst_file_sizes_on_db_open) {
+    return Status::OK();
+  }
+
  std::vector<LiveFileMetaData> metadata;
  versions_->GetLiveFilesMetaData(&metadata);

@@ -663,4 +671,4 @@ Status DB::OpenAsSecondary(
 }
 #endif  // !ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -11,7 +11,7 @@
 #include <vector>
 #include "db/db_impl/db_impl.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // A wrapper class to hold log reader, log reporter, log status.
 class LogReaderContainer {
@@ -78,7 +78,8 @@ class DBImplSecondary : public DBImpl {
  // and log_readers_ to facilitate future operations.
  Status Recover(const std::vector<ColumnFamilyDescriptor>& column_families,
                 bool read_only, bool error_if_log_file_exist,
-                 bool error_if_data_exists_in_logs) override;
+                 bool error_if_data_exists_in_logs,
+                 uint64_t* = nullptr) override;

  // Implementations of the DB interface
  using DB::Get;
@@ -327,6 +328,6 @@ class DBImplSecondary : public DBImpl {
  std::unordered_map<ColumnFamilyData*, uint64_t> cfd_to_current_log_;
 };

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 #endif  // !ROCKSDB_LITE
@@ -15,7 +15,7 @@
 #include "options/options_helper.h"
 #include "test_util/sync_point.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 // Convenience methods
 Status DBImpl::Put(const WriteOptions& o, ColumnFamilyHandle* column_family,
                   const Slice& key, const Slice& val) {
@@ -101,6 +101,7 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
         disable_memtable);

  Status status;
+  IOStatus io_s;
  if (write_options.low_pri) {
    status = ThrottleLowPriWritesIfNeeded(write_options, my_batch);
    if (!status.ok()) {
@@ -322,21 +323,22 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
    if (!two_write_queues_) {
      if (status.ok() && !write_options.disableWAL) {
        PERF_TIMER_GUARD(write_wal_time);
-        status = WriteToWAL(write_group, log_writer, log_used, need_log_sync,
-                            need_log_dir_sync, last_sequence + 1);
+        io_s = WriteToWAL(write_group, log_writer, log_used, need_log_sync,
+                          need_log_dir_sync, last_sequence + 1);
      }
    } else {
      if (status.ok() && !write_options.disableWAL) {
        PERF_TIMER_GUARD(write_wal_time);
        // LastAllocatedSequence is increased inside WriteToWAL under
        // wal_write_mutex_ to ensure ordered events in WAL
-        status = ConcurrentWriteToWAL(write_group, log_used, &last_sequence,
-                                      seq_inc);
+        io_s = ConcurrentWriteToWAL(write_group, log_used, &last_sequence,
+                                    seq_inc);
      } else {
        // Otherwise we inc seq number for memtable writes
        last_sequence = versions_->FetchAddLastAllocatedSequence(seq_inc);
      }
    }
+    status = io_s;
    assert(last_sequence != kMaxSequenceNumber);
    const SequenceNumber current_sequence = last_sequence + 1;
    last_sequence += seq_inc;
@@ -411,7 +413,11 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
  PERF_TIMER_START(write_pre_and_post_process_time);

  if (!w.CallbackFailed()) {
-    WriteStatusCheck(status);
+    if (!io_s.ok()) {
+      IOStatusCheck(io_s);
+    } else {
+      WriteStatusCheck(status);
+    }
  }

  if (need_log_sync) {
@@ -515,6 +521,7 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,

    PERF_TIMER_STOP(write_pre_and_post_process_time);

+    IOStatus io_s;
    if (w.status.ok() && !write_options.disableWAL) {
      PERF_TIMER_GUARD(write_wal_time);
      stats->AddDBStats(InternalStats::kIntStatsWriteDoneBySelf, 1);
@@ -524,12 +531,17 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
                          wal_write_group.size - 1);
        RecordTick(stats_, WRITE_DONE_BY_OTHER, wal_write_group.size - 1);
      }
-      w.status = WriteToWAL(wal_write_group, log_writer, log_used,
-                            need_log_sync, need_log_dir_sync, current_sequence);
+      io_s = WriteToWAL(wal_write_group, log_writer, log_used, need_log_sync,
+                        need_log_dir_sync, current_sequence);
+      w.status = io_s;
    }

    if (!w.CallbackFailed()) {
-      WriteStatusCheck(w.status);
+      if (!io_s.ok()) {
+        IOStatusCheck(io_s);
+      } else {
+        WriteStatusCheck(w.status);
+      }
    }

    if (need_log_sync) {
@@ -740,9 +752,10 @@ Status DBImpl::WriteImplWALOnly(
    }
    seq_inc = total_batch_cnt;
  }
+  IOStatus io_s;
  if (!write_options.disableWAL) {
-    status =
-        ConcurrentWriteToWAL(write_group, log_used, &last_sequence, seq_inc);
+    io_s = ConcurrentWriteToWAL(write_group, log_used, &last_sequence, seq_inc);
+    status = io_s;
  } else {
    // Otherwise we inc seq number to do solely the seq allocation
    last_sequence = versions_->FetchAddLastAllocatedSequence(seq_inc);
@@ -777,7 +790,11 @@ Status DBImpl::WriteImplWALOnly(
  PERF_TIMER_START(write_pre_and_post_process_time);

  if (!w.CallbackFailed()) {
-    WriteStatusCheck(status);
+    if (!io_s.ok()) {
+      IOStatusCheck(io_s);
+    } else {
+      WriteStatusCheck(status);
+    }
  }
  if (status.ok()) {
    size_t index = 0;
@@ -823,6 +840,17 @@ 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()) {
+    mutex_.Lock();
+    error_handler_.SetBGError(io_status, BackgroundErrorReason::kWriteCallback);
+    mutex_.Unlock();
+  }
+}
+
 void DBImpl::MemTableInsertStatusCheck(const Status& status) {
  // A non-OK status here indicates that the state implied by the
  // WAL has diverged from the in-memory state.  This could be
@@ -961,9 +989,9 @@ WriteBatch* DBImpl::MergeBatch(const WriteThread::WriteGroup& write_group,

 // When two_write_queues_ is disabled, this function is called from the only
 // write thread. Otherwise this must be called holding log_write_mutex_.
-Status DBImpl::WriteToWAL(const WriteBatch& merged_batch,
-                          log::Writer* log_writer, uint64_t* log_used,
-                          uint64_t* log_size) {
+IOStatus DBImpl::WriteToWAL(const WriteBatch& merged_batch,
+                            log::Writer* log_writer, uint64_t* log_used,
+                            uint64_t* log_size) {
  assert(log_size != nullptr);
  Slice log_entry = WriteBatchInternal::Contents(&merged_batch);
  *log_size = log_entry.size();
@@ -978,7 +1006,8 @@ Status DBImpl::WriteToWAL(const WriteBatch& merged_batch,
  if (UNLIKELY(needs_locking)) {
    log_write_mutex_.Lock();
  }
-  Status status = log_writer->AddRecord(log_entry);
+  IOStatus io_s = log_writer->AddRecord(log_entry);
+
  if (UNLIKELY(needs_locking)) {
    log_write_mutex_.Unlock();
  }
@@ -990,15 +1019,14 @@ Status DBImpl::WriteToWAL(const WriteBatch& merged_batch,
  // since alive_log_files_ might be modified concurrently
  alive_log_files_.back().AddSize(log_entry.size());
  log_empty_ = false;
-  return status;
+  return io_s;
 }

-Status DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
-                          log::Writer* log_writer, uint64_t* log_used,
-                          bool need_log_sync, bool need_log_dir_sync,
-                          SequenceNumber sequence) {
-  Status status;
-
+IOStatus DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
+                            log::Writer* log_writer, uint64_t* log_used,
+                            bool need_log_sync, bool need_log_dir_sync,
+                            SequenceNumber sequence) {
+  IOStatus io_s;
  assert(!write_group.leader->disable_wal);
  // Same holds for all in the batch group
  size_t write_with_wal = 0;
@@ -1016,13 +1044,13 @@ Status DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
  WriteBatchInternal::SetSequence(merged_batch, sequence);

  uint64_t log_size;
-  status = WriteToWAL(*merged_batch, log_writer, log_used, &log_size);
+  io_s = WriteToWAL(*merged_batch, log_writer, log_used, &log_size);
  if (to_be_cached_state) {
    cached_recoverable_state_ = *to_be_cached_state;
    cached_recoverable_state_empty_ = false;
  }

-  if (status.ok() && need_log_sync) {
+  if (io_s.ok() && need_log_sync) {
    StopWatch sw(env_, stats_, WAL_FILE_SYNC_MICROS);
    // It's safe to access logs_ with unlocked mutex_ here because:
    //  - we've set getting_synced=true for all logs,
@@ -1032,23 +1060,24 @@ Status DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
    //  - as long as other threads don't modify it, it's safe to read
    //    from std::deque from multiple threads concurrently.
    for (auto& log : logs_) {
-      status = log.writer->file()->Sync(immutable_db_options_.use_fsync);
-      if (!status.ok()) {
+      io_s = log.writer->file()->Sync(immutable_db_options_.use_fsync);
+      if (!io_s.ok()) {
        break;
      }
    }
-    if (status.ok() && need_log_dir_sync) {
+
+    if (io_s.ok() && need_log_dir_sync) {
      // We only sync WAL directory the first time WAL syncing is
      // requested, so that in case users never turn on WAL sync,
      // we can avoid the disk I/O in the write code path.
-      status = directories_.GetWalDir()->Fsync();
+      io_s = directories_.GetWalDir()->Fsync(IOOptions(), nullptr);
    }
  }

  if (merged_batch == &tmp_batch_) {
    tmp_batch_.Clear();
  }
-  if (status.ok()) {
+  if (io_s.ok()) {
    auto stats = default_cf_internal_stats_;
    if (need_log_sync) {
      stats->AddDBStats(InternalStats::kIntStatsWalFileSynced, 1);
@@ -1059,14 +1088,13 @@ Status DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
    stats->AddDBStats(InternalStats::kIntStatsWriteWithWal, write_with_wal);
    RecordTick(stats_, WRITE_WITH_WAL, write_with_wal);
  }
-  return status;
+  return io_s;
 }

-Status DBImpl::ConcurrentWriteToWAL(const WriteThread::WriteGroup& write_group,
-                                    uint64_t* log_used,
-                                    SequenceNumber* last_sequence,
-                                    size_t seq_inc) {
-  Status status;
+IOStatus DBImpl::ConcurrentWriteToWAL(
+    const WriteThread::WriteGroup& write_group, uint64_t* log_used,
+    SequenceNumber* last_sequence, size_t seq_inc) {
+  IOStatus io_s;

  assert(!write_group.leader->disable_wal);
  // Same holds for all in the batch group
@@ -1092,14 +1120,14 @@ Status DBImpl::ConcurrentWriteToWAL(const WriteThread::WriteGroup& write_group,

  log::Writer* log_writer = logs_.back().writer;
  uint64_t log_size;
-  status = WriteToWAL(*merged_batch, log_writer, log_used, &log_size);
+  io_s = WriteToWAL(*merged_batch, log_writer, log_used, &log_size);
  if (to_be_cached_state) {
    cached_recoverable_state_ = *to_be_cached_state;
    cached_recoverable_state_empty_ = false;
  }
  log_write_mutex_.Unlock();

-  if (status.ok()) {
+  if (io_s.ok()) {
    const bool concurrent = true;
    auto stats = default_cf_internal_stats_;
    stats->AddDBStats(InternalStats::kIntStatsWalFileBytes, log_size,
@@ -1109,7 +1137,7 @@ Status DBImpl::ConcurrentWriteToWAL(const WriteThread::WriteGroup& write_group,
                      concurrent);
    RecordTick(stats_, WRITE_WITH_WAL, write_with_wal);
  }
-  return status;
+  return io_s;
 }

 Status DBImpl::WriteRecoverableState() {
@@ -1458,7 +1486,7 @@ Status DBImpl::ThrottleLowPriWritesIfNeeded(const WriteOptions& write_options,
      return Status::OK();
    }
    if (write_options.no_slowdown) {
-      return Status::Incomplete();
+      return Status::Incomplete("Low priority write stall");
    } else {
      assert(my_batch != nullptr);
      // Rate limit those writes. The reason that we don't completely wait
@@ -1631,7 +1659,6 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
  if (creating_new_log && immutable_db_options_.recycle_log_file_num &&
      !log_recycle_files_.empty()) {
    recycle_log_number = log_recycle_files_.front();
-    log_recycle_files_.pop_front();
  }
  uint64_t new_log_number =
      creating_new_log ? versions_->NewFileNumber() : logfile_number_;
@@ -1668,6 +1695,14 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
                 ". Immutable memtables: %d.\n",
                 cfd->GetName().c_str(), new_log_number, num_imm_unflushed);
  mutex_.Lock();
+  if (recycle_log_number != 0) {
+    // Since renaming the file is done outside DB mutex, we need to ensure
+    // concurrent full purges don't delete the file while we're recycling it.
+    // To achieve that we hold the old log number in the recyclable list until
+    // after it has been renamed.
+    assert(log_recycle_files_.front() == recycle_log_number);
+    log_recycle_files_.pop_front();
+  }
  if (s.ok() && creating_new_log) {
    log_write_mutex_.Lock();
    assert(new_log != nullptr);
@@ -1785,6 +1820,8 @@ Status DB::Put(const WriteOptions& opt, ColumnFamilyHandle* column_family,
  const Slice* ts = opt.timestamp;
  assert(nullptr != ts);
  size_t ts_sz = ts->size();
+  assert(column_family->GetComparator());
+  assert(ts_sz == column_family->GetComparator()->timestamp_size());
  WriteBatch batch(key.size() + ts_sz + value.size() + 24, /*max_bytes=*/0,
                   ts_sz);
  Status s = batch.Put(column_family, key, value);
@@ -1829,4 +1866,4 @@ Status DB::Merge(const WriteOptions& opt, ColumnFamilyHandle* column_family,
  }
  return Write(opt, &batch);
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -13,7 +13,7 @@
 #include "test_util/fault_injection_test_env.h"
 #include "test_util/sync_point.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 #ifndef ROCKSDB_LITE
 class DBSecondaryTest : public DBTestBase {
@@ -45,6 +45,8 @@ class DBSecondaryTest : public DBTestBase {

  void OpenSecondary(const Options& options);

+  Status TryOpenSecondary(const Options& options);
+
  void OpenSecondaryWithColumnFamilies(
      const std::vector<std::string>& column_families, const Options& options);

@@ -70,9 +72,13 @@ class DBSecondaryTest : public DBTestBase {
 };

 void DBSecondaryTest::OpenSecondary(const Options& options) {
+  ASSERT_OK(TryOpenSecondary(options));
+}
+
+Status DBSecondaryTest::TryOpenSecondary(const Options& options) {
  Status s =
      DB::OpenAsSecondary(options, dbname_, secondary_path_, &db_secondary_);
-  ASSERT_OK(s);
+  return s;
 }

 void DBSecondaryTest::OpenSecondaryWithColumnFamilies(
@@ -858,12 +864,62 @@ TEST_F(DBSecondaryTest, CheckConsistencyWhenOpen) {
  thread.join();
  ASSERT_TRUE(called);
 }
+
+TEST_F(DBSecondaryTest, StartFromInconsistent) {
+  Options options = CurrentOptions();
+  DestroyAndReopen(options);
+  ASSERT_OK(Put("foo", "value"));
+  ASSERT_OK(Flush());
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionBuilder::CheckConsistencyBeforeReturn", [&](void* arg) {
+        ASSERT_NE(nullptr, arg);
+        *(reinterpret_cast<Status*>(arg)) =
+            Status::Corruption("Inject corruption");
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+  Options options1;
+  Status s = TryOpenSecondary(options1);
+  ASSERT_TRUE(s.IsCorruption());
+}
+
+TEST_F(DBSecondaryTest, InconsistencyDuringCatchUp) {
+  Options options = CurrentOptions();
+  DestroyAndReopen(options);
+  ASSERT_OK(Put("foo", "value"));
+  ASSERT_OK(Flush());
+
+  Options options1;
+  OpenSecondary(options1);
+
+  {
+    std::string value;
+    ASSERT_OK(db_secondary_->Get(ReadOptions(), "foo", &value));
+    ASSERT_EQ("value", value);
+  }
+
+  ASSERT_OK(Put("bar", "value1"));
+  ASSERT_OK(Flush());
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionBuilder::CheckConsistencyBeforeReturn", [&](void* arg) {
+        ASSERT_NE(nullptr, arg);
+        *(reinterpret_cast<Status*>(arg)) =
+            Status::Corruption("Inject corruption");
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+  Status s = db_secondary_->TryCatchUpWithPrimary();
+  ASSERT_TRUE(s.IsCorruption());
+}
 #endif  //! ROCKSDB_LITE

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
-  rocksdb::port::InstallStackTraceHandler();
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
--- a/Show More
+++ b/Show More