Do not swallow error returned from SaveTo() (#6801 )

Summary: With consistency check enabled, VersionBuilder::SaveTo() may return error once corruption is detected while building versions. We should handle these errors. Pull Request resolved: https://github.com/facebook/rocksdb/pull/6801 Test Plan: make check Reviewed By: siying Differential Revision: D21385045 Pulled By: riversand963 fbshipit-source-id: 98f6424e2a4699b62befa21e9fe00e70a771118e
Fix compilation of compaction_picker_test after cherry-pick
2026-07-08 07:05:23 +08:00 · 2020-05-08 09:52:53 -07:00 · 2020-05-08 09:20:38 -07:00 · 2020-05-08 08:47:34 -07:00 · 2020-05-08 08:46:12 -07:00 · 2020-05-08 08:45:41 -07:00
1019 changed files with 34777 additions and 15885 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
@@ -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

@@ -52,22 +54,109 @@ env:

 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-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 most osx, arm64 and ppc64le tests for pull requests, but build in branches
+  # 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
+    os : osx
+    env: JOB_NAME=java_test
+  - if: type = pull_request
+    os : linux
+    arch: arm64
+    env: JOB_NAME=java_test
+  - if: type = pull_request
+    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

 install:
  - if [ "${TRAVIS_OS_NAME}" == osx ]; then
@@ -81,10 +170,29 @@ install:
      sudo apt-get install -y mingw-w64 ;
    fi
  - if [[ "${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:
@@ -103,7 +211,7 @@ 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
@@ -203,16 +203,25 @@ 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 +244,7 @@ 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()
@@ -481,6 +490,11 @@ 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)
@@ -496,6 +510,9 @@ set(SOURCES
        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 +566,7 @@ 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
@@ -604,12 +622,15 @@ set(SOURCES
        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 +638,14 @@ 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 +691,7 @@ 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 +760,11 @@ 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
@@ -906,18 +933,21 @@ 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 +958,7 @@ 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 +984,13 @@ 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_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
@@ -990,8 +1021,10 @@ 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
@@ -1025,6 +1058,7 @@ 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
@@ -1033,6 +1067,7 @@ 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/thread_list_test.cc
@@ -1074,6 +1109,7 @@ 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()
@@ -1166,9 +1202,15 @@ 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})
@@ -1,5 +1,90 @@
 # Rocksdb Change Log
-## Unreleased
+## 6.9.4 (05/08/2020)
+### Bug Fixes
+* Fix a bug caused by overwrite the status with io status in block based table builder when writing data blocks. If status stores the error message (e.g., failure of verify block compression), the bug will make the io status overwrite the status.
+* Fix consistency checking error swallowing in some cases when options.force_consistency_checks = true.
+* Fix possible false NotFound status from batched MultiGet using index type kHashSearch.
+* Fix corruption caused by enabling delete triggered compaction (NewCompactOnDeletionCollectorFactory) in universal compaction mode, along with parallel compactions. The bug can result in two parallel compactions picking the same input files, resulting in the DB resurrecting older and deleted versions of some keys.
+
+### Behavior Changes
+* Disable delete triggered compaction (NewCompactOnDeletionCollectorFactory) in universal compaction mode and num_levels = 1 in order to avoid a corruption bug.
+
+## 6.9.3 (04/28/2020)
+### Bug Fixes
+* 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.
+* Fix a bug by updating CURRENT file so that it points to the correct MANIFEST file after best-efforts recovery.
+* 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.
+
+## 6.9.2 (04/16/2020)
+### Public API Change
+* Add NewFileChecksumGenCrc32cFactory to the file checksum public API, such that the builtin Crc32c based file checksum generator factory can be used by applications.
+
+### New Features
+* Log CompactOnDeletionCollectorFactory window_size and deletion_trigger parameters in the info log file for troubleshooting purposes.
+* Report kFilesMarkedForCompaction for delete triggered compactions in case of Universal and Levelled Compactions.
+
+## 6.9.1 (04/09/2020)
+### 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.
+* 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.
+
+## 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
 * Added a rocksdb::FileSystem class in include/rocksdb/file_system.h to encapsulate file creation/read/write operations, and an option DBOptions::file_system to allow a user to pass in an instance of rocksdb::FileSystem. If its a non-null value, this will take precendence over DBOptions::env for file operations. A new API rocksdb::FileSystem::Default() returns a platform default object. The DBOptions::env option and Env::Default() API will continue to be used for threading and other OS related functions, and where DBOptions::file_system is not specified, for file operations. For storage developers who are accustomed to rocksdb::Env, the interface in rocksdb::FileSystem is new and will probably undergo some changes as more storage systems are ported to it from rocksdb::Env. As of now, no env other than Posix has been ported to the new interface.
 * A new rocksdb::NewSstFileManager() API that allows the caller to pass in separate Env and FileSystem objects.
@@ -8,18 +93,35 @@

 ### Bug Fixes
 * Fix a bug that can cause unnecessary bg thread to be scheduled(#6104).
-* Fix a bug in which a snapshot read could be affected by a DeleteRange after the snapshot (#6062).
 * Fix crash caused by concurrent CF iterations and drops(#6147).
-* Fixed two performance issues related to memtable history trimming. First, a new SuperVersion is now created only if some memtables were actually trimmed. Second, trimming is only scheduled if there is at least one flushed memtable that is kept in memory for the purposes of transaction conflict checking.
-* Fix a bug in WriteBatchWithIndex::MultiGetFromBatchAndDB, which is called by Transaction::MultiGet, that causes due to stale pointer access when the number of keys is > 32
-* BlobDB no longer updates the SST to blob file mapping upon failed compactions.
-* Fixed a bug where BlobDB was comparing the `ColumnFamilyHandle` pointers themselves instead of only the column family IDs when checking whether an API call uses the default column family or not.
 * Fix a race condition for cfd->log_number_ between manifest switch and memtable switch (PR 6249) when number of column families is greater than 1.
+* Fix a bug on fractional cascading index when multiple files at the same level contain the same smallest user key, and those user keys are for merge operands. In this case, Get() the exact key may miss some merge operands.
+* Delcare kHashSearch index type feature-incompatible with index_block_restart_interval larger than 1.
+* Fixed an issue where the thread pools were not resized upon setting `max_background_jobs` dynamically through the `SetDBOptions` interface.
+* Fix a bug that can cause write threads to hang when a slowdown/stall happens and there is a mix of writers with WriteOptions::no_slowdown set/unset.
+* Fixed an issue where an incorrect "number of input records" value was used to compute the "records dropped" statistics for compactions.
+* Fix a regression bug that causes segfault when hash is used, max_open_files != -1 and total order seek is used and switched back.

 ### New Features
 * 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
+* Fixed a bug where non-L0 compaction input files were not considered to compute the `creation_time` of new compaction outputs.
+
+## 6.6.1 (01/02/2020)
+### Bug Fixes
+* Fix a bug in WriteBatchWithIndex::MultiGetFromBatchAndDB, which is called by Transaction::MultiGet, that causes due to stale pointer access when the number of keys is > 32
+* Fixed two performance issues related to memtable history trimming. First, a new SuperVersion is now created only if some memtables were actually trimmed. Second, trimming is only scheduled if there is at least one flushed memtable that is kept in memory for the purposes of transaction conflict checking.
+* BlobDB no longer updates the SST to blob file mapping upon failed compactions.
+* Fix a bug in which a snapshot read through an iterator could be affected by a DeleteRange after the snapshot (#6062).
+* Fixed a bug where BlobDB was comparing the `ColumnFamilyHandle` pointers themselves instead of only the column family IDs when checking whether an API call uses the default column family or not.
+* Delete superversions in BackgroundCallPurge.
+* Fix use-after-free and double-deleting files in BackgroundCallPurge().

 ## 6.6.0 (11/25/2019)
 ### 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}
@@ -207,6 +208,7 @@ AM_LINK = $(AM_V_CCLD)$(CXX) $^ $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS
 dummy := $(shell (export ROCKSDB_ROOT="$(CURDIR)"; export PORTABLE="$(PORTABLE)"; "$(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 +218,7 @@ 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
@@ -445,6 +447,7 @@ EXPOBJECTS = $(LIBOBJECTS) $(TESTUTIL)

 TESTS = \
 	db_basic_test \
+	db_with_timestamp_basic_test \
 	db_encryption_test \
 	db_test2 \
 	external_sst_file_basic_test \
@@ -459,6 +462,7 @@ TESTS = \
 	env_basic_test \
 	env_test \
 	env_logger_test \
+	io_posix_test \
 	hash_test \
 	random_test \
 	thread_local_test \
@@ -468,6 +472,7 @@ TESTS = \
 	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 \
@@ -492,7 +497,7 @@ 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 \
@@ -503,11 +508,13 @@ TESTS = \
 	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 \
@@ -595,6 +602,9 @@ TESTS = \
 	db_secondary_test \
 	block_cache_tracer_test \
 	block_cache_trace_analyzer_test \
+	defer_test \
+	blob_file_addition_test \
+	blob_file_garbage_test \

 ifeq ($(USE_FOLLY_DISTRIBUTED_MUTEX),1)
 	TESTS += folly_synchronization_distributed_mutex_test
@@ -939,9 +949,9 @@ check: all
 	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
@@ -952,7 +962,7 @@ check_some: $(SUBSET)

 .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 +971,31 @@ 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:
@@ -1128,16 +1138,21 @@ 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*
@@ -1286,12 +1301,18 @@ corruption_test: db/corruption_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARN
 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_encryption_test: db/db_encryption_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1301,7 +1322,10 @@ 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 +1379,7 @@ 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 +1492,9 @@ 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 +1507,9 @@ 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)

@@ -1705,6 +1735,15 @@ 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)
+	$(AM_LINK)
+
 #-------------------------------------------------
 # make install related stuff
 INSTALL_PATH ?= /usr/local
@@ -1784,11 +1823,11 @@ SHA256_CMD = sha256sum
 ZLIB_VER ?= 1.2.11
 ZLIB_SHA256 ?= c3e5e9fdd5004dcb542feda5ee4f0ff0744628baf8ed2dd5d66f8ca1197cb1a1
 ZLIB_DOWNLOAD_BASE ?= http://zlib.net
-BZIP2_VER ?= 1.0.6
-BZIP2_SHA256 ?= a2848f34fcd5d6cf47def00461fcb528a0484d8edef8208d6d2e2909dc61d9cd
-BZIP2_DOWNLOAD_BASE ?= https://downloads.sourceforge.net/project/bzip2
-SNAPPY_VER ?= 1.1.7
-SNAPPY_SHA256 ?= 3dfa02e873ff51a11ee02b9ca391807f0c8ea0529a4924afa645fbf97163f9d4
+BZIP2_VER ?= 1.0.8
+BZIP2_SHA256 ?= ab5a03176ee106d3f0fa90e381da478ddae405918153cca248e682cd0c4a2269
+BZIP2_DOWNLOAD_BASE ?= https://sourceware.org/pub/bzip2
+SNAPPY_VER ?= 1.1.8
+SNAPPY_SHA256 ?= 16b677f07832a612b0836178db7f374e414f94657c138e6993cbfc5dcc58651f
 SNAPPY_DOWNLOAD_BASE ?= https://github.com/google/snappy/archive
 LZ4_VER ?= 1.9.2
 LZ4_SHA256 ?= 658ba6191fa44c92280d4aa2c271b0f4fbc0e34d249578dd05e50e76d0e5efcc
@@ -1834,7 +1873,7 @@ endif
 libz.a:
 	-rm -rf zlib-$(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)" != "$$ZLIB_SHA256_ACTUAL" ]; then \
@@ -1848,7 +1887,7 @@ endif
 libbz2.a:
 	-rm -rf bzip2-$(BZIP2_VER)
 ifeq (,$(wildcard ./bzip2-$(BZIP2_VER).tar.gz))
-	curl --output bzip2-$(BZIP2_VER).tar.gz -L ${CURL_SSL_OPTS} ${BZIP2_DOWNLOAD_BASE}/bzip2-$(BZIP2_VER).tar.gz
+	curl --fail --output bzip2-$(BZIP2_VER).tar.gz --location ${CURL_SSL_OPTS} ${BZIP2_DOWNLOAD_BASE}/bzip2-$(BZIP2_VER).tar.gz
 endif
 	BZIP2_SHA256_ACTUAL=`$(SHA256_CMD) bzip2-$(BZIP2_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(BZIP2_SHA256)" != "$$BZIP2_SHA256_ACTUAL" ]; then \
@@ -1862,7 +1901,7 @@ 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)" != "$$SNAPPY_SHA256_ACTUAL" ]; then \
@@ -1877,7 +1916,7 @@ endif
 liblz4.a:
 	-rm -rf lz4-$(LZ4_VER)
 ifeq (,$(wildcard ./lz4-$(LZ4_VER).tar.gz))
-	curl --output lz4-$(LZ4_VER).tar.gz -L ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
+	curl --fail --output lz4-$(LZ4_VER).tar.gz --location ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
 endif
 	LZ4_SHA256_ACTUAL=`$(SHA256_CMD) lz4-$(LZ4_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(LZ4_SHA256)" != "$$LZ4_SHA256_ACTUAL" ]; then \
@@ -1891,7 +1930,7 @@ 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 +2000,35 @@ 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 +2094,7 @@ 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;
@@ -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",
@@ -117,6 +116,9 @@ cpp_library(
        "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 +172,7 @@ 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",
@@ -230,12 +233,15 @@ cpp_library(
        "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 +249,14 @@ 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 +295,7 @@ 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 +379,7 @@ 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",
@@ -477,6 +490,20 @@ 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",
@@ -689,7 +716,7 @@ ROCKS_TESTS = [
    ],
    [
        "db_blob_index_test",
-        "db/db_blob_index_test.cc",
+        "db/blob/db_blob_index_test.cc",
        "serial",
        [],
        [],
@@ -785,6 +812,13 @@ 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 +924,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "db_with_timestamp_basic_test",
+        "db/db_with_timestamp_basic_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "db_write_test",
        "db/db_write_test.cc",
@@ -904,6 +945,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "defer_test",
+        "util/defer_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "delete_scheduler_test",
        "file/delete_scheduler_test.cc",
@@ -954,8 +1002,8 @@ ROCKS_TESTS = [
        [],
    ],
    [
-        "error_handler_test",
-        "db/error_handler_test.cc",
+        "error_handler_fs_test",
+        "db/error_handler_fs_test.cc",
        "serial",
        [],
        [],
@@ -1072,6 +1120,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "io_posix_test",
+        "env/io_posix_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "iostats_context_test",
        "monitoring/iostats_context_test.cc",
@@ -1247,6 +1302,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "random_access_file_reader_test",
+        "file/random_access_file_reader_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "random_test",
        "util/random_test.cc",
@@ -1310,6 +1372,13 @@ ROCKS_TESTS = [
        [],
        [],
    ],
+    [
+        "slice_test",
+        "util/slice_test.cc",
+        "serial",
+        [],
+        [],
+    ],
    [
        "slice_transform_test",
        "util/slice_transform_test.cc",
@@ -102,3 +102,7 @@ LzLabs is using RocksDB as a storage engine in their multi-database distributed
 
 ## Crux
 [Crux](https://github.com/juxt/crux) is a document database that uses RocksDB for local [EAV](https://en.wikipedia.org/wiki/Entity%E2%80%93attribute%E2%80%93value_model) index storage to enable point-in-time bitemporal Datalog queries. The "unbundled" architecture uses Kafka to provide horizontal scalability.
+
+## Nebula Graph
+
+[Nebula Graph](https://github.com/vesoft-inc/nebula) is a distributed, scalable, lightning-fast, open source graph database capable of hosting super large scale graphs with dozens of billions of vertices (nodes) and trillions of edges, with milliseconds of latency.
@@ -29,7 +29,7 @@ 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
@@ -39,7 +39,7 @@ install:
  - 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 +47,7 @@ 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
@@ -68,7 +68,7 @@ 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
@@ -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
@@ -150,18 +151,20 @@ 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
@@ -237,6 +240,7 @@ 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.
@@ -506,6 +510,20 @@ 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 $PLATFORM_CXXFLAGS -faligned-new -x c++ - -o /dev/null 2>/dev/null <<EOF
@@ -694,6 +712,7 @@ echo "PLATFORM=$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"
@@ -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
    }
@@ -880,7 +884,7 @@ 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 +1063,6 @@ case $1 in
    ;;
  *)
    echo "Invalid determinator command"
+    exit 1
    ;;
 esac
@@ -45,7 +45,7 @@ DEFINE_int32(erase_percent, 10,

 DEFINE_bool(use_clock_cache, false, "");

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class CacheBench;
 namespace {
@@ -154,7 +154,7 @@ class CacheBench {
  }

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

    PrintEnv();
    SharedState shared(this);
@@ -257,7 +257,7 @@ class CacheBench {
    printf("----------------------------\n");
  }
 };
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE

 int main(int argc, char** argv) {
  ParseCommandLineFlags(&argc, &argv, true);
@@ -267,7 +267,7 @@ int main(int argc, char** argv) {
    exit(1);
  }

-  rocksdb::CacheBench bench;
+  ROCKSDB_NAMESPACE::CacheBench bench;
  if (FLAGS_populate_cache) {
    bench.PopulateCache();
  }
@@ -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 {

@@ -756,6 +756,6 @@ 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();
@@ -571,4 +571,4 @@ 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
@@ -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) {
@@ -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 +1,52 @@
+//  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 {
+
+SharedBlobFileMetaData::~SharedBlobFileMetaData() {
+  // TODO: add the blob file to the list of obsolete files here
+}
+
+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 +1,124 @@
+//  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:
+  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());
+  }
+
+  ~SharedBlobFileMetaData();
+
+  SharedBlobFileMetaData(const SharedBlobFileMetaData&) = delete;
+  SharedBlobFileMetaData& operator=(const 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:
+  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:
+  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());
+  }
+
+  ~BlobFileMetaData() = default;
+
+  BlobFileMetaData(const BlobFileMetaData&) = delete;
+  BlobFileMetaData& operator=(const 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:
+  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,
@@ -184,11 +186,13 @@ Status BuildTable(
    tp = builder->GetTableProperties();
    bool empty = builder->NumEntries() == 0 && tp.num_range_deletions == 0;
    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 +209,22 @@ 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
@@ -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** column_family_names,
-    const rocksdb_options_t** column_family_options,
-    rocksdb_column_family_handle_t** column_family_handles,
-    char** errptr) {
+    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) {
  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** column_family_names,
-    const rocksdb_options_t** column_family_options,
+    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(
@@ -735,8 +731,8 @@ rocksdb_t* rocksdb_open_for_read_only_column_families(
 rocksdb_t* rocksdb_open_as_secondary_column_families(
    const rocksdb_options_t* db_options, const char* name,
    const char* secondary_path, int num_column_families,
-    const char** column_family_names,
-    const rocksdb_options_t** column_family_options,
+    const char* const* column_family_names,
+    const rocksdb_options_t* const* column_family_options,
    rocksdb_column_family_handle_t** column_family_handles, char** errptr) {
  std::vector<ColumnFamilyDescriptor> column_families;
  for (int i = 0; i != num_column_families; ++i) {
@@ -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());
  }
@@ -2736,6 +2745,11 @@ 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;
+}
+
 rocksdb_ratelimiter_t* rocksdb_ratelimiter_create(
    int64_t rate_bytes_per_sec,
    int64_t refill_period_us,
@@ -2751,6 +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);
@@ -2758,7 +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;
 }

@@ -3060,11 +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);
    }
@@ -3166,7 +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(
@@ -3338,7 +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;
 }
@@ -3553,7 +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;
@@ -3561,7 +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;
@@ -3569,7 +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;
 }

@@ -3600,7 +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(
@@ -3630,7 +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;
@@ -3837,8 +3859,8 @@ rocksdb_transactiondb_t* rocksdb_transactiondb_open(
 rocksdb_transactiondb_t* rocksdb_transactiondb_open_column_families(
    const rocksdb_options_t* options,
    const rocksdb_transactiondb_options_t* txn_db_options, const char* name,
-    int num_column_families, const char** column_family_names,
-    const rocksdb_options_t** column_family_options,
+    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) {
  std::vector<ColumnFamilyDescriptor> column_families;
  for (int i = 0; i < num_column_families; i++) {
@@ -4215,8 +4237,8 @@ rocksdb_optimistictransactiondb_t* rocksdb_optimistictransactiondb_open(
 rocksdb_optimistictransactiondb_t*
 rocksdb_optimistictransactiondb_open_column_families(
    const rocksdb_options_t* db_options, const char* name,
-    int num_column_families, const char** column_family_names,
-    const rocksdb_options_t** column_family_options,
+    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) {
  std::vector<ColumnFamilyDescriptor> column_families;
  for (int i = 0; i < num_column_families; i++) {
@@ -4332,7 +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;
@@ -4360,7 +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;
@@ -4382,7 +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);
@@ -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)
@@ -160,11 +160,6 @@ Status CheckConcurrentWritesSupported(const ColumnFamilyOptions& cf_options) {
    return Status::InvalidArgument(
        "Memtable doesn't concurrent writes (allow_concurrent_memtable_write)");
  }
-  if (cf_options.max_successive_merges != 0) {
-    return Status::InvalidArgument(
-        "max_successive_merges > 0 is incompatible "
-        "with concurrent writes (allow_concurrent_memtable_write)");
-  }
  return Status::OK();
 }

@@ -476,6 +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,
@@ -512,7 +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.
@@ -606,6 +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() {
@@ -1252,6 +1286,11 @@ Status ColumnFamilyData::ValidateOptions(
  if (s.ok() && db_options.allow_concurrent_memtable_write) {
    s = CheckConcurrentWritesSupported(cf_options);
  }
+  if (s.ok() && db_options.unordered_write &&
+      cf_options.max_successive_merges != 0) {
+    s = Status::InvalidArgument(
+        "max_successive_merges > 0 is incompatible with unordered_write");
+  }
  if (s.ok()) {
    s = CheckCFPathsSupported(db_options, cf_options);
  }
@@ -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
@@ -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_;
  }
@@ -595,14 +595,15 @@ 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 +612,7 @@ TEST_P(ColumnFamilyTest, CreateCFRaceWithGetAggProperty) {

  thread.join();

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

@@ -820,7 +821,7 @@ 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 +1254,14 @@ 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 +1429,11 @@ 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 +1442,7 @@ 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 +1485,8 @@ 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 +1525,11 @@ 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 +1537,7 @@ 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 +1555,7 @@ 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 +1582,8 @@ 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 +1629,11 @@ 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 +1645,8 @@ 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 +1681,8 @@ 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 +1723,13 @@ 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 +1741,8 @@ 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 +1762,7 @@ 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 +1783,8 @@ 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 +1825,12 @@ 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 +1842,8 @@ 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 +1876,8 @@ 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 +1918,14 @@ 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 +1937,8 @@ 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 +1969,8 @@ 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 +2010,12 @@ 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 +2027,7 @@ 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 +2061,8 @@ 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 +2417,7 @@ 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 +2425,7 @@ 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 +2504,21 @@ 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 +2530,12 @@ 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 +2548,8 @@ 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 +2984,10 @@ 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 +3003,7 @@ 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 +3081,14 @@ 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 +3106,7 @@ 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 +3150,14 @@ 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 +3181,7 @@ 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 +3201,15 @@ 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 +3220,7 @@ TEST_P(ColumnFamilyTest, LogSyncConflictFlush) {

  thread.join();

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  Close();
 }
 #endif
@@ -3367,7 +3369,7 @@ TEST_P(ColumnFamilyTest, MultipleCFPathsTest) {
  }
 }

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

 #ifdef ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS
 extern "C" {
@@ -3378,7 +3380,7 @@ 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
@@ -16,7 +16,7 @@
 #include "test_util/sync_point.h"
 #include "util/string_util.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 const uint64_t kRangeTombstoneSentinel =
    PackSequenceAndType(kMaxSequenceNumber, kTypeRangeDeletion);
@@ -545,11 +545,13 @@ bool Compaction::ShouldFormSubcompactions() const {

 uint64_t Compaction::MinInputFileOldestAncesterTime() const {
  uint64_t min_oldest_ancester_time = port::kMaxUint64;
-  for (const auto& file : inputs_[0].files) {
-    uint64_t oldest_ancester_time = file->TryGetOldestAncesterTime();
-    if (oldest_ancester_time != 0) {
-      min_oldest_ancester_time =
-          std::min(min_oldest_ancester_time, oldest_ancester_time);
+  for (const auto& level_files : inputs_) {
+    for (const auto& file : level_files.files) {
+      uint64_t oldest_ancester_time = file->TryGetOldestAncesterTime();
+      if (oldest_ancester_time != 0) {
+        min_oldest_ancester_time =
+            std::min(min_oldest_ancester_time, oldest_ancester_time);
+      }
    }
  }
  return min_oldest_ancester_time;
@@ -559,4 +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) {
@@ -144,7 +144,6 @@ struct CompactionJob::SubcompactionState {

  // State during the subcompaction
  uint64_t total_bytes;
-  uint64_t num_input_records;
  uint64_t num_output_records;
  CompactionJobStats compaction_job_stats;
  uint64_t approx_size;
@@ -165,7 +164,6 @@ struct CompactionJob::SubcompactionState {
        builder(nullptr),
        current_output_file_size(0),
        total_bytes(0),
-        num_input_records(0),
        num_output_records(0),
        approx_size(size),
        grandparent_index(0),
@@ -186,7 +184,6 @@ struct CompactionJob::SubcompactionState {
    builder = std::move(o.builder);
    current_output_file_size = std::move(o.current_output_file_size);
    total_bytes = std::move(o.total_bytes);
-    num_input_records = std::move(o.num_input_records);
    num_output_records = std::move(o.num_output_records);
    compaction_job_stats = std::move(o.compaction_job_stats);
    approx_size = std::move(o.approx_size);
@@ -245,13 +242,11 @@ struct CompactionJob::CompactionState {
  Status status;

  uint64_t total_bytes;
-  uint64_t num_input_records;
  uint64_t num_output_records;

  explicit CompactionState(Compaction* c)
      : compaction(c),
        total_bytes(0),
-        num_input_records(0),
        num_output_records(0) {}

  size_t NumOutputFiles() {
@@ -289,7 +284,6 @@ struct CompactionJob::CompactionState {
 void CompactionJob::AggregateStatistics() {
  for (SubcompactionState& sc : compact_->sub_compact_states) {
    compact_->total_bytes += sc.total_bytes;
-    compact_->num_input_records += sc.num_input_records;
    compact_->num_output_records += sc.num_output_records;
  }
  if (compaction_job_stats_) {
@@ -304,7 +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,
@@ -619,8 +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()) {
@@ -719,9 +718,13 @@ 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_;
@@ -770,12 +773,12 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
  auto stream = event_logger_->LogToBuffer(log_buffer_);
  stream << "job" << job_id_ << "event"
         << "compaction_finished"
-         << "compaction_time_micros" << compaction_stats_.micros
-         << "compaction_time_cpu_micros" << compaction_stats_.cpu_micros
-         << "output_level" << compact_->compaction->output_level()
-         << "num_output_files" << compact_->NumOutputFiles()
-         << "total_output_size" << compact_->total_bytes << "num_input_records"
-         << compact_->num_input_records << "num_output_records"
+         << "compaction_time_micros" << stats.micros
+         << "compaction_time_cpu_micros" << stats.cpu_micros << "output_level"
+         << compact_->compaction->output_level() << "num_output_files"
+         << compact_->NumOutputFiles() << "total_output_size"
+         << compact_->total_bytes << "num_input_records"
+         << stats.num_input_records << "num_output_records"
         << compact_->num_output_records << "num_subcompactions"
         << compact_->sub_compact_states.size() << "output_compression"
         << CompressionTypeToString(compact_->compaction->output_compression());
@@ -991,7 +994,6 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
    }
  }

-  sub_compact->num_input_records = c_iter_stats.num_input_records;
  sub_compact->compaction_job_stats.num_input_deletion_records =
      c_iter_stats.num_input_deletion_records;
  sub_compact->compaction_job_stats.num_corrupt_keys =
@@ -1301,6 +1303,10 @@ 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;
@@ -1309,12 +1315,23 @@ 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();

@@ -1515,7 +1532,8 @@ 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
@@ -1589,6 +1607,8 @@ void CompactionJob::UpdateCompactionStats() {
    }
  }

+  uint64_t num_output_records = 0;
+
  for (const auto& sub_compact : compact_->sub_compact_states) {
    size_t num_output_files = sub_compact.outputs.size();
    if (sub_compact.builder != nullptr) {
@@ -1598,13 +1618,16 @@ void CompactionJob::UpdateCompactionStats() {
    }
    compaction_stats_.num_output_files += static_cast<int>(num_output_files);

+    num_output_records += sub_compact.num_output_records;
+
    for (const auto& out : sub_compact.outputs) {
      compaction_stats_.bytes_written += out.meta.fd.file_size;
    }
-    if (sub_compact.num_input_records > sub_compact.num_output_records) {
-      compaction_stats_.num_dropped_records +=
-          sub_compact.num_input_records - sub_compact.num_output_records;
-    }
+  }
+
+  if (compaction_stats_.num_input_records > num_output_records) {
+    compaction_stats_.num_dropped_records =
+        compaction_stats_.num_input_records - num_output_records;
  }
 }

@@ -1632,7 +1655,7 @@ void CompactionJob::UpdateCompactionJobStats(
    // input information
    compaction_job_stats_->total_input_bytes =
        stats.bytes_read_non_output_levels + stats.bytes_read_output_level;
-    compaction_job_stats_->num_input_records = compact_->num_input_records;
+    compaction_job_stats_->num_input_records = stats.num_input_records;
    compaction_job_stats_->num_input_files =
        stats.num_input_files_in_non_output_levels +
        stats.num_input_files_in_output_level;
@@ -1693,4 +1716,4 @@ 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);

@@ -1085,6 +1085,8 @@ void CompactionPicker::PickFilesMarkedForCompaction(
  Random64 rnd(/* seed */ reinterpret_cast<uint64_t>(vstorage));
  size_t random_file_index = static_cast<size_t>(rnd.Uniform(
      static_cast<uint64_t>(vstorage->FilesMarkedForCompaction().size())));
+  TEST_SYNC_POINT_CALLBACK("CompactionPicker::PickFilesMarkedForCompaction",
+                           &random_file_index);

  if (continuation(vstorage->FilesMarkedForCompaction()[random_file_index])) {
    // found the compaction!
@@ -1128,4 +1130,4 @@ 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 +70,17 @@ 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;
+      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 +95,21 @@ Compaction* FIFOCompactionPicker::PickTTLCompaction(
  }

  for (const auto& f : inputs[0].files) {
+    uint64_t creation_time = 0;
+    if (f && 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 +149,7 @@ 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 +197,9 @@ 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 +242,5 @@ 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 {
@@ -250,7 +250,6 @@ void LevelCompactionBuilder::SetupInitialFiles() {
        cf_name_, vstorage_, &start_level_, &output_level_,
        &start_level_inputs_);
    if (!start_level_inputs_.empty()) {
-      is_manual_ = true;
      compaction_reason_ = CompactionReason::kFilesMarkedForCompaction;
      return;
    }
@@ -384,7 +383,7 @@ 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 +554,4 @@ 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:
@@ -78,8 +78,17 @@ class CompactionPickerTest : public testing::Test {
    vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
  }

+  // Create a new VersionStorageInfo object so we can add mode files and then
+  // merge it with the existing VersionStorageInfo
+  void AddVersionStorage() {
+    temp_vstorage_.reset(new VersionStorageInfo(
+        &icmp_, ucmp_, options_.num_levels, ioptions_.compaction_style,
+        vstorage_.get(), false));
+  }
+
  void DeleteVersionStorage() {
    vstorage_.reset();
+    temp_vstorage_.reset();
    files_.clear();
    file_map_.clear();
    input_files_.clear();
@@ -88,17 +97,24 @@ class CompactionPickerTest : public testing::Test {
  void Add(int level, uint32_t file_number, const char* smallest,
           const char* largest, uint64_t file_size = 1, uint32_t path_id = 0,
           SequenceNumber smallest_seq = 100, SequenceNumber largest_seq = 100,
-           size_t compensated_file_size = 0) {
-    assert(level < vstorage_->num_levels());
+           size_t compensated_file_size = 0, bool marked_for_compact = false) {
+    VersionStorageInfo* vstorage;
+    if (temp_vstorage_) {
+      vstorage = temp_vstorage_.get();
+    } else {
+      vstorage = vstorage_.get();
+    }
+    assert(level < vstorage->num_levels());
    FileMetaData* f = new FileMetaData(
        file_number, path_id, file_size,
        InternalKey(smallest, smallest_seq, kTypeValue),
        InternalKey(largest, largest_seq, kTypeValue), smallest_seq,
-        largest_seq, /* marked_for_compact */ false, kInvalidBlobFileNumber,
-        kUnknownOldestAncesterTime, kUnknownFileCreationTime);
+        largest_seq, marked_for_compact, kInvalidBlobFileNumber,
+        kUnknownOldestAncesterTime, kUnknownFileCreationTime,
+        kUnknownFileChecksum, kUnknownFileChecksumFuncName);
    f->compensated_file_size =
        (compensated_file_size != 0) ? compensated_file_size : file_size;
-    vstorage_->AddFile(level, f);
+    vstorage->AddFile(level, f);
    files_.emplace_back(f);
    file_map_.insert({file_number, {f, level}});
  }
@@ -121,6 +137,11 @@ class CompactionPickerTest : public testing::Test {
  }

  void UpdateVersionStorageInfo() {
+    if (temp_vstorage_) {
+      VersionBuilder builder(FileOptions(), nullptr, vstorage_.get());
+      builder.SaveTo(temp_vstorage_.get());
+      vstorage_ = std::move(temp_vstorage_);
+    }
    vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
    vstorage_->UpdateFilesByCompactionPri(ioptions_.compaction_pri);
    vstorage_->UpdateNumNonEmptyLevels();
@@ -131,6 +152,27 @@ class CompactionPickerTest : public testing::Test {
    vstorage_->ComputeFilesMarkedForCompaction();
    vstorage_->SetFinalized();
  }
+  void AddFileToVersionStorage(int level, uint32_t file_number,
+                               const char* smallest, const char* largest,
+                               uint64_t file_size = 1, uint32_t path_id = 0,
+                               SequenceNumber smallest_seq = 100,
+                               SequenceNumber largest_seq = 100,
+                               size_t compensated_file_size = 0,
+                               bool marked_for_compact = false) {
+    VersionStorageInfo* base_vstorage = vstorage_.release();
+    vstorage_.reset(new VersionStorageInfo(&icmp_, ucmp_, options_.num_levels,
+                                           kCompactionStyleUniversal,
+                                           base_vstorage, false));
+    Add(level, file_number, smallest, largest, file_size, path_id, smallest_seq,
+        largest_seq, compensated_file_size, marked_for_compact);
+
+    VersionBuilder builder(FileOptions(), nullptr, base_vstorage);
+    builder.SaveTo(vstorage_.get());
+    UpdateVersionStorageInfo();
+  }
+
+ private:
+  std::unique_ptr<VersionStorageInfo> temp_vstorage_;
 };

 TEST_F(CompactionPickerTest, Empty) {
@@ -1732,7 +1774,164 @@ TEST_F(CompactionPickerTest, IntraL0ForEarliestSeqno) {
  ASSERT_EQ(0, compaction->output_level());
 }

-}  // namespace rocksdb
+TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap) {
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+
+  // This test covers the case where a "regular" universal compaction is
+  // scheduled first, followed by a delete triggered compaction. The latter
+  // should fail
+  NewVersionStorage(5, kCompactionStyleUniversal);
+
+  Add(0, 1U, "150", "200", kFileSize, 0, 500, 550);
+  Add(0, 2U, "201", "250", 2 * kFileSize, 0, 401, 450);
+  Add(0, 4U, "260", "300", 4 * kFileSize, 0, 260, 300);
+  Add(3, 5U, "010", "080", 8 * kFileSize, 0, 200, 251);
+  Add(4, 3U, "301", "350", 8 * kFileSize, 0, 101, 150);
+  Add(4, 6U, "501", "750", 8 * kFileSize, 0, 101, 150);
+
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+
+  ASSERT_TRUE(compaction);
+  // Validate that its a compaction to reduce sorted runs
+  ASSERT_EQ(CompactionReason::kUniversalSortedRunNum,
+            compaction->compaction_reason());
+  ASSERT_EQ(0, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(2U, compaction->num_input_files(0));
+
+  AddVersionStorage();
+  // Simulate a flush and mark the file for compaction
+  Add(0, 1U, "150", "200", kFileSize, 0, 551, 600, 0, true);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction2(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+  ASSERT_FALSE(compaction2);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedCompactionFullOverlap2) {
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+
+  // This test covers the case where a delete triggered compaction is
+  // scheduled first, followed by a "regular" compaction. The latter
+  // should fail
+  NewVersionStorage(5, kCompactionStyleUniversal);
+
+  // Mark file number 4 for compaction
+  Add(0, 4U, "260", "300", 4 * kFileSize, 0, 260, 300, 0, true);
+  Add(3, 5U, "240", "290", 8 * kFileSize, 0, 201, 250);
+  Add(4, 3U, "301", "350", 8 * kFileSize, 0, 101, 150);
+  Add(4, 6U, "501", "750", 8 * kFileSize, 0, 101, 150);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+
+  ASSERT_TRUE(compaction);
+  // Validate that its a delete triggered compaction
+  ASSERT_EQ(CompactionReason::kFilesMarkedForCompaction,
+            compaction->compaction_reason());
+  ASSERT_EQ(3, compaction->output_level());
+  ASSERT_EQ(0, compaction->start_level());
+  ASSERT_EQ(1U, compaction->num_input_files(0));
+  ASSERT_EQ(1U, compaction->num_input_files(1));
+
+  AddVersionStorage();
+  Add(0, 1U, "150", "200", kFileSize, 0, 500, 550);
+  Add(0, 2U, "201", "250", 2 * kFileSize, 0, 401, 450);
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction2(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+  ASSERT_FALSE(compaction2);
+}
+
+TEST_F(CompactionPickerTest, UniversalMarkedCompactionStartOutputOverlap) {
+  // The case where universal periodic compaction can be picked
+  // with some newer files being compacted.
+  const uint64_t kFileSize = 100000;
+
+  ioptions_.compaction_style = kCompactionStyleUniversal;
+
+  bool input_level_overlap = false;
+  bool output_level_overlap = false;
+  // Let's mark 2 files in 2 different levels for compaction. The
+  // compaction picker will randomly pick one, so use the sync point to
+  // ensure a deterministic order. Loop until both cases are covered
+  size_t random_index = 0;
+  SyncPoint::GetInstance()->SetCallBack(
+      "CompactionPicker::PickFilesMarkedForCompaction", [&](void* arg) {
+        size_t* index = static_cast<size_t*>(arg);
+        *index = random_index;
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+  while (!input_level_overlap || !output_level_overlap) {
+    // Ensure that the L0 file gets picked first
+    random_index = !input_level_overlap ? 0 : 1;
+    UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+    NewVersionStorage(5, kCompactionStyleUniversal);
+
+    Add(0, 1U, "260", "300", 4 * kFileSize, 0, 260, 300, 0, true);
+    Add(3, 2U, "010", "020", 2 * kFileSize, 0, 201, 248);
+    Add(3, 3U, "250", "270", 2 * kFileSize, 0, 202, 249);
+    Add(3, 4U, "290", "310", 2 * kFileSize, 0, 203, 250);
+    Add(3, 5U, "310", "320", 2 * kFileSize, 0, 204, 251, 0, true);
+    Add(4, 6U, "301", "350", 8 * kFileSize, 0, 101, 150);
+    Add(4, 7U, "501", "750", 8 * kFileSize, 0, 101, 150);
+    UpdateVersionStorageInfo();
+
+    std::unique_ptr<Compaction> compaction(
+        universal_compaction_picker.PickCompaction(
+            cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+
+    ASSERT_TRUE(compaction);
+    // Validate that its a delete triggered compaction
+    ASSERT_EQ(CompactionReason::kFilesMarkedForCompaction,
+              compaction->compaction_reason());
+    ASSERT_TRUE(compaction->start_level() == 0 ||
+                compaction->start_level() == 3);
+    if (compaction->start_level() == 0) {
+      // The L0 file was picked. The next compaction will detect an
+      // overlap on its input level
+      input_level_overlap = true;
+      ASSERT_EQ(3, compaction->output_level());
+      ASSERT_EQ(1U, compaction->num_input_files(0));
+      ASSERT_EQ(3U, compaction->num_input_files(1));
+    } else {
+      // The level 3 file was picked. The next compaction will pick
+      // the L0 file and will detect overlap when adding output
+      // level inputs
+      output_level_overlap = true;
+      ASSERT_EQ(4, compaction->output_level());
+      ASSERT_EQ(2U, compaction->num_input_files(0));
+      ASSERT_EQ(1U, compaction->num_input_files(1));
+    }
+
+    vstorage_->ComputeCompactionScore(ioptions_, mutable_cf_options_);
+    // After recomputing the compaction score, only one marked file will remain
+    random_index = 0;
+    std::unique_ptr<Compaction> compaction2(
+        universal_compaction_picker.PickCompaction(
+            cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+    ASSERT_FALSE(compaction2);
+    DeleteVersionStorage();
+  }
+}
+
+}  // 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().
@@ -120,8 +120,7 @@ class UniversalCompactionBuilder {
  LogBuffer* log_buffer_;

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

  // Pick a path ID to place a newly generated file, with its estimated file
  // size.
@@ -325,8 +324,7 @@ void UniversalCompactionBuilder::SortedRun::DumpSizeInfo(

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

  if (sorted_runs_.size() == 0 ||
      (vstorage_->FilesMarkedForPeriodicCompaction().empty() &&
@@ -751,7 +737,7 @@ 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);
@@ -855,6 +841,7 @@ Compaction* UniversalCompactionBuilder::PickDeleteTriggeredCompaction() {
  std::vector<CompactionInputFiles> inputs;

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

@@ -959,8 +951,8 @@ 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 +1021,7 @@ 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 +1092,6 @@ 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);
@@ -11,7 +11,7 @@
 #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()))
@@ -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;
  }
@@ -6,6 +6,7 @@
 // Copyright (c) 2011 The LevelDB Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
+
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
 #include "rocksdb/perf_context.h"
@@ -17,7 +18,7 @@
 #include "test_util/sync_point.h"
 #endif

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBBasicTest : public DBTestBase {
 public:
@@ -27,8 +28,8 @@ class DBBasicTest : public DBTestBase {
 TEST_F(DBBasicTest, OpenWhenOpen) {
  Options options = CurrentOptions();
  options.env = env_;
-  rocksdb::DB* db2 = nullptr;
-  rocksdb::Status s = DB::Open(options, dbname_, &db2);
+  ROCKSDB_NAMESPACE::DB* db2 = nullptr;
+  ROCKSDB_NAMESPACE::Status s = DB::Open(options, dbname_, &db2);

  ASSERT_EQ(Status::Code::kIOError, s.code());
  ASSERT_EQ(Status::SubCode::kNone, s.subcode());
@@ -324,7 +325,12 @@ TEST_F(DBBasicTest, CheckLock) {
    ASSERT_OK(TryReopen(options));

    // second open should fail
-    ASSERT_TRUE(!(DB::Open(options, dbname_, &localdb)).ok());
+    Status s = DB::Open(options, dbname_, &localdb);
+    ASSERT_NOK(s);
+#ifdef OS_LINUX
+    ASSERT_TRUE(s.ToString().find("lock hold by current process") !=
+                std::string::npos);
+#endif  // OS_LINUX
  } while (ChangeCompactOptions());
 }

@@ -392,7 +398,7 @@ TEST_F(DBBasicTest, FlushEmptyColumnFamily) {
  sleeping_task_low.WaitUntilDone();
 }

-TEST_F(DBBasicTest, FLUSH) {
+TEST_F(DBBasicTest, Flush) {
  do {
    CreateAndReopenWithCF({"pikachu"}, CurrentOptions());
    WriteOptions writeOpt = WriteOptions();
@@ -919,44 +925,44 @@ TEST_F(DBBasicTest, MmapAndBufferOptions) {
 #endif

 class TestEnv : public EnvWrapper {
-  public:
-   explicit TestEnv(Env* base_env) : EnvWrapper(base_env), close_count(0) {}
+ public:
+  explicit TestEnv(Env* base_env) : EnvWrapper(base_env), close_count(0) {}

-   class TestLogger : public Logger {
-    public:
-     using Logger::Logv;
-     explicit TestLogger(TestEnv* env_ptr) : Logger() { env = env_ptr; }
-     ~TestLogger() override {
-       if (!closed_) {
-         CloseHelper();
-       }
-     }
-     void Logv(const char* /*format*/, va_list /*ap*/) override {}
-
-    protected:
-     Status CloseImpl() override { return CloseHelper(); }
-
-    private:
-     Status CloseHelper() {
-       env->CloseCountInc();
-       ;
-       return Status::IOError();
-     }
-     TestEnv* env;
-   };
-
-    void CloseCountInc() { close_count++; }
-
-    int GetCloseCount() { return close_count; }
-
-    Status NewLogger(const std::string& /*fname*/,
-                     std::shared_ptr<Logger>* result) override {
-      result->reset(new TestLogger(this));
-      return Status::OK();
+  class TestLogger : public Logger {
+   public:
+    using Logger::Logv;
+    explicit TestLogger(TestEnv* env_ptr) : Logger() { env = env_ptr; }
+    ~TestLogger() override {
+      if (!closed_) {
+        CloseHelper();
+      }
    }
+    void Logv(const char* /*format*/, va_list /*ap*/) override {}
+
+   protected:
+    Status CloseImpl() override { return CloseHelper(); }

   private:
-    int close_count;
+    Status CloseHelper() {
+      env->CloseCountInc();
+      ;
+      return Status::IOError();
+    }
+    TestEnv* env;
+  };
+
+  void CloseCountInc() { close_count++; }
+
+  int GetCloseCount() { return close_count; }
+
+  Status NewLogger(const std::string& /*fname*/,
+                   std::shared_ptr<Logger>* result) override {
+    result->reset(new TestLogger(this));
+    return Status::OK();
+  }
+
+ private:
+  int close_count;
 };

 TEST_F(DBBasicTest, DBClose) {
@@ -1008,7 +1014,7 @@ TEST_F(DBBasicTest, DBCloseFlushError) {
  Options options = GetDefaultOptions();
  options.create_if_missing = true;
  options.manual_wal_flush = true;
-  options.write_buffer_size=100;
+  options.write_buffer_size = 100;
  options.env = fault_injection_env.get();

  Reopen(options);
@@ -1048,8 +1054,8 @@ TEST_P(DBMultiGetTestWithParam, MultiGetMultiCF) {
  }

  int get_sv_count = 0;
-  rocksdb::DBImpl* db = reinterpret_cast<DBImpl*>(db_);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::DBImpl* db = reinterpret_cast<DBImpl*>(db_);
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::MultiGet::AfterRefSV", [&](void* /*arg*/) {
        if (++get_sv_count == 2) {
          // After MultiGet refs a couple of CFs, flush all CFs so MultiGet
@@ -1073,7 +1079,7 @@ TEST_P(DBMultiGetTestWithParam, MultiGetMultiCF) {
          }
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  std::vector<int> cfs;
  std::vector<std::string> keys;
@@ -1139,12 +1145,12 @@ TEST_P(DBMultiGetTestWithParam, MultiGetMultiCFMutex) {
  int get_sv_count = 0;
  int retries = 0;
  bool last_try = false;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::MultiGet::LastTry", [&](void* /*arg*/) {
        last_try = true;
-        rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+        ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
      });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::MultiGet::AfterRefSV", [&](void* /*arg*/) {
        if (last_try) {
          return;
@@ -1160,7 +1166,7 @@ TEST_P(DBMultiGetTestWithParam, MultiGetMultiCFMutex) {
          }
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  std::vector<int> cfs;
  std::vector<std::string> keys;
@@ -1198,8 +1204,8 @@ TEST_P(DBMultiGetTestWithParam, MultiGetMultiCFSnapshot) {
  }

  int get_sv_count = 0;
-  rocksdb::DBImpl* db = reinterpret_cast<DBImpl*>(db_);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::DBImpl* db = reinterpret_cast<DBImpl*>(db_);
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::MultiGet::AfterRefSV", [&](void* /*arg*/) {
        if (++get_sv_count == 2) {
          for (int i = 0; i < 8; ++i) {
@@ -1219,7 +1225,7 @@ TEST_P(DBMultiGetTestWithParam, MultiGetMultiCFSnapshot) {
          }
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  std::vector<int> cfs;
  std::vector<std::string> keys;
@@ -1289,14 +1295,18 @@ TEST_F(DBBasicTest, MultiGetBatchedSimpleUnsorted) {
  } while (ChangeCompactOptions());
 }

-TEST_F(DBBasicTest, MultiGetBatchedSimpleSorted) {
+TEST_F(DBBasicTest, MultiGetBatchedSortedMultiFile) {
  do {
    CreateAndReopenWithCF({"pikachu"}, CurrentOptions());
    SetPerfLevel(kEnableCount);
+    // To expand the power of this test, generate > 1 table file and
+    // mix with memtable
    ASSERT_OK(Put(1, "k1", "v1"));
    ASSERT_OK(Put(1, "k2", "v2"));
+    Flush(1);
    ASSERT_OK(Put(1, "k3", "v3"));
    ASSERT_OK(Put(1, "k4", "v4"));
+    Flush(1);
    ASSERT_OK(Delete(1, "k4"));
    ASSERT_OK(Put(1, "k5", "v5"));
    ASSERT_OK(Delete(1, "no_key"));
@@ -1327,7 +1337,7 @@ TEST_F(DBBasicTest, MultiGetBatchedSimpleSorted) {
    ASSERT_TRUE(s[5].IsNotFound());

    SetPerfLevel(kDisable);
-  } while (ChangeCompactOptions());
+  } while (ChangeOptions());
 }

 TEST_F(DBBasicTest, MultiGetBatchedMultiLevel) {
@@ -1405,6 +1415,92 @@ TEST_F(DBBasicTest, MultiGetBatchedMultiLevel) {
  }
 }

+TEST_F(DBBasicTest, MultiGetBatchedMultiLevelMerge) {
+  Options options = CurrentOptions();
+  options.disable_auto_compactions = true;
+  options.merge_operator = MergeOperators::CreateStringAppendOperator();
+  BlockBasedTableOptions bbto;
+  bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  Reopen(options);
+  int num_keys = 0;
+
+  for (int i = 0; i < 128; ++i) {
+    ASSERT_OK(Put("key_" + std::to_string(i), "val_l2_" + std::to_string(i)));
+    num_keys++;
+    if (num_keys == 8) {
+      Flush();
+      num_keys = 0;
+    }
+  }
+  if (num_keys > 0) {
+    Flush();
+    num_keys = 0;
+  }
+  MoveFilesToLevel(2);
+
+  for (int i = 0; i < 128; i += 3) {
+    ASSERT_OK(Merge("key_" + std::to_string(i), "val_l1_" + std::to_string(i)));
+    num_keys++;
+    if (num_keys == 8) {
+      Flush();
+      num_keys = 0;
+    }
+  }
+  if (num_keys > 0) {
+    Flush();
+    num_keys = 0;
+  }
+  MoveFilesToLevel(1);
+
+  for (int i = 0; i < 128; i += 5) {
+    ASSERT_OK(Merge("key_" + std::to_string(i), "val_l0_" + std::to_string(i)));
+    num_keys++;
+    if (num_keys == 8) {
+      Flush();
+      num_keys = 0;
+    }
+  }
+  if (num_keys > 0) {
+    Flush();
+    num_keys = 0;
+  }
+  ASSERT_EQ(0, num_keys);
+
+  for (int i = 0; i < 128; i += 9) {
+    ASSERT_OK(
+        Merge("key_" + std::to_string(i), "val_mem_" + std::to_string(i)));
+  }
+
+  std::vector<std::string> keys;
+  std::vector<std::string> values;
+
+  for (int i = 32; i < 80; ++i) {
+    keys.push_back("key_" + std::to_string(i));
+  }
+
+  values = MultiGet(keys, nullptr);
+  ASSERT_EQ(values.size(), keys.size());
+  for (unsigned int j = 0; j < 48; ++j) {
+    int key = j + 32;
+    std::string value;
+    value.append("val_l2_" + std::to_string(key));
+    if (key % 3 == 0) {
+      value.append(",");
+      value.append("val_l1_" + std::to_string(key));
+    }
+    if (key % 5 == 0) {
+      value.append(",");
+      value.append("val_l0_" + std::to_string(key));
+    }
+    if (key % 9 == 0) {
+      value.append(",");
+      value.append("val_mem_" + std::to_string(key));
+    }
+    ASSERT_EQ(values[j], value);
+  }
+}
+
 // Test class for batched MultiGet with prefix extractor
 // Param bool - If true, use partitioned filters
 //              If false, use full filter block
@@ -1473,7 +1569,7 @@ TEST_P(DBMultiGetRowCacheTest, MultiGetBatched) {
  do {
    option_config_ = kRowCache;
    Options options = CurrentOptions();
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
    CreateAndReopenWithCF({"pikachu"}, options);
    SetPerfLevel(kEnableCount);
    ASSERT_OK(Put(1, "k1", "v1"));
@@ -1579,13 +1675,13 @@ TEST_F(DBBasicTest, GetAllKeyVersions) {
    ASSERT_OK(Delete(std::to_string(i)));
  }
  std::vector<KeyVersion> key_versions;
-  ASSERT_OK(rocksdb::GetAllKeyVersions(db_, Slice(), Slice(),
-                                       std::numeric_limits<size_t>::max(),
-                                       &key_versions));
+  ASSERT_OK(ROCKSDB_NAMESPACE::GetAllKeyVersions(
+      db_, Slice(), Slice(), std::numeric_limits<size_t>::max(),
+      &key_versions));
  ASSERT_EQ(kNumInserts + kNumDeletes + kNumUpdates, key_versions.size());
-  ASSERT_OK(rocksdb::GetAllKeyVersions(db_, handles_[0], Slice(), Slice(),
-                                       std::numeric_limits<size_t>::max(),
-                                       &key_versions));
+  ASSERT_OK(ROCKSDB_NAMESPACE::GetAllKeyVersions(
+      db_, handles_[0], Slice(), Slice(), std::numeric_limits<size_t>::max(),
+      &key_versions));
  ASSERT_EQ(kNumInserts + kNumDeletes + kNumUpdates, key_versions.size());

  // Check non-default column family
@@ -1598,9 +1694,9 @@ TEST_F(DBBasicTest, GetAllKeyVersions) {
  for (size_t i = 0; i != kNumDeletes - 1; ++i) {
    ASSERT_OK(Delete(1, std::to_string(i)));
  }
-  ASSERT_OK(rocksdb::GetAllKeyVersions(db_, handles_[1], Slice(), Slice(),
-                                       std::numeric_limits<size_t>::max(),
-                                       &key_versions));
+  ASSERT_OK(ROCKSDB_NAMESPACE::GetAllKeyVersions(
+      db_, handles_[1], Slice(), Slice(), std::numeric_limits<size_t>::max(),
+      &key_versions));
  ASSERT_EQ(kNumInserts + kNumDeletes + kNumUpdates - 3, key_versions.size());
 }
 #endif  // !ROCKSDB_LITE
@@ -1611,8 +1707,8 @@ TEST_F(DBBasicTest, MultiGetIOBufferOverrun) {
  BlockBasedTableOptions table_options;
  table_options.pin_l0_filter_and_index_blocks_in_cache = true;
  table_options.block_size = 16 * 1024;
-  assert(table_options.block_size >
-          BlockBasedTable::kMultiGetReadStackBufSize);
+  ASSERT_TRUE(table_options.block_size >
+            BlockBasedTable::kMultiGetReadStackBufSize);
  options.table_factory.reset(new BlockBasedTableFactory(table_options));
  Reopen(options);

@@ -1644,6 +1740,205 @@ TEST_F(DBBasicTest, MultiGetIOBufferOverrun) {
                     keys.data(), values.data(), statuses.data(), true);
 }

+TEST_F(DBBasicTest, IncrementalRecoveryNoCorrupt) {
+  Options options = CurrentOptions();
+  DestroyAndReopen(options);
+  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
+  size_t num_cfs = handles_.size();
+  ASSERT_EQ(3, num_cfs);
+  WriteOptions write_opts;
+  write_opts.disableWAL = true;
+  for (size_t cf = 0; cf != num_cfs; ++cf) {
+    for (size_t i = 0; i != 10000; ++i) {
+      std::string key_str = Key(static_cast<int>(i));
+      std::string value_str = std::to_string(cf) + "_" + std::to_string(i);
+
+      ASSERT_OK(Put(static_cast<int>(cf), key_str, value_str));
+      if (0 == (i % 1000)) {
+        ASSERT_OK(Flush(static_cast<int>(cf)));
+      }
+    }
+  }
+  for (size_t cf = 0; cf != num_cfs; ++cf) {
+    ASSERT_OK(Flush(static_cast<int>(cf)));
+  }
+  Close();
+  options.best_efforts_recovery = true;
+  ReopenWithColumnFamilies({kDefaultColumnFamilyName, "pikachu", "eevee"},
+                           options);
+  num_cfs = handles_.size();
+  ASSERT_EQ(3, num_cfs);
+  for (size_t cf = 0; cf != num_cfs; ++cf) {
+    for (int i = 0; i != 10000; ++i) {
+      std::string key_str = Key(static_cast<int>(i));
+      std::string expected_value_str =
+          std::to_string(cf) + "_" + std::to_string(i);
+      ASSERT_EQ(expected_value_str, Get(static_cast<int>(cf), key_str));
+    }
+  }
+}
+
+TEST_F(DBBasicTest, BestEffortsRecoveryWithVersionBuildingFailure) {
+  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.best_efforts_recovery = true;
+  Status s = TryReopen(options);
+  ASSERT_TRUE(s.IsCorruption());
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+#ifndef ROCKSDB_LITE
+namespace {
+class TableFileListener : public EventListener {
+ public:
+  void OnTableFileCreated(const TableFileCreationInfo& info) override {
+    InstrumentedMutexLock lock(&mutex_);
+    cf_to_paths_[info.cf_name].push_back(info.file_path);
+  }
+  std::vector<std::string>& GetFiles(const std::string& cf_name) {
+    InstrumentedMutexLock lock(&mutex_);
+    return cf_to_paths_[cf_name];
+  }
+
+ private:
+  InstrumentedMutex mutex_;
+  std::unordered_map<std::string, std::vector<std::string>> cf_to_paths_;
+};
+}  // namespace
+
+TEST_F(DBBasicTest, RecoverWithMissingFiles) {
+  Options options = CurrentOptions();
+  DestroyAndReopen(options);
+  TableFileListener* listener = new TableFileListener();
+  // Disable auto compaction to simplify SST file name tracking.
+  options.disable_auto_compactions = true;
+  options.listeners.emplace_back(listener);
+  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
+  std::vector<std::string> all_cf_names = {kDefaultColumnFamilyName, "pikachu",
+                                           "eevee"};
+  size_t num_cfs = handles_.size();
+  ASSERT_EQ(3, num_cfs);
+  for (size_t cf = 0; cf != num_cfs; ++cf) {
+    ASSERT_OK(Put(static_cast<int>(cf), "a", "0_value"));
+    ASSERT_OK(Flush(static_cast<int>(cf)));
+    ASSERT_OK(Put(static_cast<int>(cf), "b", "0_value"));
+    ASSERT_OK(Flush(static_cast<int>(cf)));
+    ASSERT_OK(Put(static_cast<int>(cf), "c", "0_value"));
+    ASSERT_OK(Flush(static_cast<int>(cf)));
+  }
+
+  // Delete files
+  for (size_t i = 0; i < all_cf_names.size(); ++i) {
+    std::vector<std::string>& files = listener->GetFiles(all_cf_names[i]);
+    ASSERT_EQ(3, files.size());
+    for (int j = static_cast<int>(files.size() - 1); j >= static_cast<int>(i);
+         --j) {
+      ASSERT_OK(env_->DeleteFile(files[j]));
+    }
+  }
+  options.best_efforts_recovery = true;
+  ReopenWithColumnFamilies(all_cf_names, options);
+  // Verify data
+  ReadOptions read_opts;
+  read_opts.total_order_seek = true;
+  {
+    std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts, handles_[0]));
+    iter->SeekToFirst();
+    ASSERT_FALSE(iter->Valid());
+    iter.reset(db_->NewIterator(read_opts, handles_[1]));
+    iter->SeekToFirst();
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_EQ("a", iter->key());
+    iter->Next();
+    ASSERT_FALSE(iter->Valid());
+    iter.reset(db_->NewIterator(read_opts, handles_[2]));
+    iter->SeekToFirst();
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_EQ("a", iter->key());
+    iter->Next();
+    ASSERT_TRUE(iter->Valid());
+    ASSERT_EQ("b", iter->key());
+    iter->Next();
+    ASSERT_FALSE(iter->Valid());
+  }
+}
+
+TEST_F(DBBasicTest, RecoverWithNoCurrentFile) {
+  Options options = CurrentOptions();
+  options.env = env_;
+  DestroyAndReopen(options);
+  CreateAndReopenWithCF({"pikachu"}, options);
+  options.best_efforts_recovery = true;
+  ReopenWithColumnFamilies({kDefaultColumnFamilyName, "pikachu"}, options);
+  ASSERT_EQ(2, handles_.size());
+  ASSERT_OK(Put("foo", "value"));
+  ASSERT_OK(Put(1, "bar", "value"));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Flush(1));
+  Close();
+  ASSERT_OK(env_->DeleteFile(CurrentFileName(dbname_)));
+  ReopenWithColumnFamilies({kDefaultColumnFamilyName, "pikachu"}, options);
+  std::vector<std::string> cf_names;
+  ASSERT_OK(DB::ListColumnFamilies(DBOptions(options), dbname_, &cf_names));
+  ASSERT_EQ(2, cf_names.size());
+  for (const auto& name : cf_names) {
+    ASSERT_TRUE(name == kDefaultColumnFamilyName || name == "pikachu");
+  }
+}
+
+TEST_F(DBBasicTest, SkipWALIfMissingTableFiles) {
+  Options options = CurrentOptions();
+  DestroyAndReopen(options);
+  TableFileListener* listener = new TableFileListener();
+  options.listeners.emplace_back(listener);
+  CreateAndReopenWithCF({"pikachu"}, options);
+  std::vector<std::string> kAllCfNames = {kDefaultColumnFamilyName, "pikachu"};
+  size_t num_cfs = handles_.size();
+  ASSERT_EQ(2, num_cfs);
+  for (int cf = 0; cf < static_cast<int>(kAllCfNames.size()); ++cf) {
+    ASSERT_OK(Put(cf, "a", "0_value"));
+    ASSERT_OK(Flush(cf));
+    ASSERT_OK(Put(cf, "b", "0_value"));
+  }
+  // Delete files
+  for (size_t i = 0; i < kAllCfNames.size(); ++i) {
+    std::vector<std::string>& files = listener->GetFiles(kAllCfNames[i]);
+    ASSERT_EQ(1, files.size());
+    for (int j = static_cast<int>(files.size() - 1); j >= static_cast<int>(i);
+         --j) {
+      ASSERT_OK(env_->DeleteFile(files[j]));
+    }
+  }
+  options.best_efforts_recovery = true;
+  ReopenWithColumnFamilies(kAllCfNames, options);
+  // Verify WAL is not applied
+  ReadOptions read_opts;
+  read_opts.total_order_seek = true;
+  std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts, handles_[0]));
+  iter->SeekToFirst();
+  ASSERT_FALSE(iter->Valid());
+  iter.reset(db_->NewIterator(read_opts, handles_[1]));
+  iter->SeekToFirst();
+  ASSERT_TRUE(iter->Valid());
+  ASSERT_EQ("a", iter->key());
+  iter->Next();
+  ASSERT_FALSE(iter->Valid());
+}
+#endif  // !ROCKSDB_LITE
+
 class DBBasicTestWithParallelIO
    : public DBTestBase,
      public testing::WithParamInterface<std::tuple<bool, bool, bool, bool>> {
@@ -1675,10 +1970,17 @@ class DBBasicTestWithParallelIO
      compression_types = GetSupportedCompressions();
      // Not every platform may have compression libraries available, so
      // dynamically pick based on what's available
-      if (compression_types.size() == 0) {
-        compression_enabled_ = false;
+      CompressionType tmp_type = kNoCompression;
+      for (auto c_type : compression_types) {
+        if (c_type != kNoCompression) {
+          tmp_type = c_type;
+          break;
+        }
+      }
+      if (tmp_type != kNoCompression) {
+        options.compression = tmp_type;
      } else {
-        options.compression = compression_types[0];
+        compression_enabled_ = false;
      }
    }
 #else
@@ -1686,7 +1988,7 @@ class DBBasicTestWithParallelIO
    if (!Snappy_Supported()) {
      compression_enabled_ = false;
    }
-#endif //ROCKSDB_LITE
+#endif  // ROCKSDB_LITE

    table_options.block_cache = uncompressed_cache_;
    if (table_options.block_cache == nullptr) {
@@ -2011,200 +2313,102 @@ TEST_P(DBBasicTestWithParallelIO, MultiGet) {
  }
 }

-INSTANTIATE_TEST_CASE_P(
-    ParallelIO, DBBasicTestWithParallelIO,
-    // Params are as follows -
-    // Param 0 - Compressed cache enabled
-    // Param 1 - Uncompressed cache enabled
-    // Param 2 - Data compression enabled
-    // Param 3 - ReadOptions::fill_cache
-    ::testing::Combine(::testing::Bool(), ::testing::Bool(),
-                       ::testing::Bool(), ::testing::Bool()));
+TEST_P(DBBasicTestWithParallelIO, MultiGetWithChecksumMismatch) {
+  std::vector<std::string> key_data(10);
+  std::vector<Slice> keys;
+  // We cannot resize a PinnableSlice vector, so just set initial size to
+  // largest we think we will need
+  std::vector<PinnableSlice> values(10);
+  std::vector<Status> statuses;
+  int read_count = 0;
+  ReadOptions ro;
+  ro.fill_cache = fill_cache();

-class DBBasicTestWithTimestampWithParam
-    : public DBTestBase,
-      public testing::WithParamInterface<bool> {
- public:
-  DBBasicTestWithTimestampWithParam()
-      : DBTestBase("/db_basic_test_with_timestamp") {}
-
- protected:
-  class TestComparator : public Comparator {
-   private:
-    const Comparator* cmp_without_ts_;
-
-   public:
-    explicit TestComparator(size_t ts_sz)
-        : Comparator(ts_sz), cmp_without_ts_(nullptr) {
-      cmp_without_ts_ = BytewiseComparator();
-    }
-
-    const char* Name() const override { return "TestComparator"; }
-
-    void FindShortSuccessor(std::string*) const override {}
-
-    void FindShortestSeparator(std::string*, const Slice&) const override {}
-
-    int Compare(const Slice& a, const Slice& b) const override {
-      int r = CompareWithoutTimestamp(a, b);
-      if (r != 0 || 0 == timestamp_size()) {
-        return r;
-      }
-      return CompareTimestamp(
-          Slice(a.data() + a.size() - timestamp_size(), timestamp_size()),
-          Slice(b.data() + b.size() - timestamp_size(), timestamp_size()));
-    }
-
-    int CompareWithoutTimestamp(const Slice& a, const Slice& b) const override {
-      assert(a.size() >= timestamp_size());
-      assert(b.size() >= timestamp_size());
-      Slice k1 = StripTimestampFromUserKey(a, timestamp_size());
-      Slice k2 = StripTimestampFromUserKey(b, timestamp_size());
-
-      return cmp_without_ts_->Compare(k1, k2);
-    }
-
-    int CompareTimestamp(const Slice& ts1, const Slice& ts2) const override {
-      if (!ts1.data() && !ts2.data()) {
-        return 0;
-      } else if (ts1.data() && !ts2.data()) {
-        return 1;
-      } else if (!ts1.data() && ts2.data()) {
-        return -1;
-      }
-      assert(ts1.size() == ts2.size());
-      uint64_t low1 = 0;
-      uint64_t low2 = 0;
-      uint64_t high1 = 0;
-      uint64_t high2 = 0;
-      auto* ptr1 = const_cast<Slice*>(&ts1);
-      auto* ptr2 = const_cast<Slice*>(&ts2);
-      if (!GetFixed64(ptr1, &low1) || !GetFixed64(ptr1, &high1) ||
-          !GetFixed64(ptr2, &low2) || !GetFixed64(ptr2, &high2)) {
-        assert(false);
-      }
-      if (high1 < high2) {
-        return 1;
-      } else if (high1 > high2) {
-        return -1;
-      }
-      if (low1 < low2) {
-        return 1;
-      } else if (low1 > low2) {
-        return -1;
-      }
-      return 0;
-    }
-  };
-
-  Slice EncodeTimestamp(uint64_t low, uint64_t high, std::string* ts) {
-    assert(nullptr != ts);
-    ts->clear();
-    PutFixed64(ts, low);
-    PutFixed64(ts, high);
-    assert(ts->size() == sizeof(low) + sizeof(high));
-    return Slice(*ts);
-  }
-};
-
-TEST_P(DBBasicTestWithTimestampWithParam, PutAndGet) {
-  const int kNumKeysPerFile = 8192;
-  const size_t kNumTimestamps = 6;
-  bool memtable_only = GetParam();
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.env = env_;
-  options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeysPerFile));
-  std::string tmp;
-  size_t ts_sz = EncodeTimestamp(0, 0, &tmp).size();
-  TestComparator test_cmp(ts_sz);
-  options.comparator = &test_cmp;
-  BlockBasedTableOptions bbto;
-  bbto.filter_policy.reset(NewBloomFilterPolicy(
-      10 /*bits_per_key*/, false /*use_block_based_builder*/));
-  bbto.whole_key_filtering = true;
-  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
-
-  std::vector<CompressionType> compression_types;
-  compression_types.push_back(kNoCompression);
-  if (Zlib_Supported()) {
-    compression_types.push_back(kZlibCompression);
-  }
-#if LZ4_VERSION_NUMBER >= 10400  // r124+
-  compression_types.push_back(kLZ4Compression);
-  compression_types.push_back(kLZ4HCCompression);
-#endif  // LZ4_VERSION_NUMBER >= 10400
-  if (ZSTD_Supported()) {
-    compression_types.push_back(kZSTD);
-  }
-
-  // Switch compression dictionary on/off to check key extraction
-  // correctness in kBuffered state
-  std::vector<uint32_t> max_dict_bytes_list = {0, 1 << 14};  // 0 or 16KB
-
-  for (auto compression_type : compression_types) {
-    for (uint32_t max_dict_bytes : max_dict_bytes_list) {
-      options.compression = compression_type;
-      options.compression_opts.max_dict_bytes = max_dict_bytes;
-      if (compression_type == kZSTD) {
-        options.compression_opts.zstd_max_train_bytes = max_dict_bytes;
-      }
-      options.target_file_size_base = 1 << 26;  // 64MB
-
-      DestroyAndReopen(options);
-      CreateAndReopenWithCF({"pikachu"}, options);
-      size_t num_cfs = handles_.size();
-      ASSERT_EQ(2, num_cfs);
-      std::vector<std::string> write_ts_strs(kNumTimestamps);
-      std::vector<std::string> read_ts_strs(kNumTimestamps);
-      std::vector<Slice> write_ts_list;
-      std::vector<Slice> read_ts_list;
-
-      for (size_t i = 0; i != kNumTimestamps; ++i) {
-        write_ts_list.emplace_back(
-            EncodeTimestamp(i * 2, 0, &write_ts_strs[i]));
-        read_ts_list.emplace_back(
-            EncodeTimestamp(1 + i * 2, 0, &read_ts_strs[i]));
-        const Slice& write_ts = write_ts_list.back();
-        WriteOptions wopts;
-        wopts.timestamp = &write_ts;
-        for (int cf = 0; cf != static_cast<int>(num_cfs); ++cf) {
-          for (size_t j = 0; j != (kNumKeysPerFile - 1) / kNumTimestamps; ++j) {
-            ASSERT_OK(Put(
-                cf, "key" + std::to_string(j),
-                "value_" + std::to_string(j) + "_" + std::to_string(i), wopts));
-          }
-          if (!memtable_only) {
-            ASSERT_OK(Flush(cf));
-          }
+  SyncPoint::GetInstance()->SetCallBack(
+      "RetrieveMultipleBlocks:VerifyChecksum", [&](void* status) {
+        Status* s = static_cast<Status*>(status);
+        read_count++;
+        if (read_count == 2) {
+          *s = Status::Corruption();
        }
-      }
-      const auto& verify_db_func = [&]() {
-        for (size_t i = 0; i != kNumTimestamps; ++i) {
-          ReadOptions ropts;
-          ropts.timestamp = &read_ts_list[i];
-          for (int cf = 0; cf != static_cast<int>(num_cfs); ++cf) {
-            ColumnFamilyHandle* cfh = handles_[cf];
-            for (size_t j = 0; j != (kNumKeysPerFile - 1) / kNumTimestamps;
-                 ++j) {
-              std::string value;
-              ASSERT_OK(
-                  db_->Get(ropts, cfh, "key" + std::to_string(j), &value));
-              ASSERT_EQ("value_" + std::to_string(j) + "_" + std::to_string(i),
-                        value);
-            }
-          }
-        }
-      };
-      verify_db_func();
-    }
-  }
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  // Warm up the cache first
+  key_data.emplace_back(Key(0));
+  keys.emplace_back(Slice(key_data.back()));
+  key_data.emplace_back(Key(50));
+  keys.emplace_back(Slice(key_data.back()));
+  statuses.resize(keys.size());
+
+  dbfull()->MultiGet(ro, dbfull()->DefaultColumnFamily(), keys.size(),
+                     keys.data(), values.data(), statuses.data(), true);
+  ASSERT_TRUE(CheckValue(0, values[0].ToString()));
+  // ASSERT_TRUE(CheckValue(50, values[1].ToString()));
+  ASSERT_EQ(statuses[0], Status::OK());
+  ASSERT_EQ(statuses[1], Status::Corruption());
+
+  SyncPoint::GetInstance()->DisableProcessing();
 }

-INSTANTIATE_TEST_CASE_P(Timestamp, DBBasicTestWithTimestampWithParam,
-                        ::testing::Bool());
+TEST_P(DBBasicTestWithParallelIO, MultiGetWithMissingFile) {
+  std::vector<std::string> key_data(10);
+  std::vector<Slice> keys;
+  // We cannot resize a PinnableSlice vector, so just set initial size to
+  // largest we think we will need
+  std::vector<PinnableSlice> values(10);
+  std::vector<Status> statuses;
+  ReadOptions ro;
+  ro.fill_cache = fill_cache();

-}  // namespace rocksdb
+  SyncPoint::GetInstance()->SetCallBack(
+      "TableCache::MultiGet:FindTable", [&](void* status) {
+        Status* s = static_cast<Status*>(status);
+        *s = Status::IOError();
+      });
+  // DB open will create table readers unless we reduce the table cache
+  // capacity.
+  // SanitizeOptions will set max_open_files to minimum of 20. Table cache
+  // is allocated with max_open_files - 10 as capacity. So override
+  // max_open_files to 11 so table cache capacity will become 1. This will
+  // prevent file open during DB open and force the file to be opened
+  // during MultiGet
+  SyncPoint::GetInstance()->SetCallBack(
+      "SanitizeOptions::AfterChangeMaxOpenFiles", [&](void* arg) {
+        int* max_open_files = (int*)arg;
+        *max_open_files = 11;
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  Reopen(CurrentOptions());
+
+  // Warm up the cache first
+  key_data.emplace_back(Key(0));
+  keys.emplace_back(Slice(key_data.back()));
+  key_data.emplace_back(Key(50));
+  keys.emplace_back(Slice(key_data.back()));
+  statuses.resize(keys.size());
+
+  dbfull()->MultiGet(ro, dbfull()->DefaultColumnFamily(), keys.size(),
+                     keys.data(), values.data(), statuses.data(), true);
+  ASSERT_EQ(statuses[0], Status::IOError());
+  ASSERT_EQ(statuses[1], Status::IOError());
+
+  SyncPoint::GetInstance()->DisableProcessing();
+}
+
+INSTANTIATE_TEST_CASE_P(ParallelIO, DBBasicTestWithParallelIO,
+                        // Params are as follows -
+                        // Param 0 - Compressed cache enabled
+                        // Param 1 - Uncompressed cache enabled
+                        // Param 2 - Data compression enabled
+                        // Param 3 - ReadOptions::fill_cache
+                        ::testing::Combine(::testing::Bool(), ::testing::Bool(),
+                                           ::testing::Bool(),
+                                           ::testing::Bool()));
+
+
+}  // namespace ROCKSDB_NAMESPACE

 #ifdef ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS
 extern "C" {
@@ -2215,7 +2419,7 @@ 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();
@@ -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());
@@ -520,7 +520,7 @@ TEST_F(DBBlockCacheTest, IndexAndFilterBlocksCachePriority) {
 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 +576,7 @@ 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 +705,7 @@ 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 +752,10 @@ 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();
 }
@@ -12,16 +12,10 @@
 #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 +88,7 @@ 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 +150,7 @@ 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 +216,7 @@ 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 +267,7 @@ 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 +531,7 @@ 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 +563,7 @@ 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 +607,7 @@ 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 +654,18 @@ 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 +676,16 @@ 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 +801,7 @@ 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 +900,7 @@ 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 +929,7 @@ 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 +968,8 @@ 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);
@@ -1031,6 +1030,14 @@ TEST_F(DBBloomFilterTest, MemtablePrefixBloomOutOfDomain) {
 }

 #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 +1047,8 @@ 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 +1329,7 @@ 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);
@@ -1447,13 +1455,13 @@ 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 +1901,10 @@ 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") {}
@@ -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_);
@@ -172,6 +172,6 @@ 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));
      }
@@ -890,9 +894,9 @@ 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);
  }
@@ -1005,12 +1009,36 @@ Status DBImpl::SetDBOptions(
      }
    }
    if (s.ok()) {
-      if (new_options.max_background_compactions >
-          mutable_db_options_.max_background_compactions) {
-        env_->IncBackgroundThreadsIfNeeded(
-            new_options.max_background_compactions, Env::Priority::LOW);
+      const BGJobLimits current_bg_job_limits =
+          GetBGJobLimits(immutable_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_compactions,
+          new_options.max_background_jobs, /* parallelize_compactions */ true);
+
+      const bool max_flushes_increased =
+          new_bg_job_limits.max_flushes > current_bg_job_limits.max_flushes;
+      const bool max_compactions_increased =
+          new_bg_job_limits.max_compactions >
+          current_bg_job_limits.max_compactions;
+
+      if (max_flushes_increased || max_compactions_increased) {
+        if (max_flushes_increased) {
+          env_->IncBackgroundThreadsIfNeeded(new_bg_job_limits.max_flushes,
+                                             Env::Priority::HIGH);
+        }
+
+        if (max_compactions_increased) {
+          env_->IncBackgroundThreadsIfNeeded(new_bg_job_limits.max_compactions,
+                                             Env::Priority::LOW);
+        }
+
        MaybeScheduleFlushOrCompaction();
      }
+
      if (new_options.stats_dump_period_sec !=
          mutable_db_options_.stats_dump_period_sec) {
        if (thread_dump_stats_) {
@@ -1019,7 +1047,7 @@ 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));
@@ -1035,7 +1063,7 @@ 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));
@@ -1123,25 +1151,25 @@ 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) {
@@ -1193,14 +1221,23 @@ 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");

@@ -1225,7 +1262,7 @@ Status DBImpl::LockWAL() {
    // future writes
    WriteStatusCheck(status);
  }
-  return status;
+  return std::move(status);
 }

 Status DBImpl::UnlockWAL() {
@@ -1472,16 +1509,37 @@ 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);
@@ -1559,10 +1617,14 @@ 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)) {
@@ -1570,9 +1632,10 @@ 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();
@@ -1581,9 +1644,9 @@ 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()) &&
@@ -1602,8 +1665,8 @@ 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,
@@ -1651,12 +1714,34 @@ 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());
@@ -1687,6 +1772,9 @@ 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;
@@ -1701,8 +1789,9 @@ std::vector<Status> DBImpl::MultiGet(
    merge_context.Clear();
    Status& s = stat_list[i];
    std::string* value = &(*values)[i];
+    std::string* timestamp = timestamps ? &(*timestamps)[i] : nullptr;

-    LookupKey lkey(keys[i], consistent_seqnum);
+    LookupKey lkey(keys[i], consistent_seqnum, read_options.timestamp);
    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family[i]);
    SequenceNumber max_covering_tombstone_seq = 0;
    auto mgd_iter = multiget_cf_data.find(cfh->cfd()->GetID());
@@ -1714,13 +1803,13 @@ 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);
      }
@@ -1728,8 +1817,9 @@ 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);
    }
@@ -1891,14 +1981,39 @@ 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) {
@@ -1969,7 +2084,8 @@ 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;
    }
@@ -2001,7 +2117,8 @@ 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++;
@@ -2019,11 +2136,22 @@ 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];
@@ -2101,7 +2229,7 @@ void DBImpl::MultiGetImpl(
                            ? 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;

@@ -2277,7 +2405,8 @@ 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;
@@ -2411,7 +2540,8 @@ 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
@@ -2424,6 +2554,7 @@ 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());

@@ -2536,7 +2667,7 @@ 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(),
@@ -3293,27 +3424,67 @@ 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 {
@@ -3329,31 +3500,12 @@ 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();
@@ -3416,12 +3568,8 @@ 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() {}
@@ -3465,24 +3613,15 @@ 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) {
@@ -3545,6 +3684,11 @@ 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;
@@ -3768,8 +3912,9 @@ 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.
@@ -3794,8 +3939,9 @@ 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.
@@ -3820,7 +3966,7 @@ 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);

@@ -3848,7 +3994,7 @@ 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);
@@ -4325,8 +4471,8 @@ 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()) {
@@ -4471,13 +4617,21 @@ 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;
@@ -4488,4 +4642,4 @@ 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,
@@ -826,9 +846,9 @@ 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)
@@ -1110,13 +1130,25 @@ 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 +1362,7 @@ 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 +1387,10 @@ 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 +1413,7 @@ 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 +1485,10 @@ 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 +1499,25 @@ 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 +1632,7 @@ 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();

@@ -2010,13 +2049,13 @@ 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 +2138,4 @@ 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;
    }
@@ -2604,6 +2666,7 @@ 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 +2688,11 @@ 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 +2733,8 @@ 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 +2746,11 @@ 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 +2837,7 @@ 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 +2847,11 @@ 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 +2877,11 @@ 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 +3188,4 @@ 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()) {
@@ -618,9 +619,9 @@ 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 +665,69 @@ 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);
@@ -218,4 +220,4 @@ 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);
  }
  {
@@ -468,6 +470,11 @@ 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 +670,4 @@ 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() {
@@ -1292,7 +1320,7 @@ Status DBImpl::HandleWriteBufferFull(WriteContext* write_context) {
  // suboptimal but still correct.
  ROCKS_LOG_INFO(
      immutable_db_options_.info_log,
-      "Flushing column family with largest mem table size. Write buffer is "
+      "Flushing column family with oldest memtable entry. Write buffer is "
      "using %" ROCKSDB_PRIszt " bytes out of a total of %" ROCKSDB_PRIszt ".",
      write_buffer_manager_->memory_usage(),
      write_buffer_manager_->buffer_size());
@@ -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();
 }
@@ -14,7 +14,7 @@
 #include "file/filename.h"
 #include "rocksdb/env.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 void DumpDBFileSummary(const ImmutableDBOptions& options,
                       const std::string& dbname) {
@@ -120,4 +120,4 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
  Header(options.info_log, "Write Ahead Log file in %s: %s\n",
         options.wal_dir.c_str(), wal_info.c_str());
 }
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -8,7 +8,7 @@

 #include "options/db_options.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {
 void DumpDBFileSummary(const ImmutableDBOptions& options,
                       const std::string& dbname);
-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -9,7 +9,7 @@
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBTestInPlaceUpdate : public DBTestBase {
 public:
@@ -73,7 +73,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerSize) {
    options.env = env_;
    options.write_buffer_size = 100000;
    options.inplace_callback =
-      rocksdb::DBTestInPlaceUpdate::updateInPlaceSmallerSize;
+        ROCKSDB_NAMESPACE::DBTestInPlaceUpdate::updateInPlaceSmallerSize;
    options.allow_concurrent_memtable_write = false;
    Reopen(options);
    CreateAndReopenWithCF({"pikachu"}, options);
@@ -102,7 +102,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerVarintSize) {
    options.env = env_;
    options.write_buffer_size = 100000;
    options.inplace_callback =
-      rocksdb::DBTestInPlaceUpdate::updateInPlaceSmallerVarintSize;
+        ROCKSDB_NAMESPACE::DBTestInPlaceUpdate::updateInPlaceSmallerVarintSize;
    options.allow_concurrent_memtable_write = false;
    Reopen(options);
    CreateAndReopenWithCF({"pikachu"}, options);
@@ -131,7 +131,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackLargeNewValue) {
    options.env = env_;
    options.write_buffer_size = 100000;
    options.inplace_callback =
-      rocksdb::DBTestInPlaceUpdate::updateInPlaceLargerSize;
+        ROCKSDB_NAMESPACE::DBTestInPlaceUpdate::updateInPlaceLargerSize;
    options.allow_concurrent_memtable_write = false;
    Reopen(options);
    CreateAndReopenWithCF({"pikachu"}, options);
@@ -158,7 +158,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackNoAction) {
    options.env = env_;
    options.write_buffer_size = 100000;
    options.inplace_callback =
-        rocksdb::DBTestInPlaceUpdate::updateInPlaceNoAction;
+        ROCKSDB_NAMESPACE::DBTestInPlaceUpdate::updateInPlaceNoAction;
    options.allow_concurrent_memtable_write = false;
    Reopen(options);
    CreateAndReopenWithCF({"pikachu"}, options);
@@ -168,10 +168,10 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackNoAction) {
    ASSERT_EQ(Get(1, "key"), "NOT_FOUND");
  } while (ChangeCompactOptions());
 }
-}  // 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 {

 class DBIOFailureTest : public DBTestBase {
 public:
@@ -270,14 +270,14 @@ TEST_F(DBIOFailureTest, FlushSstRangeSyncError) {
  Status s;

  std::atomic<int> range_sync_called(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "SpecialEnv::SStableFile::RangeSync", [&](void* arg) {
        if (range_sync_called.fetch_add(1) == 0) {
          Status* st = static_cast<Status*>(arg);
          *st = Status::IOError("range sync dummy error");
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  Random rnd(301);
  std::string rnd_str =
@@ -302,7 +302,7 @@ TEST_F(DBIOFailureTest, FlushSstRangeSyncError) {
  ASSERT_NOK(Put(1, "foo2", "bar3"));
  ASSERT_EQ("bar", Get(1, "foo"));

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_GE(1, range_sync_called.load());

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
@@ -350,14 +350,14 @@ TEST_F(DBIOFailureTest, CompactSstRangeSyncError) {
  dbfull()->TEST_WaitForFlushMemTable(handles_[1]);

  std::atomic<int> range_sync_called(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "SpecialEnv::SStableFile::RangeSync", [&](void* arg) {
        if (range_sync_called.fetch_add(1) == 0) {
          Status* st = static_cast<Status*>(arg);
          *st = Status::IOError("range sync dummy error");
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  ASSERT_OK(dbfull()->SetOptions(handles_[1],
                                 {
@@ -369,7 +369,7 @@ TEST_F(DBIOFailureTest, CompactSstRangeSyncError) {
  ASSERT_NOK(Put(1, "foo2", "bar3"));
  ASSERT_EQ("bar", Get(1, "foo"));

-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
  ASSERT_GE(1, range_sync_called.load());

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
@@ -389,7 +389,7 @@ TEST_F(DBIOFailureTest, FlushSstCloseError) {
  CreateAndReopenWithCF({"pikachu"}, options);
  Status s;
  std::atomic<int> close_called(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "SpecialEnv::SStableFile::Close", [&](void* arg) {
        if (close_called.fetch_add(1) == 0) {
          Status* st = static_cast<Status*>(arg);
@@ -397,7 +397,7 @@ TEST_F(DBIOFailureTest, FlushSstCloseError) {
        }
      });

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

  ASSERT_OK(Put(1, "foo", "bar"));
  ASSERT_OK(Put(1, "foo1", "bar1"));
@@ -409,7 +409,7 @@ TEST_F(DBIOFailureTest, FlushSstCloseError) {
  ASSERT_EQ("bar2", Get(1, "foo"));
  ASSERT_EQ("bar1", Get(1, "foo1"));

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

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
  ASSERT_EQ("bar2", Get(1, "foo"));
@@ -441,7 +441,7 @@ TEST_F(DBIOFailureTest, CompactionSstCloseError) {
  dbfull()->TEST_WaitForCompact();

  std::atomic<int> close_called(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "SpecialEnv::SStableFile::Close", [&](void* arg) {
        if (close_called.fetch_add(1) == 0) {
          Status* st = static_cast<Status*>(arg);
@@ -449,7 +449,7 @@ TEST_F(DBIOFailureTest, CompactionSstCloseError) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
  ASSERT_OK(dbfull()->SetOptions(handles_[1],
                                 {
                                     {"disable_auto_compactions", "false"},
@@ -460,7 +460,7 @@ TEST_F(DBIOFailureTest, CompactionSstCloseError) {
  ASSERT_NOK(Put(1, "foo2", "bar3"));
  ASSERT_EQ("bar3", Get(1, "foo"));

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

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
  ASSERT_EQ("bar3", Get(1, "foo"));
@@ -480,7 +480,7 @@ TEST_F(DBIOFailureTest, FlushSstSyncError) {
  CreateAndReopenWithCF({"pikachu"}, options);
  Status s;
  std::atomic<int> sync_called(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "SpecialEnv::SStableFile::Sync", [&](void* arg) {
        if (sync_called.fetch_add(1) == 0) {
          Status* st = static_cast<Status*>(arg);
@@ -488,7 +488,7 @@ TEST_F(DBIOFailureTest, FlushSstSyncError) {
        }
      });

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

  ASSERT_OK(Put(1, "foo", "bar"));
  ASSERT_OK(Put(1, "foo1", "bar1"));
@@ -500,7 +500,7 @@ TEST_F(DBIOFailureTest, FlushSstSyncError) {
  ASSERT_EQ("bar2", Get(1, "foo"));
  ASSERT_EQ("bar1", Get(1, "foo1"));

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

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
  ASSERT_EQ("bar2", Get(1, "foo"));
@@ -533,7 +533,7 @@ TEST_F(DBIOFailureTest, CompactionSstSyncError) {
  dbfull()->TEST_WaitForCompact();

  std::atomic<int> sync_called(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "SpecialEnv::SStableFile::Sync", [&](void* arg) {
        if (sync_called.fetch_add(1) == 0) {
          Status* st = static_cast<Status*>(arg);
@@ -541,7 +541,7 @@ TEST_F(DBIOFailureTest, CompactionSstSyncError) {
        }
      });

-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
  ASSERT_OK(dbfull()->SetOptions(handles_[1],
                                 {
                                     {"disable_auto_compactions", "false"},
@@ -552,17 +552,17 @@ TEST_F(DBIOFailureTest, CompactionSstSyncError) {
  ASSERT_NOK(Put(1, "foo2", "bar3"));
  ASSERT_EQ("bar3", Get(1, "foo"));

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

  ReopenWithColumnFamilies({"default", "pikachu"}, options);
  ASSERT_EQ("bar3", Get(1, "foo"));
 }
 #endif  // !(defined NDEBUG) || !defined(OS_WIN)
 #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();
 }
@@ -31,20 +31,7 @@
 #include "util/string_util.h"
 #include "util/user_comparator_wrapper.h"

-namespace rocksdb {
-
-#if 0
-static void DumpInternalIter(Iterator* iter) {
-  for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
-    ParsedInternalKey k;
-    if (!ParseInternalKey(iter->key(), &k)) {
-      fprintf(stderr, "Corrupt '%s'\n", EscapeString(iter->key()).c_str());
-    } else {
-      fprintf(stderr, "@ '%s'\n", k.DebugString().c_str());
-    }
-  }
-}
-#endif
+namespace ROCKSDB_NAMESPACE {

 DBIter::DBIter(Env* _env, const ReadOptions& read_options,
               const ImmutableCFOptions& cf_options,
@@ -62,6 +49,8 @@ DBIter::DBIter(Env* _env, const ReadOptions& read_options,
      read_callback_(read_callback),
      sequence_(s),
      statistics_(cf_options.statistics),
+      max_skip_(max_sequential_skip_in_iterations),
+      max_skippable_internal_keys_(read_options.max_skippable_internal_keys),
      num_internal_keys_skipped_(0),
      iterate_lower_bound_(read_options.iterate_lower_bound),
      iterate_upper_bound_(read_options.iterate_upper_bound),
@@ -73,23 +62,26 @@ DBIter::DBIter(Env* _env, const ReadOptions& read_options,
                                ? read_options.prefix_same_as_start
                                : false),
      pin_thru_lifetime_(read_options.pin_data),
-      total_order_seek_(read_options.total_order_seek),
+      expect_total_order_inner_iter_(prefix_extractor_ == nullptr ||
+                                     read_options.total_order_seek ||
+                                     read_options.auto_prefix_mode),
      allow_blob_(allow_blob),
      is_blob_(false),
      arena_mode_(arena_mode),
      range_del_agg_(&cf_options.internal_comparator, s),
      db_impl_(db_impl),
      cfd_(cfd),
-      start_seqnum_(read_options.iter_start_seqnum) {
+      start_seqnum_(read_options.iter_start_seqnum),
+      timestamp_ub_(read_options.timestamp),
+      timestamp_size_(timestamp_ub_ ? timestamp_ub_->size() : 0) {
  RecordTick(statistics_, NO_ITERATOR_CREATED);
-  max_skip_ = max_sequential_skip_in_iterations;
-  max_skippable_internal_keys_ = read_options.max_skippable_internal_keys;
  if (pin_thru_lifetime_) {
    pinned_iters_mgr_.StartPinning();
  }
  if (iter_.iter()) {
    iter_.iter()->SetPinnedItersMgr(&pinned_iters_mgr_);
  }
+  assert(timestamp_size_ == user_comparator_.timestamp_size());
 }

 Status DBIter::GetProperty(std::string prop_name, std::string* prop) {
@@ -181,7 +173,7 @@ void DBIter::Next() {
 //       a delete marker or a sequence number higher than sequence_
 //       saved_key_ MUST have a proper user_key before calling this function
 //
-// The prefix parameter, if not null, indicates that we need to iterator
+// The prefix parameter, if not null, indicates that we need to iterate
 // within the prefix, and the iterator needs to be made invalid, if no
 // more entry for the prefix can be found.
 bool DBIter::FindNextUserEntry(bool skipping_saved_key, const Slice* prefix) {
@@ -202,13 +194,14 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
  // or equal to saved_key_. We could skip these entries either because
  // sequence numbers were too high or because skipping_saved_key = true.
  // What saved_key_ contains throughout this method:
-  //  - if skipping_saved_key        : saved_key_ contains the key that we need
-  //  to skip,
-  //                         and we haven't seen any keys greater than that,
-  //  - if num_skipped > 0 : saved_key_ contains the key that we have skipped
-  //                         num_skipped times, and we haven't seen any keys
-  //                         greater than that,
-  //  - none of the above  : saved_key_ can contain anything, it doesn't matter.
+  //  - if skipping_saved_key : saved_key_ contains the key that we need
+  //                            to skip, and we haven't seen any keys greater
+  //                            than that,
+  //  - if num_skipped > 0    : saved_key_ contains the key that we have skipped
+  //                            num_skipped times, and we haven't seen any keys
+  //                            greater than that,
+  //  - none of the above     : saved_key_ can contain anything, it doesn't
+  //                            matter.
  uint64_t num_skipped = 0;
  // For write unprepared, the target sequence number in reseek could be larger
  // than the snapshot, and thus needs to be skipped again. This could result in
@@ -230,9 +223,13 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
    is_key_seqnum_zero_ = (ikey_.sequence == 0);

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

@@ -247,20 +244,25 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
      return false;
    }

-    if (IsVisible(ikey_.sequence)) {
+    assert(ikey_.user_key.size() >= timestamp_size_);
+    Slice ts;
+    if (timestamp_size_ > 0) {
+      ts = ExtractTimestampFromUserKey(ikey_.user_key, timestamp_size_);
+    }
+    if (IsVisible(ikey_.sequence, ts)) {
      // If the previous entry is of seqnum 0, the current entry will not
      // possibly be skipped. This condition can potentially be relaxed to
      // prev_key.seq <= ikey_.sequence. We are cautious because it will be more
      // prone to bugs causing the same user key with the same sequence number.
      if (!is_prev_key_seqnum_zero && skipping_saved_key &&
-          user_comparator_.Compare(ikey_.user_key, saved_key_.GetUserKey()) <=
-              0) {
+          user_comparator_.CompareWithoutTimestamp(
+              ikey_.user_key, saved_key_.GetUserKey()) <= 0) {
        num_skipped++;  // skip this entry
        PERF_COUNTER_ADD(internal_key_skipped_count, 1);
      } else {
        assert(!skipping_saved_key ||
-               user_comparator_.Compare(ikey_.user_key,
-                                        saved_key_.GetUserKey()) > 0);
+               user_comparator_.CompareWithoutTimestamp(
+                   ikey_.user_key, saved_key_.GetUserKey()) > 0);
        num_skipped = 0;
        reseek_done = false;
        switch (ikey_.type) {
@@ -290,9 +292,7 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
            if (start_seqnum_ > 0) {
              // we are taking incremental snapshot here
              // incremental snapshots aren't supported on DB with range deletes
-              assert(!(
-                (ikey_.type == kTypeBlobIndex) && (start_seqnum_ > 0)
-              ));
+              assert(ikey_.type != kTypeBlobIndex);
              if (ikey_.sequence >= start_seqnum_) {
                saved_key_.SetInternalKey(ikey_);
                valid_ = true;
@@ -323,7 +323,7 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
                  ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
                  status_ = Status::NotSupported(
                      "Encounter unexpected blob index. Please open DB with "
-                      "rocksdb::blob_db::BlobDB instead.");
+                      "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
                  valid_ = false;
                  return false;
                }
@@ -368,9 +368,9 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
      // This key was inserted after our snapshot was taken.
      // If this happens too many times in a row for the same user key, we want
      // to seek to the target sequence number.
-      int cmp =
-          user_comparator_.Compare(ikey_.user_key, saved_key_.GetUserKey());
-      if (cmp == 0 || (skipping_saved_key && cmp <= 0)) {
+      int cmp = user_comparator_.CompareWithoutTimestamp(
+          ikey_.user_key, saved_key_.GetUserKey());
+      if (cmp == 0 || (skipping_saved_key && cmp < 0)) {
        num_skipped++;
      } else {
        saved_key_.SetUserKey(
@@ -389,7 +389,7 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
    // reseek previously.
    //
    // TODO(lth): If we reseek to sequence number greater than ikey_.sequence,
-    // than it does not make sense to reseek as we would actually land further
+    // then it does not make sense to reseek as we would actually land further
    // away from the desired key. There is opportunity for optimization here.
    if (num_skipped > max_skip_ && !reseek_done) {
      is_key_seqnum_zero_ = false;
@@ -400,8 +400,17 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
        // We're looking for the next user-key but all we see are the same
        // user-key with decreasing sequence numbers. Fast forward to
        // sequence number 0 and type deletion (the smallest type).
-        AppendInternalKey(&last_key, ParsedInternalKey(saved_key_.GetUserKey(),
-                                                       0, kTypeDeletion));
+        if (timestamp_size_ == 0) {
+          AppendInternalKey(
+              &last_key,
+              ParsedInternalKey(saved_key_.GetUserKey(), 0, kTypeDeletion));
+        } else {
+          std::string min_ts(timestamp_size_, static_cast<char>(0));
+          AppendInternalKeyWithDifferentTimestamp(
+              &last_key,
+              ParsedInternalKey(saved_key_.GetUserKey(), 0, kTypeDeletion),
+              min_ts);
+        }
        // Don't set skipping_saved_key = false because we may still see more
        // user-keys equal to saved_key_.
      } else {
@@ -410,9 +419,17 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
        // Note that this only covers a case when a higher key was overwritten
        // many times since our snapshot was taken, not the case when a lot of
        // different keys were inserted after our snapshot was taken.
-        AppendInternalKey(&last_key,
-                          ParsedInternalKey(saved_key_.GetUserKey(), sequence_,
-                                            kValueTypeForSeek));
+        if (timestamp_size_ == 0) {
+          AppendInternalKey(
+              &last_key, ParsedInternalKey(saved_key_.GetUserKey(), sequence_,
+                                           kValueTypeForSeek));
+        } else {
+          AppendInternalKeyWithDifferentTimestamp(
+              &last_key,
+              ParsedInternalKey(saved_key_.GetUserKey(), sequence_,
+                                kValueTypeForSeek),
+              *timestamp_ub_);
+        }
      }
      iter_.Seek(last_key);
      RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
@@ -495,7 +512,7 @@ bool DBIter::MergeValuesNewToOld() {
        ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
        status_ = Status::NotSupported(
            "Encounter unexpected blob index. Please open DB with "
-            "rocksdb::blob_db::BlobDB instead.");
+            "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
      } else {
        status_ =
            Status::NotSupported("Blob DB does not support merge operator.");
@@ -531,6 +548,13 @@ bool DBIter::MergeValuesNewToOld() {
 }

 void DBIter::Prev() {
+  if (timestamp_size_ > 0) {
+    valid_ = false;
+    status_ = Status::NotSupported(
+        "SeekToLast/SeekForPrev/Prev currently not supported with timestamp.");
+    return;
+  }
+
  assert(valid_);
  assert(status_.ok());

@@ -567,7 +591,7 @@ bool DBIter::ReverseToForward() {
  // When moving backwards, iter_ is positioned on _previous_ key, which may
  // not exist or may have different prefix than the current key().
  // If that's the case, seek iter_ to current key.
-  if ((prefix_extractor_ != nullptr && !total_order_seek_) || !iter_.Valid()) {
+  if (!expect_total_order_inner_iter() || !iter_.Valid()) {
    IterKey last_key;
    last_key.SetInternalKey(ParsedInternalKey(
        saved_key_.GetUserKey(), kMaxSequenceNumber, kValueTypeForSeek));
@@ -603,15 +627,14 @@ bool DBIter::ReverseToBackward() {
  // key, which may not exist or may have prefix different from current.
  // If that's the case, seek to saved_key_.
  if (current_entry_is_merged_ &&
-      ((prefix_extractor_ != nullptr && !total_order_seek_) ||
-       !iter_.Valid())) {
+      (!expect_total_order_inner_iter() || !iter_.Valid())) {
    IterKey last_key;
    // Using kMaxSequenceNumber and kValueTypeForSeek
    // (not kValueTypeForSeekForPrev) to seek to a key strictly smaller
    // than saved_key_.
    last_key.SetInternalKey(ParsedInternalKey(
        saved_key_.GetUserKey(), kMaxSequenceNumber, kValueTypeForSeek));
-    if (prefix_extractor_ != nullptr && !total_order_seek_) {
+    if (!expect_total_order_inner_iter()) {
      iter_.SeekForPrev(last_key.GetInternalKey());
    } else {
      // Some iterators may not support SeekForPrev(), so we avoid using it
@@ -710,7 +733,13 @@ bool DBIter::FindValueForCurrentKey() {
      return false;
    }

-    if (!IsVisible(ikey.sequence) ||
+    assert(ikey.user_key.size() >= timestamp_size_);
+    Slice ts;
+    if (timestamp_size_ > 0) {
+      ts = Slice(ikey.user_key.data() + ikey.user_key.size() - timestamp_size_,
+                 timestamp_size_);
+    }
+    if (!IsVisible(ikey.sequence, ts) ||
        !user_comparator_.Equal(ikey.user_key, saved_key_.GetUserKey())) {
      break;
    }
@@ -797,7 +826,7 @@ bool DBIter::FindValueForCurrentKey() {
          ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
          status_ = Status::NotSupported(
              "Encounter unexpected blob index. Please open DB with "
-              "rocksdb::blob_db::BlobDB instead.");
+              "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
        } else {
          status_ =
              Status::NotSupported("Blob DB does not support merge operator.");
@@ -820,7 +849,7 @@ bool DBIter::FindValueForCurrentKey() {
        ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
        status_ = Status::NotSupported(
            "Encounter unexpected blob index. Please open DB with "
-            "rocksdb::blob_db::BlobDB instead.");
+            "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
        valid_ = false;
        return false;
      }
@@ -866,6 +895,13 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
    if (!ParseKey(&ikey)) {
      return false;
    }
+    assert(ikey.user_key.size() >= timestamp_size_);
+    Slice ts;
+    if (timestamp_size_ > 0) {
+      ts = Slice(ikey.user_key.data() + ikey.user_key.size() - timestamp_size_,
+                 timestamp_size_);
+    }
+
    if (!user_comparator_.Equal(ikey.user_key, saved_key_.GetUserKey())) {
      // No visible values for this key, even though FindValueForCurrentKey()
      // has seen some. This is possible if we're using a tailing iterator, and
@@ -874,7 +910,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
      return true;
    }

-    if (IsVisible(ikey.sequence)) {
+    if (IsVisible(ikey.sequence, ts)) {
      break;
    }

@@ -891,7 +927,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
    ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
    status_ = Status::NotSupported(
        "Encounter unexpected blob index. Please open DB with "
-        "rocksdb::blob_db::BlobDB instead.");
+        "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
    valid_ = false;
    return false;
  }
@@ -953,7 +989,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
        ROCKS_LOG_ERROR(logger_, "Encounter unexpected blob index.");
        status_ = Status::NotSupported(
            "Encounter unexpected blob index. Please open DB with "
-            "rocksdb::blob_db::BlobDB instead.");
+            "ROCKSDB_NAMESPACE::blob_db::BlobDB instead.");
      } else {
        status_ =
            Status::NotSupported("Blob DB does not support merge operator.");
@@ -978,8 +1014,8 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
  // Make sure we leave iter_ in a good state. If it's valid and we don't care
  // about prefixes, that's already good enough. Otherwise it needs to be
  // seeked to the current key.
-  if ((prefix_extractor_ != nullptr && !total_order_seek_) || !iter_.Valid()) {
-    if (prefix_extractor_ != nullptr && !total_order_seek_) {
+  if (!expect_total_order_inner_iter() || !iter_.Valid()) {
+    if (!expect_total_order_inner_iter()) {
      iter_.SeekForPrev(last_key);
    } else {
      iter_.Seek(last_key);
@@ -1014,7 +1050,13 @@ bool DBIter::FindUserKeyBeforeSavedKey() {
    }

    assert(ikey.sequence != kMaxSequenceNumber);
-    if (!IsVisible(ikey.sequence)) {
+    assert(ikey.user_key.size() >= timestamp_size_);
+    Slice ts;
+    if (timestamp_size_ > 0) {
+      ts = Slice(ikey.user_key.data() + ikey.user_key.size() - timestamp_size_,
+                 timestamp_size_);
+    }
+    if (!IsVisible(ikey.sequence, ts)) {
      PERF_COUNTER_ADD(internal_recent_skipped_count, 1);
    } else {
      PERF_COUNTER_ADD(internal_key_skipped_count, 1);
@@ -1059,10 +1101,18 @@ bool DBIter::TooManyInternalKeysSkipped(bool increment) {
  return false;
 }

-bool DBIter::IsVisible(SequenceNumber sequence) {
+bool DBIter::IsVisible(SequenceNumber sequence, const Slice& ts) {
+  // Remember that comparator orders preceding timestamp as larger.
+  int cmp_ts = timestamp_ub_ != nullptr
+                   ? user_comparator_.CompareTimestamp(ts, *timestamp_ub_)
+                   : 0;
+  if (cmp_ts > 0) {
+    return false;
+  }
  if (read_callback_ == nullptr) {
    return sequence <= sequence_;
  } else {
+    // TODO(yanqin): support timestamp in read_callback_.
    return read_callback_->IsVisible(sequence);
  }
 }
@@ -1071,14 +1121,16 @@ void DBIter::SetSavedKeyToSeekTarget(const Slice& target) {
  is_key_seqnum_zero_ = false;
  SequenceNumber seq = sequence_;
  saved_key_.Clear();
-  saved_key_.SetInternalKey(target, seq);
+  saved_key_.SetInternalKey(target, seq, kValueTypeForSeek, timestamp_ub_);

  if (iterate_lower_bound_ != nullptr &&
-      user_comparator_.Compare(saved_key_.GetUserKey(), *iterate_lower_bound_) <
-          0) {
+      user_comparator_.CompareWithoutTimestamp(
+          saved_key_.GetUserKey(), /*a_has_ts=*/true, *iterate_lower_bound_,
+          /*b_has_ts=*/false) < 0) {
    // Seek key is smaller than the lower bound.
    saved_key_.Clear();
-    saved_key_.SetInternalKey(*iterate_lower_bound_, seq);
+    saved_key_.SetInternalKey(*iterate_lower_bound_, seq, kValueTypeForSeek,
+                              timestamp_ub_);
  }
 }

@@ -1129,18 +1181,16 @@ void DBIter::Seek(const Slice& target) {

  // Now the inner iterator is placed to the target position. From there,
  // we need to find out the next key that is visible to the user.
-  //
  ClearSavedValue();
  if (prefix_same_as_start_) {
    // The case where the iterator needs to be invalidated if it has exausted
    // keys within the same prefix of the seek key.
    assert(prefix_extractor_ != nullptr);
-    Slice target_prefix;
-    target_prefix = prefix_extractor_->Transform(target);
+    Slice target_prefix = prefix_extractor_->Transform(target);
    FindNextUserEntry(false /* not skipping saved_key */,
                      &target_prefix /* prefix */);
    if (valid_) {
-      // Remember the prefix of the seek key for the future Prev() call to
+      // Remember the prefix of the seek key for the future Next() call to
      // check.
      prefix_.SetUserKey(target_prefix);
    }
@@ -1170,6 +1220,13 @@ void DBIter::SeekForPrev(const Slice& target) {
  }
 #endif  // ROCKSDB_LITE

+  if (timestamp_size_ > 0) {
+    valid_ = false;
+    status_ = Status::NotSupported(
+        "SeekToLast/SeekForPrev/Prev currently not supported with timestamp.");
+    return;
+  }
+
  status_ = Status::OK();
  ReleaseTempPinnedData();
  ResetInternalKeysSkippedCounter();
@@ -1196,8 +1253,7 @@ void DBIter::SeekForPrev(const Slice& target) {
    // The case where the iterator needs to be invalidated if it has exausted
    // keys within the same prefix of the seek key.
    assert(prefix_extractor_ != nullptr);
-    Slice target_prefix;
-    target_prefix = prefix_extractor_->Transform(target);
+    Slice target_prefix = prefix_extractor_->Transform(target);
    PrevInternal(&target_prefix);
    if (valid_) {
      // Remember the prefix of the seek key for the future Prev() call to
@@ -1224,7 +1280,7 @@ void DBIter::SeekToFirst() {
  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, env_);
  // Don't use iter_::Seek() if we set a prefix extractor
  // because prefix seek will be used.
-  if (prefix_extractor_ != nullptr && !total_order_seek_) {
+  if (!expect_total_order_inner_iter()) {
    max_skip_ = std::numeric_limits<uint64_t>::max();
  }
  status_ = Status::OK();
@@ -1264,6 +1320,13 @@ void DBIter::SeekToFirst() {
 }

 void DBIter::SeekToLast() {
+  if (timestamp_size_ > 0) {
+    valid_ = false;
+    status_ = Status::NotSupported(
+        "SeekToLast/SeekForPrev/Prev currently not supported with timestamp.");
+    return;
+  }
+
  if (iterate_upper_bound_ != nullptr) {
    // Seek to last key strictly less than ReadOptions.iterate_upper_bound.
    SeekForPrev(*iterate_upper_bound_);
@@ -1277,7 +1340,7 @@ void DBIter::SeekToLast() {
  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, env_);
  // Don't use iter_::Seek() if we set a prefix extractor
  // because prefix seek will be used.
-  if (prefix_extractor_ != nullptr && !total_order_seek_) {
+  if (!expect_total_order_inner_iter()) {
    max_skip_ = std::numeric_limits<uint64_t>::max();
  }
  status_ = Status::OK();
@@ -1323,4 +1386,4 @@ Iterator* NewDBIterator(Env* env, const ReadOptions& read_options,
  return db_iter;
 }

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -20,15 +20,15 @@
 #include "table/iterator_wrapper.h"
 #include "util/autovector.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 // This file declares the factory functions of DBIter, in its original form
 // or a wrapped form with class ArenaWrappedDBIter, which is defined here.
 // Class DBIter, which is declared and implemented inside db_iter.cc, is
-// a iterator that converts internal keys (yielded by an InternalIterator)
+// an iterator that converts internal keys (yielded by an InternalIterator)
 // that were live at the specified sequence number into appropriate user
 // keys.
-// Each internal key is consist of a user key, a sequence number, and a value
+// Each internal key consists of a user key, a sequence number, and a value
 // type. DBIter deals with multiple key versions, tombstones, merge operands,
 // etc, and exposes an Iterator.
 // For example, DBIter may wrap following InternalIterator:
@@ -133,14 +133,12 @@ class DBIter final : public Iterator {
    local_stats_.BumpGlobalStatistics(statistics_);
    iter_.DeleteIter(arena_mode_);
  }
-  virtual void SetIter(InternalIterator* iter) {
+  void SetIter(InternalIterator* iter) {
    assert(iter_.iter() == nullptr);
    iter_.Set(iter);
    iter_.iter()->SetPinnedItersMgr(&pinned_iters_mgr_);
  }
-  virtual ReadRangeDelAggregator* GetRangeDelAggregator() {
-    return &range_del_agg_;
-  }
+  ReadRangeDelAggregator* GetRangeDelAggregator() { return &range_del_agg_; }

  bool Valid() const override { return valid_; }
  Slice key() const override {
@@ -148,7 +146,8 @@ class DBIter final : public Iterator {
    if (start_seqnum_ > 0) {
      return saved_key_.GetInternalKey();
    } else {
-      return saved_key_.GetUserKey();
+      const Slice ukey_and_ts = saved_key_.GetUserKey();
+      return Slice(ukey_and_ts.data(), ukey_and_ts.size() - timestamp_size_);
    }
  }
  Slice value() const override {
@@ -171,6 +170,13 @@ class DBIter final : public Iterator {
      return status_;
    }
  }
+  Slice timestamp() const override {
+    assert(valid_);
+    assert(timestamp_size_ > 0);
+    const Slice ukey_and_ts = saved_key_.GetUserKey();
+    assert(timestamp_size_ < ukey_and_ts.size());
+    return ExtractTimestampFromUserKey(ukey_and_ts, timestamp_size_);
+  }
  bool IsBlob() const {
    assert(valid_ && (allow_blob_ || !is_blob_));
    return is_blob_;
@@ -180,11 +186,13 @@ class DBIter final : public Iterator {

  void Next() final override;
  void Prev() final override;
+  // 'target' does not contain timestamp, even if user timestamp feature is
+  // enabled.
  void Seek(const Slice& target) final override;
  void SeekForPrev(const Slice& target) final override;
  void SeekToFirst() final override;
  void SeekToLast() final override;
-  Env* env() { return env_; }
+  Env* env() const { return env_; }
  void set_sequence(uint64_t s) {
    sequence_ = s;
    if (read_callback_) {
@@ -202,10 +210,10 @@ class DBIter final : public Iterator {
  bool ReverseToBackward();
  // Set saved_key_ to the seek key to target, with proper sequence number set.
  // It might get adjusted if the seek key is smaller than iterator lower bound.
-  void SetSavedKeyToSeekTarget(const Slice& /*target*/);
+  void SetSavedKeyToSeekTarget(const Slice& target);
  // Set saved_key_ to the seek key to target, with proper sequence number set.
  // It might get adjusted if the seek key is larger than iterator upper bound.
-  void SetSavedKeyToSeekForPrevTarget(const Slice& /*target*/);
+  void SetSavedKeyToSeekForPrevTarget(const Slice& target);
  bool FindValueForCurrentKey();
  bool FindValueForCurrentKeyUsingSeek();
  bool FindUserKeyBeforeSavedKey();
@@ -221,9 +229,9 @@ class DBIter final : public Iterator {

  // If prefix is not null, we need to set the iterator to invalid if no more
  // entry can be found within the prefix.
-  void PrevInternal(const Slice* /*prefix*/);
+  void PrevInternal(const Slice* prefix);
  bool TooManyInternalKeysSkipped(bool increment = true);
-  bool IsVisible(SequenceNumber sequence);
+  bool IsVisible(SequenceNumber sequence, const Slice& ts);

  // Temporarily pin the blocks that we encounter until ReleaseTempPinnedData()
  // is called
@@ -257,6 +265,11 @@ class DBIter final : public Iterator {
    num_internal_keys_skipped_ = 0;
  }

+  bool expect_total_order_inner_iter() {
+    assert(expect_total_order_inner_iter_ || prefix_extractor_ != nullptr);
+    return expect_total_order_inner_iter_;
+  }
+
  const SliceTransform* prefix_extractor_;
  Env* const env_;
  Logger* logger_;
@@ -302,7 +315,9 @@ class DBIter final : public Iterator {
  // Means that we will pin all data blocks we read as long the Iterator
  // is not deleted, will be true if ReadOptions::pin_data is true
  const bool pin_thru_lifetime_;
-  const bool total_order_seek_;
+  // Expect the inner iterator to maintain a total order.
+  // prefix_extractor_ must be non-NULL if the value is false.
+  const bool expect_total_order_inner_iter_;
  bool allow_blob_;
  bool is_blob_;
  bool arena_mode_;
@@ -322,7 +337,10 @@ class DBIter final : public Iterator {
  // for diff snapshots we want the lower bound on the seqnum;
  // if this value > 0 iterator will return internal keys
  SequenceNumber start_seqnum_;
+  const Slice* const timestamp_ub_;
+  const size_t timestamp_size_;
 };
+
 // Return a new iterator that converts internal keys (yielded by
 // "*internal_iter") that were live at the specified `sequence` number
 // into appropriate user keys.
@@ -335,4 +353,4 @@ extern Iterator* NewDBIterator(
    ReadCallback* read_callback, DBImpl* db_impl = nullptr,
    ColumnFamilyData* cfd = nullptr, bool allow_blob = false);

-}  // namespace rocksdb
+}  // namespace ROCKSDB_NAMESPACE
@@ -29,7 +29,7 @@ bool FLAGS_verbose = false;

 #endif

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 class DBIteratorStressTest : public testing::Test {
 public:
@@ -645,7 +645,7 @@ TEST_F(DBIteratorStressTest, StressTest) {
            << "\n  mutated on the fly: " << num_recently_removed << std::endl;
 }

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

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -22,7 +22,7 @@
 #include "util/string_util.h"
 #include "utilities/merge_operators.h"

-namespace rocksdb {
+namespace ROCKSDB_NAMESPACE {

 static uint64_t TestGetTickerCount(const Options& options,
                                   Tickers ticker_type) {
@@ -691,7 +691,7 @@ TEST_F(DBIteratorTest, DBIteratorEmpty) {
 TEST_F(DBIteratorTest, DBIteratorUseSkipCountSkips) {
  ReadOptions ro;
  Options options;
-  options.statistics = rocksdb::CreateDBStatistics();
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
  options.merge_operator = MergeOperators::CreateFromStringId("stringappend");

  TestIterator* internal_iter = new TestIterator(BytewiseComparator());
@@ -746,7 +746,7 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
      }
      internal_iter->Finish();

-      options.statistics = rocksdb::CreateDBStatistics();
+      options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
      std::unique_ptr<Iterator> db_iter(NewDBIterator(
          env_, ro, cf_options, mutable_cf_options, BytewiseComparator(),
          internal_iter, i + 2, options.max_sequential_skip_in_iterations,
@@ -2481,7 +2481,7 @@ TEST_F(DBIteratorTest, DBIteratorTestDifferentialSnapshots) {
    ReadOptions ro;
    ro.iter_start_seqnum=5;
    Options options;
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();

    TestIterator* internal_iter = new TestIterator(BytewiseComparator());
    for (size_t i = 0; i < 10; ++i) {
@@ -2516,7 +2516,7 @@ TEST_F(DBIteratorTest, DBIteratorTestDifferentialSnapshots) {
    ReadOptions ro;
    ro.iter_start_seqnum=5;
    Options options;
-    options.statistics = rocksdb::CreateDBStatistics();
+    options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();

    TestIterator* internal_iter = new TestIterator(BytewiseComparator());
    for (size_t i = 0; i < 10; ++i) {
@@ -2661,10 +2661,10 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace1) {
  // Test call back inserts a key in the end of the mem table after
  // MergeIterator::Prev() realized the mem table iterator is at its end
  // and before an SeekToLast() is called.
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev",
      [&](void* /*arg*/) { internal_iter2_->Add("z", kTypeValue, "7", 12u); });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2683,7 +2683,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace1) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace2) {
@@ -2697,12 +2697,12 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace2) {
  // Test call back inserts entries for update a key in the end of the
  // mem table after MergeIterator::Prev() realized the mem tableiterator is at
  // its end and before an SeekToLast() is called.
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* /*arg*/) {
        internal_iter2_->Add("z", kTypeValue, "7", 12u);
        internal_iter2_->Add("z", kTypeValue, "7", 11u);
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2721,7 +2721,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace2) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace3) {
@@ -2735,7 +2735,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace3) {
  // Test call back inserts entries for update a key in the end of the
  // mem table after MergeIterator::Prev() realized the mem table iterator is at
  // its end and before an SeekToLast() is called.
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* /*arg*/) {
        internal_iter2_->Add("z", kTypeValue, "7", 16u, true);
        internal_iter2_->Add("z", kTypeValue, "7", 15u, true);
@@ -2744,7 +2744,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace3) {
        internal_iter2_->Add("z", kTypeValue, "7", 12u, true);
        internal_iter2_->Add("z", kTypeValue, "7", 11u, true);
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2763,7 +2763,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace3) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace4) {
@@ -2779,7 +2779,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace4) {
  // Test call back inserts entries for update a key before "z" in
  // mem table after MergeIterator::Prev() calls mem table iterator's
  // Seek() and before calling Prev()
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* arg) {
        IteratorWrapper* it = reinterpret_cast<IteratorWrapper*>(arg);
        if (it->key().starts_with("z")) {
@@ -2791,7 +2791,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace4) {
          internal_iter2_->Add("x", kTypeValue, "7", 11u, true);
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2814,7 +2814,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace4) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace5) {
@@ -2830,7 +2830,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace5) {
  // Test call back inserts entries for update a key before "z" in
  // mem table after MergeIterator::Prev() calls mem table iterator's
  // Seek() and before calling Prev()
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* arg) {
        IteratorWrapper* it = reinterpret_cast<IteratorWrapper*>(arg);
        if (it->key().starts_with("z")) {
@@ -2838,7 +2838,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace5) {
          internal_iter2_->Add("x", kTypeValue, "7", 15u, true);
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2861,7 +2861,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace5) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace6) {
@@ -2877,14 +2877,14 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace6) {
  // Test call back inserts an entry for update a key before "z" in
  // mem table after MergeIterator::Prev() calls mem table iterator's
  // Seek() and before calling Prev()
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* arg) {
        IteratorWrapper* it = reinterpret_cast<IteratorWrapper*>(arg);
        if (it->key().starts_with("z")) {
          internal_iter2_->Add("x", kTypeValue, "7", 16u, true);
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2907,7 +2907,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace6) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace7) {
@@ -2926,7 +2926,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace7) {
  // Test call back inserts entries for update a key before "z" in
  // mem table after MergeIterator::Prev() calls mem table iterator's
  // Seek() and before calling Prev()
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* arg) {
        IteratorWrapper* it = reinterpret_cast<IteratorWrapper*>(arg);
        if (it->key().starts_with("z")) {
@@ -2938,7 +2938,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace7) {
          internal_iter2_->Add("x", kTypeValue, "7", 11u, true);
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2961,7 +2961,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace7) {
  ASSERT_EQ(db_iter_->key().ToString(), "a");
  ASSERT_EQ(db_iter_->value().ToString(), "4");

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

 TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace8) {
@@ -2979,7 +2979,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace8) {
  // Test call back inserts two keys before "z" in mem table after
  // MergeIterator::Prev() calls mem table iterator's Seek() and
  // before calling Prev()
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
      "MergeIterator::Prev:BeforePrev", [&](void* arg) {
        IteratorWrapper* it = reinterpret_cast<IteratorWrapper*>(arg);
        if (it->key().starts_with("z")) {
@@ -2987,7 +2987,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace8) {
          internal_iter2_->Add("y", kTypeValue, "7", 17u, true);
        }
      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();

  db_iter_->Prev();
  ASSERT_TRUE(db_iter_->Valid());
@@ -2998,7 +2998,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIteratorDataRace8) {
  ASSERT_EQ(db_iter_->key().ToString(), "d");
  ASSERT_EQ(db_iter_->value().ToString(), "7");

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


@@ -3167,7 +3167,7 @@ TEST_F(DBIteratorTest, ReverseToForwardWithDisappearingKeys) {
  EXPECT_LT(internal_iter->steps(), 20);
 }

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

 int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
--- a/Show More
+++ b/Show More