update HISTORY.md and bump version for 6.19.4

Summary:
In current RocksDB, in recover the information form WAL, we do the consistency check for each column family when one WAL file is corrupted and PointInTimeRecovery is set. However, it will report a false positive alert on "SST file is ahead of WALs" when one of the CF current log number is greater than the corrupted WAL number (CF contains the data beyond the corrupted WAl) due to a new column family creation during flush. In this case, a new WAL is created (it is empty) during a flush. Also, due to some reason (e.g., storage issue or crash happens before SyncCloseLog is called), the old WAL is corrupted. The new CF has no data, therefore, it does not have the consistency issue.

Fix: when checking cfd->GetLogNumber() > corrupted_wal_number also check cfd->GetLiveSstFilesSize() > 0. So the CFs with no SST file data will skip the check here.

Note potential ignored inconsistency caused due to fix: empty CF can also be caused by write+delete. In this case, after flush, there is no SST files being generated. However, this CF still have the log in the WAL. When the WAL is corrupted, the DB might be inconsistent.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8207

Test Plan: added unit test, make crash_test

Reviewed By: riversand963

Differential Revision: D27898839

Pulled By: zhichao-cao

fbshipit-source-id: 931fc2d8b92dd00b4169bf84b94e712fd688a83e

.circleci/config.yml

@@ -103,6 +103,13 @@ commands:
             git clone --single-branch --branch master --depth 1 https://github.com/google/gtest-parallel.git ~/gtest-parallel
             echo 'export PATH=$HOME/gtest-parallel:$PATH' >> $BASH_ENV
 
+  install-compression-libs:
+    steps:
+      - run:
+          name: Install compression libs
+          command: |
+            sudo apt-get update -y && sudo apt-get install -y libsnappy-dev zlib1g-dev libbz2-dev liblz4-dev libzstd-dev
+
 executors:
   windows-2xlarge:
     machine:
@@ -565,7 +572,25 @@ jobs:
             gtest-parallel $(</tmp/test_list) --output_dir=/tmp | cat  # pipe to cat to continuously output status on circleci UI. Otherwise, no status will be printed while the job is running.
       - post-steps
 
+  build-format-compatible:
+    machine:
+      image: ubuntu-1604:202007-01
+    resource_class: 2xlarge
+    steps:
+      - pre-steps
+      - install-gflags
+      - install-compression-libs
+      - run:
+          name: "test"
+          command: |
+            export TEST_TMPDIR=/dev/shm/rocksdb
+            rm -rf /dev/shm/rocksdb
+            mkdir /dev/shm/rocksdb
+            tools/check_format_compatible.sh
+      - post-steps
+
 workflows:
+  version: 2
   build-linux:
     jobs:
       - build-linux
@@ -666,3 +691,13 @@ workflows:
   build-cmake-mingw:
     jobs:
       - build-cmake-mingw
+  nightly:
+    triggers:
+      - schedule:
+          cron: "0 0 * * *"
+          filters:
+            branches:
+              only:
+                - master
+    jobs:
+      - build-format-compatible

.github/workflows/sanity_check.yml

@@ -32,7 +32,7 @@ jobs:
     - name: Download clang-format-diff.py
       uses: wei/wget@v1
       with:
-        args: https://raw.githubusercontent.com/llvm/llvm-project/master/clang/tools/clang-format/clang-format-diff.py
+        args: https://raw.githubusercontent.com/llvm/llvm-project/main/clang/tools/clang-format/clang-format-diff.py
 
     - name: Check format
       run: VERBOSE_CHECK=1 make check-format

.gitignore

@@ -93,3 +93,5 @@ clang-format-diff.py
 
 fuzz/proto/gen/
 fuzz/crash-*
+
+cmake-build-*

.travis.yml

@@ -196,30 +196,30 @@ script:
   - if [ `command -v ccache` ]; then ccache -C; fi
   - case $TEST_GROUP in
     platform_dependent)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=only make -j4 all_but_some_tests check_some
+      OPT=-DTRAVIS LIB_MODE=shared V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=only make -j4 all_but_some_tests check_some
       ;;
     1)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_END=backupable_db_test make -j4 check_some
+      OPT=-DTRAVIS LIB_MODE=shared V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_END=backupable_db_test make -j4 check_some
       ;;
     2)
-      OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" V=1 make -j4 tools && OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=backupable_db_test ROCKSDBTESTS_END=db_universal_compaction_test make -j4 check_some
+      OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" LIB_MODE=shared V=1 make -j4 tools && OPT="-DTRAVIS -DROCKSDB_NAMESPACE=alternative_rocksdb_ns" LIB_MODE=shared V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=backupable_db_test ROCKSDBTESTS_END=db_universal_compaction_test make -j4 check_some
       ;;
     3)
       OPT=-DTRAVIS LIB_MODE=shared V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=db_universal_compaction_test ROCKSDBTESTS_END=table_properties_collector_test make -j4 check_some
       ;;
     4)
-      OPT=-DTRAVIS V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=table_properties_collector_test make -j4 check_some
+      OPT=-DTRAVIS LIB_MODE=shared V=1 ROCKSDBTESTS_PLATFORM_DEPENDENT=exclude ROCKSDBTESTS_START=table_properties_collector_test make -j4 check_some
       ;;
     esac
   - case $JOB_NAME in
     java_test)
-      OPT=-DTRAVIS V=1 make rocksdbjava jtest
+      OPT=-DTRAVIS LIB_MODE=shared V=1 make rocksdbjava jtest
       ;;
     lite_build)
-      OPT='-DTRAVIS -DROCKSDB_LITE' V=1 make -j4 all
+      OPT='-DTRAVIS -DROCKSDB_LITE' LIB_MODE=shared V=1 make -j4 all
       ;;
     examples)
-      OPT=-DTRAVIS V=1 make -j4 static_lib && cd examples && make -j4
+      OPT=-DTRAVIS LIB_MODE=shared V=1 make -j4 static_lib && cd examples && make -j4
       ;;
     cmake-mingw)
       sudo update-alternatives --set x86_64-w64-mingw32-g++ /usr/bin/x86_64-w64-mingw32-g++-posix;
@@ -235,10 +235,10 @@ script:
       mkdir build && cd build && cmake -DCMAKE_BUILD_TYPE=Release -DWITH_TESTS=0 -DWITH_GFLAGS=0 -DWITH_BENCHMARK_TOOLS=0 -DWITH_TOOLS=0 -DWITH_CORE_TOOLS=1 .. && make -j4 && cd  .. && rm -rf build && mkdir build && cd build && cmake -DJNI=1 .. -DCMAKE_BUILD_TYPE=Release $OPT && make -j4 rocksdb rocksdbjni
       ;;
     make-gcc4.8)
-      OPT=-DTRAVIS V=1 SKIP_LINK=1 make -j4 all && [ "Linking broken because libgflags compiled with newer ABI" ]
+      OPT=-DTRAVIS LIB_MODE=shared V=1 SKIP_LINK=1 make -j4 all && [ "Linking broken because libgflags compiled with newer ABI" ]
       ;;
     status_checked)
-      OPT=-DTRAVIS V=1 ASSERT_STATUS_CHECKED=1 make -j4 check_some
+      OPT=-DTRAVIS LIB_MODE=shared V=1 ASSERT_STATUS_CHECKED=1 make -j4 check_some
       ;;
     esac
 notifications:

CMakeLists.txt

@@ -88,10 +88,9 @@ if( NOT DEFINED CMAKE_CXX_STANDARD )
 endif()
 
 include(CMakeDependentOption)
-CMAKE_DEPENDENT_OPTION(WITH_GFLAGS "build with GFlags" ON
-  "NOT MSVC;NOT MINGW" OFF)
 
 if(MSVC)
+  option(WITH_GFLAGS "build with GFlags" OFF)
   option(WITH_XPRESS "build with windows built in compression" OFF)
   include(${CMAKE_CURRENT_SOURCE_DIR}/thirdparty.inc)
 else()
@@ -107,6 +106,11 @@ else()
     endif()
   endif()
 
+  if(MINGW)
+    option(WITH_GFLAGS "build with GFlags" OFF)
+  else()
+    option(WITH_GFLAGS "build with GFlags" ON)
+  endif()
   set(GFLAGS_LIB)
   if(WITH_GFLAGS)
     # Config with namespace available since gflags 2.2.2
@@ -170,23 +174,25 @@ else()
   endif()
 endif()
 
-string(TIMESTAMP TS "%Y/%m/%d %H:%M:%S" UTC)
-set(GIT_DATE_TIME "${TS}" CACHE STRING "the time we first built rocksdb")
+string(TIMESTAMP TS "%Y-%m-%d %H:%M:%S" UTC)
+set(BUILD_DATE "${TS}" CACHE STRING "the time we first built rocksdb")
 
 find_package(Git)
 
 if(GIT_FOUND AND EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/.git")
-  if(WIN32)
-    execute_process(COMMAND $ENV{COMSPEC} /C ${GIT_EXECUTABLE} -C ${CMAKE_CURRENT_SOURCE_DIR} rev-parse HEAD OUTPUT_VARIABLE GIT_SHA)
-  else()
-    execute_process(COMMAND ${GIT_EXECUTABLE} -C ${CMAKE_CURRENT_SOURCE_DIR} rev-parse HEAD OUTPUT_VARIABLE GIT_SHA)
+  execute_process(WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}" OUTPUT_VARIABLE GIT_SHA COMMAND "${GIT_EXECUTABLE}" rev-parse HEAD )
+  execute_process(WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}" RESULT_VARIABLE GIT_MOD COMMAND "${GIT_EXECUTABLE}" diff-index HEAD --quiet)
+  execute_process(WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}" OUTPUT_VARIABLE GIT_DATE COMMAND "${GIT_EXECUTABLE}" log -1 --date=format:"%Y-%m-%d %T" --format="%ad")
+  execute_process(WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}" OUTPUT_VARIABLE GIT_TAG RESULT_VARIABLE rv COMMAND "${GIT_EXECUTABLE}" symbolic-ref -q --short HEAD OUTPUT_STRIP_TRAILING_WHITESPACE)
+  if (rv AND NOT rv EQUAL 0)
+    execute_process(WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}" OUTPUT_VARIABLE GIT_TAG COMMAND "${GIT_EXECUTABLE}" describe --tags --exact-match OUTPUT_STRIP_TRAILING_WHITESPACE)
   endif()
 else()
   set(GIT_SHA 0)
+  set(GIT_MOD 1)
 endif()
-
-string(REGEX REPLACE "[^0-9a-f]+" "" GIT_SHA "${GIT_SHA}")
-
+string(REGEX REPLACE "[^0-9a-fA-F]+" "" GIT_SHA "${GIT_SHA}")
+string(REGEX REPLACE "[^0-9: /-]+" "" GIT_DATE "${GIT_DATE}")
 
 option(WITH_MD_LIBRARY "build with MD" ON)
 if(WIN32 AND MSVC)
@@ -199,15 +205,16 @@ endif()
 
 set(BUILD_VERSION_CC ${CMAKE_BINARY_DIR}/build_version.cc)
 configure_file(util/build_version.cc.in ${BUILD_VERSION_CC} @ONLY)
-add_library(build_version OBJECT ${BUILD_VERSION_CC})
-target_include_directories(build_version PRIVATE
-  ${CMAKE_CURRENT_SOURCE_DIR}/util)
+
 if(MSVC)
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /Zi /nologo /EHsc /GS /Gd /GR /GF /fp:precise /Zc:wchar_t /Zc:forScope /errorReport:queue")
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /FC /d2Zi+ /W4 /wd4127 /wd4800 /wd4996 /wd4351 /wd4100 /wd4204 /wd4324")
 else()
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -W -Wextra -Wall")
   set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wsign-compare -Wshadow -Wno-unused-parameter -Wno-unused-variable -Woverloaded-virtual -Wnon-virtual-dtor -Wno-missing-field-initializers -Wno-strict-aliasing")
+  if(CMAKE_SYSTEM_PROCESSOR MATCHES "x86_64")
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wstrict-prototypes")
+  endif()
   if(MINGW)
     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-format -fno-asynchronous-unwind-tables")
     add_definitions(-D_POSIX_C_SOURCE=1)
@@ -243,14 +250,14 @@ if(CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64")
   endif(HAS_ALTIVEC)
 endif(CMAKE_SYSTEM_PROCESSOR MATCHES "^(powerpc|ppc)64")
 
-if(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|AARCH64")
+if(CMAKE_SYSTEM_PROCESSOR MATCHES "arm64|aarch64|AARCH64")
         CHECK_C_COMPILER_FLAG("-march=armv8-a+crc+crypto" HAS_ARMV8_CRC)
   if(HAS_ARMV8_CRC)
     message(STATUS " HAS_ARMV8_CRC yes")
     set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -march=armv8-a+crc+crypto -Wno-unused-function")
     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -march=armv8-a+crc+crypto -Wno-unused-function")
   endif(HAS_ARMV8_CRC)
-endif(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|AARCH64")
+endif(CMAKE_SYSTEM_PROCESSOR MATCHES "arm64|aarch64|AARCH64")
 
 option(PORTABLE "build a portable binary" OFF)
 option(FORCE_SSE42 "force building with SSE4.2, even when PORTABLE=ON" OFF)
@@ -466,11 +473,6 @@ if(CMAKE_SYSTEM_NAME MATCHES "Cygwin")
   add_definitions(-fno-builtin-memcmp -DCYGWIN)
 elseif(CMAKE_SYSTEM_NAME MATCHES "Darwin")
   add_definitions(-DOS_MACOSX)
-  if(CMAKE_SYSTEM_PROCESSOR MATCHES arm)
-    add_definitions(-DIOS_CROSS_COMPILE -DROCKSDB_LITE)
-    # no debug info for IOS, that will make our library big
-    add_definitions(-DNDEBUG)
-  endif()
 elseif(CMAKE_SYSTEM_NAME MATCHES "Linux")
   add_definitions(-DOS_LINUX)
 elseif(CMAKE_SYSTEM_NAME MATCHES "SunOS")
@@ -642,6 +644,7 @@ set(SOURCES
         db/write_batch_base.cc
         db/write_controller.cc
         db/write_thread.cc
+        env/composite_env.cc
         env/env.cc
         env/env_chroot.cc
         env/env_encryption.cc
@@ -653,6 +656,7 @@ set(SOURCES
         file/file_prefetch_buffer.cc
         file/file_util.cc
         file/filename.cc
+        file/line_file_reader.cc
         file/random_access_file_reader.cc
         file/read_write_util.cc
         file/readahead_raf.cc
@@ -766,6 +770,7 @@ set(SOURCES
         util/murmurhash.cc
         util/random.cc
         util/rate_limiter.cc
+        util/ribbon_config.cc
         util/slice.cc
         util/file_checksum_helper.cc
         util/status.cc
@@ -830,8 +835,7 @@ set(SOURCES
         utilities/transactions/write_unprepared_txn_db.cc
         utilities/ttl/db_ttl_impl.cc
         utilities/write_batch_with_index/write_batch_with_index.cc
-        utilities/write_batch_with_index/write_batch_with_index_internal.cc
-        $<TARGET_OBJECTS:build_version>)
+        utilities/write_batch_with_index/write_batch_with_index_internal.cc)
 
 list(APPEND SOURCES
   utilities/transactions/lock/range/range_tree/lib/locktree/concurrent_tree.cc
@@ -922,12 +926,12 @@ else()
   set(SYSTEM_LIBS ${CMAKE_THREAD_LIBS_INIT})
 endif()
 
-add_library(${ROCKSDB_STATIC_LIB} STATIC ${SOURCES})
+add_library(${ROCKSDB_STATIC_LIB} STATIC ${SOURCES} ${BUILD_VERSION_CC})
 target_link_libraries(${ROCKSDB_STATIC_LIB} PRIVATE
   ${THIRDPARTY_LIBS} ${SYSTEM_LIBS})
 
 if(ROCKSDB_BUILD_SHARED)
-  add_library(${ROCKSDB_SHARED_LIB} SHARED ${SOURCES})
+  add_library(${ROCKSDB_SHARED_LIB} SHARED ${SOURCES} ${BUILD_VERSION_CC})
   target_link_libraries(${ROCKSDB_SHARED_LIB} PRIVATE
     ${THIRDPARTY_LIBS} ${SYSTEM_LIBS})
 
@@ -945,7 +949,7 @@ if(ROCKSDB_BUILD_SHARED)
                           LINKER_LANGUAGE CXX
                           VERSION ${rocksdb_VERSION}
                           SOVERSION ${rocksdb_VERSION_MAJOR}
-                          OUTPUT_NAME "rocksdb")
+                          OUTPUT_NAME "rocksdb${ARTIFACT_SUFFIX}")
   endif()
 endif()
 
@@ -1066,6 +1070,8 @@ if(WITH_TESTS)
         db/blob/blob_file_garbage_test.cc
         db/blob/blob_file_reader_test.cc
         db/blob/db_blob_basic_test.cc
+        db/blob/db_blob_compaction_test.cc
+        db/blob/db_blob_corruption_test.cc
         db/blob/db_blob_index_test.cc
         db/column_family_test.cc
         db/compact_files_test.cc
@@ -1088,6 +1094,7 @@ if(WITH_TESTS)
         db/db_iter_test.cc
         db/db_iter_stress_test.cc
         db/db_iterator_test.cc
+        db/db_kv_checksum_test.cc
         db/db_log_iter_test.cc
         db/db_memtable_test.cc
         db/db_merge_operator_test.cc
@@ -1095,7 +1102,7 @@ if(WITH_TESTS)
         db/db_options_test.cc
         db/db_properties_test.cc
         db/db_range_del_test.cc
-        db/db_impl/db_secondary_test.cc
+        db/db_secondary_test.cc
         db/db_sst_test.cc
         db/db_statistics_test.cc
         db/db_table_properties_test.cc
@@ -1301,40 +1308,40 @@ if(WITH_TESTS)
 endif()
 
 if(WITH_BENCHMARK_TOOLS)
-  add_executable(db_bench
+  add_executable(db_bench${ARTIFACT_SUFFIX}
     tools/db_bench.cc
     tools/db_bench_tool.cc)
-  target_link_libraries(db_bench
+  target_link_libraries(db_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${THIRDPARTY_LIBS})
 
-  add_executable(cache_bench
+  add_executable(cache_bench${ARTIFACT_SUFFIX}
     cache/cache_bench.cc)
-  target_link_libraries(cache_bench
+  target_link_libraries(cache_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${GFLAGS_LIB})
 
-  add_executable(memtablerep_bench
+  add_executable(memtablerep_bench${ARTIFACT_SUFFIX}
     memtable/memtablerep_bench.cc)
-  target_link_libraries(memtablerep_bench
+  target_link_libraries(memtablerep_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${GFLAGS_LIB})
 
-  add_executable(range_del_aggregator_bench
+  add_executable(range_del_aggregator_bench${ARTIFACT_SUFFIX}
     db/range_del_aggregator_bench.cc)
-  target_link_libraries(range_del_aggregator_bench
+  target_link_libraries(range_del_aggregator_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${GFLAGS_LIB})
 
-  add_executable(table_reader_bench
+  add_executable(table_reader_bench${ARTIFACT_SUFFIX}
     table/table_reader_bench.cc)
-  target_link_libraries(table_reader_bench
+  target_link_libraries(table_reader_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} testharness ${GFLAGS_LIB})
 
-  add_executable(filter_bench
+  add_executable(filter_bench${ARTIFACT_SUFFIX}
     util/filter_bench.cc)
-  target_link_libraries(filter_bench
+  target_link_libraries(filter_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${GFLAGS_LIB})
 
-  add_executable(hash_table_bench
+  add_executable(hash_table_bench${ARTIFACT_SUFFIX}
     utilities/persistent_cache/hash_table_bench.cc)
-  target_link_libraries(hash_table_bench
+  target_link_libraries(hash_table_bench${ARTIFACT_SUFFIX}
     ${ROCKSDB_LIB} ${GFLAGS_LIB})
 endif()
 

HISTORY.md

@@ -1,16 +1,76 @@
 # Rocksdb Change Log
-## 6.17.3 (02/18/2021)
+## 6.19.4 (04/23/2021)
 ### Bug Fixes
-* Fix `WRITE_PREPARED`, `WRITE_UNPREPARED` TransactionDB `MultiGet()` may return uncommitted data with snapshot.
+* Fixed a bug where ingested files were written with incorrect boundary key metadata. In rare cases this could have led to a level's files being wrongly ordered and queries for the boundary keys returning wrong results.
+* Fixed the false-positive alert when recovering from the WAL file. Avoid reporting "SST file is ahead of WAL" on a newly created empty column family, if the previous WAL file is corrupted.
+
+### Behavior Changes
+* Due to the fix of false-postive alert of "SST file is ahead of WAL", all the CFs with no SST file (CF empty) will bypass the consistency check. We fixed a false-positive, but introduced a very rare true-negative which will be triggered in the following conditions: A CF with some delete operations in the last a few queries which will result in an empty CF (those are flushed to SST file and a compaction triggered which combines this file and all other SST files and generates an empty CF, or there is another reason to write a manifest entry for this CF after a flush that generates no SST file from an empty CF). The deletion entries are logged in a WAL and this WAL was corrupted, while the CF's log number points to the next WAL (due to the flush). Therefore, the DB can only recover to the point without these trailing deletions and cause the inconsistent DB status.
+
+## 6.19.3 (04/19/2021)
+### Bug Fixes
+* Fixed a bug in handling file rename error in distributed/network file systems when the server succeeds but client returns error. The bug can cause CURRENT file to point to non-existing MANIFEST file, thus DB cannot be opened.
+
+## 6.19.2 (04/08/2021)
+### Bug Fixes
+* Fixed a backward iteration bug with partitioned filter enabled: not including the prefix of the last key of the previous filter partition in current filter partition can cause wrong iteration result.
+
+## 6.19.1 (04/01/2021)
+### Bug Fixes
+* Fixed crash (divide by zero) when compression dictionary is applied to a file containing only range tombstones.
+
+## 6.19.0 (03/21/2021)
+### Bug Fixes
+* Fixed the truncation error found in APIs/tools when dumping block-based SST files in a human-readable format. After fix, the block-based table can be fully dumped as a readable file.
+* When hitting a write slowdown condition, no write delay (previously 1 millisecond) is imposed until `delayed_write_rate` is actually exceeded, with an initial burst allowance of 1 millisecond worth of bytes. Also, beyond the initial burst allowance, `delayed_write_rate` is now more strictly enforced, especially with multiple column families.
+
+### Public API change
+* Changed default `BackupableDBOptions::share_files_with_checksum` to `true` and deprecated `false` because of potential for data loss. Note that accepting this change in behavior can temporarily increase backup data usage because files are not shared between backups using the two different settings. Also removed obsolete option kFlagMatchInterimNaming.
+* Add a new option BlockBasedTableOptions::max_auto_readahead_size. RocksDB does auto-readahead for iterators on noticing more than two reads for a table file if user doesn't provide readahead_size. The readahead starts at 8KB and doubles on every additional read upto max_auto_readahead_size and now max_auto_readahead_size can be configured dynamically as well. Found that 256 KB readahead size provides the best performance, based on experiments, for auto readahead. Experiment data is in PR #3282. If value is set 0 then no automatic prefetching will be done by rocksdb. Also changing the value will only affect files opened after the change.
+* Add suppport to extend DB::VerifyFileChecksums API to also verify blob files checksum.
+* When using the new BlobDB, the amount of data written by flushes/compactions is now broken down into table files and blob files in the compaction statistics; namely, Write(GB) denotes the amount of data written to table files, while Wblob(GB) means the amount of data written to blob files.
+* New default BlockBasedTableOptions::format_version=5 to enable new Bloom filter implementation by default, compatible with RocksDB versions >= 6.6.0.
+* Add new SetBufferSize API to WriteBufferManager to allow dynamic management of memory allotted to all write buffers.  This allows user code to adjust memory monitoring provided by WriteBufferManager as process memory needs change datasets grow and shrink.
+* Clarified the required semantics of Read() functions in FileSystem and Env APIs. Please ensure any custom implementations are compliant.
+* For the new integrated BlobDB implementation, compaction statistics now include the amount of data read from blob files during compaction (due to garbage collection or compaction filters). Write amplification metrics have also been extended to account for data read from blob files.
+* Add EqualWithoutTimestamp() to Comparator.
+* Extend support to track blob files in SSTFileManager whenever a blob file is created/deleted. Blob files will be scheduled to delete via SSTFileManager and SStFileManager will now take blob files in account while calculating size and space limits along with SST files.
+* Add new Append and PositionedAppend API with checksum handoff to legacy Env.
+
+### New Features
+* Support compaction filters for the new implementation of BlobDB. Add `FilterBlobByKey()` to `CompactionFilter`. Subclasses can override this method so that compaction filters can determine whether the actual blob value has to be read during compaction. Use a new `kUndetermined` in `CompactionFilter::Decision` to indicated that further action is necessary for compaction filter to make a decision.
+* Add support to extend retrieval of checksums for blob files from the MANIFEST when checkpointing. During backup, rocksdb can detect corruption in blob files  during file copies.
+* Add new options for db_bench --benchmarks: flush, waitforcompaction, compact0, compact1.
+* Add an option to BackupEngine::GetBackupInfo to include the name and size of each backed-up file. Especially in the presence of file sharing among backups, this offers detailed insight into backup space usage.
+* Enable backward iteration on keys with user-defined timestamps.
+* Add statistics and info log for error handler: counters for bg error, bg io error, bg retryable io error, auto resume count, auto resume total retry number, and auto resume sucess; Histogram for auto resume retry count in each recovery call. Note that, each auto resume attempt will have one or multiple retries.
+
+### Behavior Changes
+* During flush, only WAL sync retryable IO error is mapped to hard error, which will stall the writes. When WAL is used but only SST file write has retryable IO error, it will be mapped to soft error and write will not be affected.
+
+## 6.18.0 (02/19/2021)
+### Behavior Changes
+* When retryable IO error occurs during compaction, it is mapped to soft error and set the BG error. However, auto resume is not called to clean the soft error since compaction will reschedule by itself. In this change, When retryable IO error occurs during compaction, BG error is not set. User will be informed the error via EventHelper.
+* Introduce a new trace file format for query tracing and replay and trace file version is bump up to 0.2. A payload map is added as the first portion of the payload. We will not have backward compatible issues when adding new entries to trace records. Added the iterator_upper_bound and iterator_lower_bound in Seek and SeekForPrev tracing function. Added them as the new payload member for iterator tracing.
+
+### New Features
+* Add support for key-value integrity protection in live updates from the user buffers provided to `WriteBatch` through the write to RocksDB's in-memory update buffer (memtable). This is intended to detect some cases of in-memory data corruption, due to either software or hardware errors. Users can enable protection by constructing their `WriteBatch` with `protection_bytes_per_key == 8`.
+* Add support for updating `full_history_ts_low` option in manual compaction, which is for old timestamp data GC.
+* Add a mechanism for using Makefile to build external plugin code into the RocksDB libraries/binaries. This intends to simplify compatibility and distribution for plugins (e.g., special-purpose `FileSystem`s) whose source code resides outside the RocksDB repo. See "plugin/README.md" for developer details, and "PLUGINS.md" for a listing of available plugins.
+* Added memory pre-fetching for experimental Ribbon filter, which especially optimizes performance with batched MultiGet.
+* A new, experimental version of BlobDB (key-value separation) is now available. The new implementation is integrated into the RocksDB core, i.e. it is accessible via the usual `rocksdb::DB` API, as opposed to the separate `rocksdb::blob_db::BlobDB` interface used by the earlier version, and can be configured on a per-column family basis using the configuration options `enable_blob_files`, `min_blob_size`, `blob_file_size`, `blob_compression_type`, `enable_blob_garbage_collection`, and `blob_garbage_collection_age_cutoff`. It extends RocksDB's consistency guarantees to blobs, and offers more features and better performance. Note that some features, most notably `Merge`, compaction filters, and backup/restore are not yet supported, and there is no support for migrating a database created by the old implementation.
 
-## 6.17.2 (02/05/2021)
 ### Bug Fixes
 * Since 6.15.0, `TransactionDB` returns error `Status`es from calls to `DeleteRange()` and calls to `Write()` where the `WriteBatch` contains a range deletion. Previously such operations may have succeeded while not providing the expected transactional guarantees. There are certain cases where range deletion can still be used on such DBs; see the API doc on `TransactionDB::DeleteRange()` for details.
 * `OptimisticTransactionDB` now returns error `Status`es from calls to `DeleteRange()` and calls to `Write()` where the `WriteBatch` contains a range deletion. Previously such operations may have succeeded while not providing the expected transactional guarantees.
+* Fix `WRITE_PREPARED`, `WRITE_UNPREPARED` TransactionDB `MultiGet()` may return uncommitted data with snapshot.
+* In DB::OpenForReadOnly, if any error happens while checking Manifest file path, it was overridden by Status::NotFound. It has been fixed and now actual error is returned.
+
+### Public API Change
+* Added a "only_mutable_options" flag to the ConfigOptions.  When this flag is "true", the Configurable functions and convenience methods (such as GetDBOptionsFromString) will only deal with options that are marked as mutable.  When this flag is true, only options marked as mutable can be configured (a Status::InvalidArgument will be returned) and options not marked as mutable will not be returned or compared.  The default is "false", meaning to compare all options.
+* Add new Append and PositionedAppend APIs to FileSystem to bring the data verification information (data checksum information) from upper layer (e.g., WritableFileWriter) to the storage layer. In this way, the customized FileSystem is able to verify the correctness of data being written to the storage on time. Add checksum_handoff_file_types to DBOptions. User can use this option to control which file types (Currently supported file tyes: kWALFile, kTableFile, kDescriptorFile.) should use the new Append and PositionedAppend APIs to handoff the verification information. Currently, RocksDB only use crc32c to calculate the checksum for write handoff.
+* Add an option, `CompressionOptions::max_dict_buffer_bytes`, to limit the in-memory buffering for selecting samples for generating/training a dictionary. The limit is currently loosely adhered to.
 
-## 6.17.1 (01/28/2021)
-### Behavior Changes
-* When retryable IO error occurs during compaction, it is mapped to soft error and set the BG error. However, auto resume is not called to clean the soft error since compaction will reschedule by itself. In this change, When retryable IO error occurs during compaction, BG error is not set. User will be informed the error via EventHelper.
 
 ## 6.17.0 (01/15/2021)
 ### Behavior Changes
@@ -23,6 +83,8 @@
 
 ### Public API Change
 * Add a public API WriteBufferManager::dummy_entries_in_cache_usage() which reports the size of dummy entries stored in cache (passed to WriteBufferManager). Dummy entries are used to account for DataBlocks.
+* Add a SystemClock class that contains the time-related methods from Env.  The original methods in Env may be deprecated in a future release.  This class will allow easier testing, development, and expansion of time-related features.
+* Add a public API GetRocksBuildProperties and GetRocksBuildInfoAsString to get properties about the current build.  These properties may include settings related to the GIT settings (branch, timestamp).  This change also sets the "build date" based on the GIT properties, rather than the actual build time, thereby enabling more reproducible builds.
 
 ## 6.16.0 (12/18/2020)
 ### Behavior Changes

Makefile

@@ -269,6 +269,12 @@ dummy := $(shell (export ROCKSDB_ROOT="$(CURDIR)"; \
 # this file is generated by the previous line to set build flags and sources
 include make_config.mk
 
+ROCKSDB_PLUGIN_MKS = $(foreach plugin, $(ROCKSDB_PLUGINS), plugin/$(plugin)/*.mk)
+include $(ROCKSDB_PLUGIN_MKS)
+ROCKSDB_PLUGIN_SOURCES = $(foreach plugin, $(ROCKSDB_PLUGINS), $(foreach source, $($(plugin)_SOURCES), plugin/$(plugin)/$(source)))
+ROCKSDB_PLUGIN_HEADERS = $(foreach plugin, $(ROCKSDB_PLUGINS), $(foreach header, $($(plugin)_HEADERS), plugin/$(plugin)/$(header)))
+PLATFORM_LDFLAGS += $(foreach plugin, $(ROCKSDB_PLUGINS), $($(plugin)_LDFLAGS))
+
 export JAVAC_ARGS
 CLEAN_FILES += make_config.mk rocksdb.pc
 
@@ -430,6 +436,10 @@ default: all
 WARNING_FLAGS = -W -Wextra -Wall -Wsign-compare -Wshadow \
   -Wunused-parameter
 
+ifeq (,$(filter amd64, $(MACHINE)))
+	C_WARNING_FLAGS = -Wstrict-prototypes
+endif
+
 ifdef USE_CLANG
 	# Used by some teams in Facebook
 	WARNING_FLAGS += -Wshift-sign-overflow
@@ -474,38 +484,14 @@ ifeq ($(NO_THREEWAY_CRC32C), 1)
 	CXXFLAGS += -DNO_THREEWAY_CRC32C
 endif
 
-CFLAGS += $(WARNING_FLAGS) -I. -I./include $(PLATFORM_CCFLAGS) $(OPT)
+CFLAGS += $(C_WARNING_FLAGS) $(WARNING_FLAGS) -I. -I./include $(PLATFORM_CCFLAGS) $(OPT)
 CXXFLAGS += $(WARNING_FLAGS) -I. -I./include $(PLATFORM_CXXFLAGS) $(OPT) -Woverloaded-virtual -Wnon-virtual-dtor -Wno-missing-field-initializers
 
 LDFLAGS += $(PLATFORM_LDFLAGS)
 
-# If NO_UPDATE_BUILD_VERSION is set we don't update util/build_version.cc, but
-# the file needs to already exist or else the build will fail
-ifndef NO_UPDATE_BUILD_VERSION
-date := $(shell date +%F)
-ifdef FORCE_GIT_SHA
-	git_sha := $(FORCE_GIT_SHA)
-else
-	git_sha := $(shell git rev-parse HEAD 2>/dev/null)
-endif
-gen_build_version = sed -e s/@@GIT_SHA@@/$(git_sha)/ -e s/@@GIT_DATE_TIME@@/$(date)/ util/build_version.cc.in
-
-# Record the version of the source that we are compiling.
-# We keep a record of the git revision in this file.  It is then built
-# as a regular source file as part of the compilation process.
-# One can run "strings executable_filename | grep _build_" to find
-# the version of the source that we used to build the executable file.
-FORCE:
-util/build_version.cc: FORCE
-	$(AM_V_GEN)rm -f $@-t
-	$(AM_V_at)$(gen_build_version) > $@-t
-	$(AM_V_at)if test -f $@; then					\
-	  cmp -s $@-t $@ && rm -f $@-t || mv -f $@-t $@;		\
-	else mv -f $@-t $@; fi
-endif
-
 OBJ_DIR?=.
 LIB_OBJECTS = $(patsubst %.cc, $(OBJ_DIR)/%.o, $(LIB_SOURCES))
+LIB_OBJECTS += $(patsubst %.cc, $(OBJ_DIR)/%.o, $(ROCKSDB_PLUGIN_SOURCES))
 ifeq ($(HAVE_POWER8),1)
 LIB_OBJECTS += $(patsubst %.c, $(OBJ_DIR)/%.o, $(LIB_SOURCES_C))
 LIB_OBJECTS += $(patsubst %.S, $(OBJ_DIR)/%.o, $(LIB_SOURCES_ASM))
@@ -616,6 +602,8 @@ ifdef ASSERT_STATUS_CHECKED
 		db_log_iter_test \
 		db_bloom_filter_test \
 		db_blob_basic_test \
+		db_blob_compaction_test \
+		db_blob_corruption_test \
 		db_blob_index_test \
 		db_block_cache_test \
 		db_compaction_test \
@@ -625,6 +613,7 @@ ifdef ASSERT_STATUS_CHECKED
 		db_inplace_update_test \
 		db_io_failure_test \
 		db_iterator_test \
+		db_kv_checksum_test \
 		db_logical_block_size_cache_test \
 		db_memtable_test \
 		db_merge_operand_test \
@@ -778,7 +767,6 @@ endif
 TESTS_PLATFORM_DEPENDENT := \
 	db_basic_test \
 	db_blob_basic_test \
-	db_with_timestamp_basic_test \
 	db_encryption_test \
 	db_test2 \
 	external_sst_file_basic_test \
@@ -869,6 +857,37 @@ ROCKSDB_MAJOR = $(shell egrep "ROCKSDB_MAJOR.[0-9]" include/rocksdb/version.h |
 ROCKSDB_MINOR = $(shell egrep "ROCKSDB_MINOR.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
 ROCKSDB_PATCH = $(shell egrep "ROCKSDB_PATCH.[0-9]" include/rocksdb/version.h | cut -d ' ' -f 3)
 
+# If NO_UPDATE_BUILD_VERSION is set we don't update util/build_version.cc, but
+# the file needs to already exist or else the build will fail
+ifndef NO_UPDATE_BUILD_VERSION
+
+# By default, use the current date-time as the date.  If there are no changes,
+# we will use the last commit date instead.
+build_date := $(shell date "+%Y-%m-%d %T")
+
+ifdef FORCE_GIT_SHA
+	git_sha := $(FORCE_GIT_SHA)
+	git_mod := 1
+	git_date := $(build_date)
+else
+	git_sha := $(shell git rev-parse HEAD 2>/dev/null)
+	git_tag  := $(shell git symbolic-ref -q --short HEAD 2> /dev/null || git describe --tags --exact-match 2>/dev/null)
+	git_mod  := $(shell git diff-index HEAD --quiet 2>/dev/null; echo $$?)
+	git_date := $(shell git log -1 --date=format:"%Y-%m-%d %T" --format="%ad" 2>/dev/null)
+endif
+gen_build_version = sed -e s/@GIT_SHA@/$(git_sha)/ -e s:@GIT_TAG@:"$(git_tag)": -e s/@GIT_MOD@/"$(git_mod)"/ -e s/@BUILD_DATE@/"$(build_date)"/ -e s/@GIT_DATE@/"$(git_date)"/ util/build_version.cc.in
+
+# Record the version of the source that we are compiling.
+# We keep a record of the git revision in this file.  It is then built
+# as a regular source file as part of the compilation process.
+# One can run "strings executable_filename | grep _build_" to find
+# the version of the source that we used to build the executable file.
+util/build_version.cc: $(filter-out $(OBJ_DIR)/util/build_version.o, $(LIB_OBJECTS)) util/build_version.cc.in
+	$(AM_V_GEN)rm -f $@-t
+	$(AM_V_at)$(gen_build_version) > $@
+endif
+CLEAN_FILES += util/build_version.cc
+
 default: all
 
 #-----------------------------------------------
@@ -1032,7 +1051,7 @@ gen_parallel_tests:
 # 107.816 PASS t/DBTest.EncodeDecompressedBlockSizeTest
 #
 slow_test_regexp = \
-	^.*SnapshotConcurrentAccessTest.*$$|^t/run-table_test-HarnessTest.Randomized$$|^t/run-db_test-.*(?:FileCreationRandomFailure|EncodeDecompressedBlockSizeTest)$$|^.*RecoverFromCorruptedWALWithoutFlush$$
+	^.*SnapshotConcurrentAccessTest.*$$|^.*SeqAdvanceConcurrentTest.*$$|^t/run-table_test-HarnessTest.Randomized$$|^t/run-db_test-.*(?:FileCreationRandomFailure|EncodeDecompressedBlockSizeTest)$$|^.*RecoverFromCorruptedWALWithoutFlush$$
 prioritize_long_running_tests =						\
   perl -pe 's,($(slow_test_regexp)),100 $$1,'				\
     | sort -k1,1gr							\
@@ -1411,7 +1430,7 @@ $(STATIC_TEST_LIBRARY): $(TEST_OBJECTS)
 	$(AM_V_AR)rm -f $@ $(SHARED_TEST_LIBRARY)
 	$(AM_V_at)$(AR) $(ARFLAGS) $@ $^
 
-$(STATIC_TOOLS_LIBRARY): $(BENCH_OBJECTS) $(TOOL_OBJECTS)
+$(STATIC_TOOLS_LIBRARY): $(TOOL_OBJECTS)
 	$(AM_V_AR)rm -f $@ $(SHARED_TOOLS_LIBRARY)
 	$(AM_V_at)$(AR) $(ARFLAGS) $@ $^
 
@@ -1560,6 +1579,9 @@ db_basic_test: $(OBJ_DIR)/db/db_basic_test.o $(TEST_LIBRARY) $(LIBRARY)
 db_blob_basic_test: $(OBJ_DIR)/db/blob/db_blob_basic_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
+db_blob_compaction_test: $(OBJ_DIR)/db/blob/db_blob_compaction_test.o $(TEST_LIBRARY) $(LIBRARY)
+	$(AM_LINK)
+
 db_with_timestamp_basic_test: $(OBJ_DIR)/db/db_with_timestamp_basic_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
@@ -1608,6 +1630,9 @@ db_inplace_update_test: $(OBJ_DIR)/db/db_inplace_update_test.o $(TEST_LIBRARY) $
 db_iterator_test: $(OBJ_DIR)/db/db_iterator_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
+db_kv_checksum_test: $(OBJ_DIR)/db/db_kv_checksum_test.o $(TEST_LIBRARY) $(LIBRARY)
+	$(AM_LINK)
+
 db_memtable_test: $(OBJ_DIR)/db/db_memtable_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
@@ -2000,7 +2025,7 @@ range_tombstone_fragmenter_test: $(OBJ_DIR)/db/range_tombstone_fragmenter_test.o
 sst_file_reader_test: $(OBJ_DIR)/table/sst_file_reader_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
-db_secondary_test: $(OBJ_DIR)/db/db_impl/db_secondary_test.o $(TEST_LIBRARY) $(LIBRARY)
+db_secondary_test: $(OBJ_DIR)/db/db_secondary_test.o $(TEST_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
 block_cache_tracer_test: $(OBJ_DIR)/trace_replay/block_cache_tracer_test.o $(TEST_LIBRARY) $(LIBRARY)
@@ -2048,6 +2073,8 @@ io_tracer_parser_test: $(OBJ_DIR)/tools/io_tracer_parser_test.o $(OBJ_DIR)/tools
 io_tracer_parser: $(OBJ_DIR)/tools/io_tracer_parser.o $(TOOLS_LIBRARY) $(LIBRARY)
 	$(AM_LINK)
 
+db_blob_corruption_test: $(OBJ_DIR)/db/blob/db_blob_corruption_test.o $(TEST_LIBRARY) $(LIBRARY)
+	$(AM_LINK)
 #-------------------------------------------------
 # make install related stuff
 PREFIX ?= /usr/local
@@ -2072,6 +2099,10 @@ install-headers: gen-pc
 	for header in `$(FIND) "include/rocksdb" -type f -name *.h`; do \
 		install -C -m 644 $$header $(DESTDIR)/$(PREFIX)/$$header; \
 	done
+	for header in $(ROCKSDB_PLUGIN_HEADERS); do \
+		install -d $(DESTDIR)/$(PREFIX)/include/rocksdb/`dirname $$header`; \
+		install -C -m 644 $$header $(DESTDIR)/$(PREFIX)/include/rocksdb/$$header; \
+	done
 	install -C -m 644 rocksdb.pc $(INSTALL_LIBDIR)/pkgconfig/rocksdb.pc
 
 install-static: install-headers $(LIBRARY)
@@ -2510,6 +2541,8 @@ build_subset_tests: $(ROCKSDBTESTS_SUBSET)
 # working solution.
 ifneq ($(MAKECMDGOALS),clean)
 ifneq ($(MAKECMDGOALS),format)
+ifneq ($(MAKECMDGOALS),check-format)
+ifneq ($(MAKECMDGOALS),check-buck-targets)
 ifneq ($(MAKECMDGOALS),jclean)
 ifneq ($(MAKECMDGOALS),jtest)
 ifneq ($(MAKECMDGOALS),rocksdbjavastatic)
@@ -2525,3 +2558,5 @@ endif
 endif
 endif
 endif
+endif
+endif

PLUGINS.md

@@ -0,0 +1,3 @@
+This is the list of all known third-party plugins for RocksDB. If something is missing, please open a pull request to add it.
+
+* [Dedupfs](https://github.com/ajkr/dedupfs): an example for plugin developers to reference

TARGETS

@@ -1,4 +1,5 @@
-# This file @generated by `python3 buckifier/buckify_rocksdb.py`
+# This file @generated by:
+#$ python3 buckifier/buckify_rocksdb.py
 # --> DO NOT EDIT MANUALLY <--
 # This file is a Facebook-specific integration for buck builds, so can
 # only be validated by Facebook employees.
@@ -24,7 +25,7 @@ ROCKSDB_EXTERNAL_DEPS = [
     ("zlib", None, "z"),
     ("gflags", None, "gflags"),
     ("lz4", None, "lz4"),
-    ("zstd", None),
+    ("zstd", None, "zstd"),
 ]
 
 ROCKSDB_OS_DEPS = [
@@ -86,10 +87,12 @@ ROCKSDB_PREPROCESSOR_FLAGS = [
 
     # Added missing flags from output of build_detect_platform
     "-DROCKSDB_BACKTRACE",
+]
 
-    # Directories with files for #include
-    "-I" + REPO_PATH + "include/",
-    "-I" + REPO_PATH,
+# Directories with files for #include
+ROCKSDB_INCLUDE_PATHS = [
+    "",
+    "include",
 ]
 
 ROCKSDB_ARCH_PREPROCESSOR_FLAGS = {
@@ -206,6 +209,7 @@ cpp_library(
         "db/write_batch_base.cc",
         "db/write_controller.cc",
         "db/write_thread.cc",
+        "env/composite_env.cc",
         "env/env.cc",
         "env/env_chroot.cc",
         "env/env_encryption.cc",
@@ -220,6 +224,7 @@ cpp_library(
         "file/file_prefetch_buffer.cc",
         "file/file_util.cc",
         "file/filename.cc",
+        "file/line_file_reader.cc",
         "file/random_access_file_reader.cc",
         "file/read_write_util.cc",
         "file/readahead_raf.cc",
@@ -340,6 +345,7 @@ cpp_library(
         "util/murmurhash.cc",
         "util/random.cc",
         "util/rate_limiter.cc",
+        "util/ribbon_config.cc",
         "util/slice.cc",
         "util/status.cc",
         "util/string_util.cc",
@@ -424,6 +430,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
     link_whole = False,
@@ -510,6 +517,7 @@ cpp_library(
         "db/write_batch_base.cc",
         "db/write_controller.cc",
         "db/write_thread.cc",
+        "env/composite_env.cc",
         "env/env.cc",
         "env/env_chroot.cc",
         "env/env_encryption.cc",
@@ -524,6 +532,7 @@ cpp_library(
         "file/file_prefetch_buffer.cc",
         "file/file_util.cc",
         "file/filename.cc",
+        "file/line_file_reader.cc",
         "file/random_access_file_reader.cc",
         "file/read_write_util.cc",
         "file/readahead_raf.cc",
@@ -644,6 +653,7 @@ cpp_library(
         "util/murmurhash.cc",
         "util/random.cc",
         "util/rate_limiter.cc",
+        "util/ribbon_config.cc",
         "util/slice.cc",
         "util/status.cc",
         "util/string_util.cc",
@@ -728,6 +738,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
     link_whole = True,
@@ -751,6 +762,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [":rocksdb_lib"],
     external_deps = ROCKSDB_EXTERNAL_DEPS + [
         ("googletest", None, "gtest"),
@@ -772,6 +784,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [":rocksdb_lib"],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
     link_whole = False,
@@ -799,6 +812,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = ROCKSDB_LIB_DEPS,
     external_deps = ROCKSDB_EXTERNAL_DEPS,
 )
@@ -810,6 +824,7 @@ cpp_binary(
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     compiler_flags = ROCKSDB_COMPILER_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [":rocksdb_test_lib"],
 ) if not is_opt_mode else None
 
@@ -831,6 +846,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [":rocksdb_test_lib"],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
     link_whole = False,
@@ -1146,6 +1162,20 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "db_blob_compaction_test",
+        "db/blob/db_blob_compaction_test.cc",
+        "serial",
+        [],
+        [],
+    ],
+    [
+        "db_blob_corruption_test",
+        "db/blob/db_blob_corruption_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "db_blob_index_test",
         "db/blob/db_blob_index_test.cc",
@@ -1237,6 +1267,13 @@ ROCKS_TESTS = [
         [],
         [],
     ],
+    [
+        "db_kv_checksum_test",
+        "db/db_kv_checksum_test.cc",
+        "serial",
+        [],
+        [],
+    ],
     [
         "db_log_iter_test",
         "db/db_log_iter_test.cc",
@@ -1295,7 +1332,7 @@ ROCKS_TESTS = [
     ],
     [
         "db_secondary_test",
-        "db/db_impl/db_secondary_test.cc",
+        "db/db_secondary_test.cc",
         "serial",
         [],
         [],
@@ -2090,6 +2127,7 @@ ROCKS_TESTS = [
         os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
         compiler_flags = ROCKSDB_COMPILER_FLAGS + extra_compiler_flags,
         preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+        include_paths = ROCKSDB_INCLUDE_PATHS,
         deps = [":rocksdb_test_lib"] + extra_deps,
         external_deps = ROCKSDB_EXTERNAL_DEPS + [
             ("googletest", None, "gtest"),

USERS.md

@@ -26,6 +26,9 @@ Learn more about those use cases in a Tech Talk by Ankit Gupta and Naveen Somasu
 ## Yahoo
 Yahoo is using RocksDB as a storage engine for their biggest distributed data store Sherpa. Learn more about it here: http://yahooeng.tumblr.com/post/120730204806/sherpa-scales-new-heights
 
+## Baidu
+[Apache Doris](http://doris.apache.org/master/en/) is a MPP analytical database engine released by Baidu. It [uses RocksDB](http://doris.apache.org/master/en/administrator-guide/operation/tablet-meta-tool.html) to manage its tablet's metadata.
+
 ## CockroachDB
 CockroachDB is an open-source geo-replicated transactional database. They are using RocksDB as their storage engine. Check out their github: https://github.com/cockroachdb/cockroach
 
@@ -67,7 +70,7 @@ Pinterest's Object Retrieval System uses RocksDB for storage: https://www.youtub
 [VWO's](https://vwo.com/) Smart Code checker and URL helper uses RocksDB to store all the URLs where VWO's Smart Code is installed.
 
 ## quasardb
-[quasardb](https://www.quasardb.net) is a high-performance, distributed, transactional key-value database that integrates well with in-memory analytics engines such as Apache Spark. 
+[quasardb](https://www.quasardb.net) is a high-performance, distributed, transactional key-value database that integrates well with in-memory analytics engines such as Apache Spark.
 quasardb uses a heavily tuned RocksDB as its persistence layer.
 
 ## Netflix
@@ -96,10 +99,10 @@ LzLabs is using RocksDB as a storage engine in their multi-database distributed
 
 ## IOTA Foundation
  [IOTA Foundation](https://www.iota.org/) is using RocksDB in the [IOTA Reference Implementation (IRI)](https://github.com/iotaledger/iri) to store the local state of the Tangle. The Tangle is the first open-source distributed ledger powering the future of the Internet of Things.
- 
+
 ## Avrio Project
  [Avrio Project](http://avrio-project.github.io/avrio.network/) is using RocksDB in [Avrio ](https://github.com/avrio-project/avrio) to store blocks, account balances and data and other blockchain-releated data. Avrio is a multiblockchain decentralized cryptocurrency empowering monetary transactions.
- 
+
 ## 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.
 

buckifier/buckify_rocksdb.py

@@ -24,10 +24,10 @@ from util import ColorString
 # (This generates a TARGET file without user-specified dependency for unit
 # tests.)
 # $python3 buckifier/buckify_rocksdb.py \
-#        '{"fake": { \
-#                      "extra_deps": [":test_dep", "//fakes/module:mock1"],  \
-#                      "extra_compiler_flags": ["-DROCKSDB_LITE", "-Os"], \
-#                  } \
+#        '{"fake": {
+#                      "extra_deps": [":test_dep", "//fakes/module:mock1"],
+#                      "extra_compiler_flags": ["-DROCKSDB_LITE", "-Os"]
+#                  }
 #         }'
 # (Generated TARGETS file has test_dep and mock1 as dependencies for RocksDB
 # unit tests, and will use the extra_compiler_flags to compile the unit test
@@ -129,7 +129,13 @@ def generate_targets(repo_path, deps_map):
     if src_mk is None or cc_files is None or parallel_tests is None:
         return False
 
-    TARGETS = TARGETSBuilder("%s/TARGETS" % repo_path)
+    extra_argv = ""
+    if len(sys.argv) >= 2:
+        # Heuristically quote and canonicalize whitespace for inclusion
+        # in how the file was generated.
+        extra_argv = " '{0}'".format(" ".join(sys.argv[1].split()))
+
+    TARGETS = TARGETSBuilder("%s/TARGETS" % repo_path, extra_argv)
 
     # rocksdb_lib
     TARGETS.add_library(

buckifier/targets_builder.py

@@ -25,10 +25,11 @@ def pretty_list(lst, indent=8):
 
 
 class TARGETSBuilder(object):
-    def __init__(self, path):
+    def __init__(self, path, extra_argv):
         self.path = path
         self.targets_file = open(path, 'wb')
-        header = targets_cfg.rocksdb_target_header_template
+        header = targets_cfg.rocksdb_target_header_template.format(
+            extra_argv=extra_argv)
         self.targets_file.write(header.encode("utf-8"))
         self.total_lib = 0
         self.total_bin = 0
@@ -86,6 +87,7 @@ cpp_binary(
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     compiler_flags = ROCKSDB_COMPILER_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [":rocksdb_test_lib"],
 ) if not is_opt_mode else None
 

buckifier/targets_cfg.py

@@ -5,7 +5,8 @@ from __future__ import print_function
 from __future__ import unicode_literals
 
 rocksdb_target_header_template = \
-    """# This file \100generated by `python3 buckifier/buckify_rocksdb.py`
+    """# This file \100generated by:
+#$ python3 buckifier/buckify_rocksdb.py{extra_argv}
 # --> DO NOT EDIT MANUALLY <--
 # This file is a Facebook-specific integration for buck builds, so can
 # only be validated by Facebook employees.
@@ -31,7 +32,7 @@ ROCKSDB_EXTERNAL_DEPS = [
     ("zlib", None, "z"),
     ("gflags", None, "gflags"),
     ("lz4", None, "lz4"),
-    ("zstd", None),
+    ("zstd", None, "zstd"),
 ]
 
 ROCKSDB_OS_DEPS = [
@@ -93,17 +94,19 @@ ROCKSDB_PREPROCESSOR_FLAGS = [
 
     # Added missing flags from output of build_detect_platform
     "-DROCKSDB_BACKTRACE",
-
-    # Directories with files for #include
-    "-I" + REPO_PATH + "include/",
-    "-I" + REPO_PATH,
 ]
 
-ROCKSDB_ARCH_PREPROCESSOR_FLAGS = {
+# Directories with files for #include
+ROCKSDB_INCLUDE_PATHS = [
+    "",
+    "include",
+]
+
+ROCKSDB_ARCH_PREPROCESSOR_FLAGS = {{
     "x86_64": [
         "-DHAVE_PCLMUL",
     ],
-}
+}}
 
 build_mode = read_config("fbcode", "build_mode")
 
@@ -144,6 +147,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [{deps}],
     external_deps = ROCKSDB_EXTERNAL_DEPS{extra_external_deps},
     link_whole = {link_whole},
@@ -160,6 +164,7 @@ cpp_library(
     os_deps = ROCKSDB_OS_DEPS,
     os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = ROCKSDB_LIB_DEPS,
     external_deps = ROCKSDB_EXTERNAL_DEPS,
 )
@@ -172,6 +177,7 @@ cpp_binary(
     arch_preprocessor_flags = ROCKSDB_ARCH_PREPROCESSOR_FLAGS,
     compiler_flags = ROCKSDB_COMPILER_FLAGS,
     preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+    include_paths = ROCKSDB_INCLUDE_PATHS,
     deps = [{deps}],
     external_deps = ROCKSDB_EXTERNAL_DEPS,
 )
@@ -202,6 +208,7 @@ ROCKS_TESTS = [
         os_preprocessor_flags = ROCKSDB_OS_PREPROCESSOR_FLAGS,
         compiler_flags = ROCKSDB_COMPILER_FLAGS + extra_compiler_flags,
         preprocessor_flags = ROCKSDB_PREPROCESSOR_FLAGS,
+        include_paths = ROCKSDB_INCLUDE_PATHS,
         deps = [":rocksdb_test_lib"] + extra_deps,
         external_deps = ROCKSDB_EXTERNAL_DEPS + [
             ("googletest", None, "gtest"),

build_tools/format-diff.sh

@@ -52,7 +52,8 @@ else
     else
       echo "You didn't have clang-format-diff.py and/or clang-format available in your computer!"
       echo "You can download clang-format-diff.py by running: "
-      echo "    curl --location https://tinyurl.com/y2kvokof -o ${REPO_ROOT}/clang-format-diff.py"
+      echo "    curl --location https://raw.githubusercontent.com/llvm/llvm-project/main/clang/tools/clang-format/clang-format-diff.py -o ${REPO_ROOT}/clang-format-diff.py"
+      echo "You should make sure the downloaded script is not compromised."
       echo "You can download clang-format by running:"
       echo "    brew install clang-format"
       echo "  Or"
@@ -82,7 +83,8 @@ else
       echo "You have clang-format-diff.py for Python 2 but are using a Python 3"
       echo "interpreter (${PYTHON:-python3})."
       echo "You can download clang-format-diff.py for Python 3 by running: "
-      echo "    curl --location https://tinyurl.com/y2kvokof -o ${REPO_ROOT}/clang-format-diff.py"
+      echo "    curl --location https://raw.githubusercontent.com/llvm/llvm-project/main/clang/tools/clang-format/clang-format-diff.py -o ${REPO_ROOT}/clang-format-diff.py"
+      echo "You should make sure the downloaded script is not compromised."
       exit 130
     fi
     CLANG_FORMAT_DIFF="${PYTHON:-python3} $CFD_PATH"

build_tools/rocksdb-lego-determinator

@@ -3,7 +3,7 @@
 # to determine next steps to run
 
 # Usage:
-# EMAIL=<email> ONCALL=<email> TRIGGER=<trigger> SUBSCRIBER=<email> rocks_ci.py <test-name>
+# EMAIL=<email> ONCALL=<email> TRIGGER=<trigger> SUBSCRIBER=<email> WORKINGDIR=<working_dir> rocksdb-lego-determinator <test-name>
 #
 # Input         Value
 # -------------------------------------------------------------------------
@@ -11,7 +11,7 @@
 # ONCALL        Email address to raise a task on failure
 # TRIGGER       Trigger conditions for email. Valid values are fail, warn, all
 # SUBSCRIBER    Email addresss to add as subscriber for task
-#
+# WORKINGDIR    Working directory
 
 #
 # Report configuration
@@ -53,13 +53,19 @@ if [[ ! -z $REPORT_EMAIL || ! -z $CREATE_TASK ]]; then
   ]"
 fi
 
+# Working directory for the following command, default to current directory
+WORKING_DIR=.
+if [ ! -z $WORKINGDIR ]; then
+  WORKING_DIR=$WORKINGDIR
+fi
+
 #
 # Helper variables
 #
 CLEANUP_ENV="
 {
     'name':'Cleanup environment',
-    'shell':'rm -rf /dev/shm/rocksdb && mkdir /dev/shm/rocksdb && (chmod +t /dev/shm || true)  && make clean',
+    'shell':'cd $WORKING_DIR; rm -rf /dev/shm/rocksdb && mkdir /dev/shm/rocksdb && (chmod +t /dev/shm || true) && make clean',
     'user':'root'
 }"
 
@@ -112,6 +118,7 @@ SETUP_JAVA_ENV="export $HTTP_PROXY; export JAVA_HOME=/usr/local/jdk-8u60-64/; ex
 PARSER="'parser':'python build_tools/error_filter.py $1'"
 
 CONTRUN_NAME="ROCKSDB_CONTRUN_NAME"
+SKIP_FORMAT_CHECKS="SKIP_FORMAT_BUCK_CHECKS=1"
 
 # This code is getting called under various scenarios. What we care about is to
 # understand when it's called from nightly contruns because in that case we'll
@@ -153,7 +160,7 @@ UNIT_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and test RocksDB debug version',
-                'shell':'$SHM $DEBUG make $PARALLELISM check || $CONTRUN_NAME=check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM check || $CONTRUN_NAME=check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -176,7 +183,7 @@ UNIT_TEST_NON_SHM_COMMANDS="[
             {
                 'name':'Build and test RocksDB debug version',
                 'timeout': 86400,
-                'shell':'$NON_SHM $DEBUG make $PARALLELISM check || $CONTRUN_NAME=non_shm_check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $NON_SHM $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM check || $CONTRUN_NAME=non_shm_check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -197,7 +204,7 @@ RELEASE_BUILD_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build RocksDB release',
-                'shell':'make $PARALLEL_j release || $CONTRUN_NAME=release $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; make $PARALLEL_j release || $CONTRUN_NAME=release $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -218,7 +225,7 @@ UNIT_TEST_COMMANDS_481="[
             $CLEANUP_ENV,
             {
                 'name':'Build and test RocksDB debug version',
-                'shell':'$SHM $GCC_481 $DEBUG make $PARALLELISM check || $CONTRUN_NAME=unit_gcc_481_check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $GCC_481 $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM check || $CONTRUN_NAME=unit_gcc_481_check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -239,7 +246,7 @@ RELEASE_BUILD_COMMANDS_481="[
             $CLEANUP_ENV,
             {
                 'name':'Build RocksDB release on GCC 4.8.1',
-                'shell':'$GCC_481 make $PARALLEL_j release || $CONTRUN_NAME=release_gcc481 $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $GCC_481 make $PARALLEL_j release || $CONTRUN_NAME=release_gcc481 $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -260,7 +267,7 @@ CLANG_UNIT_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and test RocksDB debug',
-                'shell':'$CLANG $SHM $DEBUG make $PARALLELISM check || $CONTRUN_NAME=clang_check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $CLANG $SHM $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM check || $CONTRUN_NAME=clang_check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -281,7 +288,7 @@ CLANG_RELEASE_BUILD_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build RocksDB release',
-                'shell':'$CLANG make $PARALLEL_j release|| $CONTRUN_NAME=clang_release $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $CLANG make $PARALLEL_j release|| $CONTRUN_NAME=clang_release $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -302,7 +309,7 @@ CLANG_ANALYZE_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'RocksDB build and analyze',
-                'shell':'$CLANG $SHM $DEBUG make $PARALLEL_j analyze || $CONTRUN_NAME=clang_analyze $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $CLANG $SHM $DEBUG make $PARALLEL_j analyze || $CONTRUN_NAME=clang_analyze $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -323,7 +330,7 @@ CODE_COV_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build, test and collect code coverage info',
-                'shell':'$SHM $DEBUG make $PARALLELISM coverage || $CONTRUN_NAME=coverage $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM coverage || $CONTRUN_NAME=coverage $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -344,7 +351,7 @@ UNITY_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build, test unity test',
-                'shell':'$SHM $DEBUG V=1 make J=1 unity_test || $CONTRUN_NAME=unity_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG V=1 make J=1 unity_test || $CONTRUN_NAME=unity_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -365,7 +372,7 @@ LITE_BUILD_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build RocksDB debug version',
-                'shell':'make J=1 LITE=1 all check || $CONTRUN_NAME=lite $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SKIP_FORMAT_CHECKS make J=1 LITE=1 all check || $CONTRUN_NAME=lite $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -385,7 +392,7 @@ REPORT_LITE_BINARY_SIZE_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Report RocksDB Lite binary size to scuba',
-                'shell':'tools/report_lite_binary_size.sh',
+                'shell':'cd $WORKING_DIR; tools/report_lite_binary_size.sh',
                 'user':'root',
             },
         ],
@@ -404,14 +411,14 @@ STRESS_CRASH_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and run RocksDB debug stress tests',
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
             {
                 'name':'Build and run RocksDB debug crash tests',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test || $CONTRUN_NAME=crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test || $CONTRUN_NAME=crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -434,14 +441,14 @@ BLACKBOX_STRESS_CRASH_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and run RocksDB debug stress tests',
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
             {
                 'name':'Build and run RocksDB debug blackbox crash tests',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 blackbox_crash_test || $CONTRUN_NAME=blackbox_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 blackbox_crash_test || $CONTRUN_NAME=blackbox_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -464,14 +471,14 @@ WHITEBOX_STRESS_CRASH_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and run RocksDB debug stress tests',
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
             {
                 'name':'Build and run RocksDB debug whitebox crash tests',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 whitebox_crash_test || $CONTRUN_NAME=whitebox_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 whitebox_crash_test || $CONTRUN_NAME=whitebox_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -494,14 +501,14 @@ STRESS_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and run RocksDB debug stress tests',
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
             {
                 'name':'Build and run RocksDB debug crash tests with atomic flush',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test_with_atomic_flush || $CONTRUN_NAME=crash_test_with_atomic_flush $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test_with_atomic_flush || $CONTRUN_NAME=crash_test_with_atomic_flush $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -524,14 +531,14 @@ STRESS_CRASH_TEST_WITH_TXN_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and run RocksDB debug stress tests',
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
             {
                 'name':'Build and run RocksDB debug crash tests with txn',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test_with_txn || $CONTRUN_NAME=crash_test_with_txn $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH make J=1 crash_test_with_txn || $CONTRUN_NAME=crash_test_with_txn $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -553,7 +560,7 @@ WRITE_STRESS_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build and run RocksDB write stress tests',
-                'shell':'make write_stress && python tools/write_stress_runner.py --runtime_sec=3600 --db=/tmp/rocksdb_write_stress || $CONTRUN_NAME=write_stress $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; make write_stress && python tools/write_stress_runner.py --runtime_sec=3600 --db=/tmp/rocksdb_write_stress || $CONTRUN_NAME=write_stress $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             }
@@ -576,7 +583,7 @@ ASAN_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Test RocksDB debug under ASAN',
-'shell':'set -o pipefail && ($SHM $ASAN $DEBUG make $PARALLELISM asan_check || $CONTRUN_NAME=asan_check $TASK_CREATION_TOOL) |& /usr/facebook/ops/scripts/asan_symbolize.py -d',
+'shell':'cd $WORKING_DIR; set -o pipefail && ($SHM $ASAN $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM asan_check || $CONTRUN_NAME=asan_check $TASK_CREATION_TOOL) |& /usr/facebook/ops/scripts/asan_symbolize.py -d',
                 'user':'root',
                 $PARSER
             }
@@ -599,7 +606,7 @@ ASAN_CRASH_TEST_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug asan_crash_test',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 asan_crash_test || $CONTRUN_NAME=asan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $SKIP_FORMAT_CHECKS make J=1 asan_crash_test || $CONTRUN_NAME=asan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -623,7 +630,7 @@ ASAN_BLACKBOX_CRASH_TEST_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug blackbox asan_crash_test',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 blackbox_asan_crash_test || $CONTRUN_NAME=blackbox_asan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $SKIP_FORMAT_CHECKS make J=1 blackbox_asan_crash_test || $CONTRUN_NAME=blackbox_asan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -647,7 +654,7 @@ ASAN_WHITEBOX_CRASH_TEST_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug whitebox asan_crash_test',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 whitebox_asan_crash_test || $CONTRUN_NAME=whitebox_asan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $SKIP_FORMAT_CHECKS make J=1 whitebox_asan_crash_test || $CONTRUN_NAME=whitebox_asan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -671,7 +678,7 @@ ASAN_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug asan_crash_test_with_atomic_flush',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 asan_crash_test_with_atomic_flush || $CONTRUN_NAME=asan_crash_test_with_atomic_flush $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $SKIP_FORMAT_CHECKS make J=1 asan_crash_test_with_atomic_flush || $CONTRUN_NAME=asan_crash_test_with_atomic_flush $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -695,7 +702,7 @@ ASAN_CRASH_TEST_WITH_TXN_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug asan_crash_test_with_txn',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 asan_crash_test_with_txn || $CONTRUN_NAME=asan_crash_test_with_txn $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $SKIP_FORMAT_CHECKS make J=1 asan_crash_test_with_txn || $CONTRUN_NAME=asan_crash_test_with_txn $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -717,7 +724,7 @@ UBSAN_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Test RocksDB debug under UBSAN',
-                'shell':'set -o pipefail && $SHM $UBSAN $CLANG $DEBUG make $PARALLELISM ubsan_check || $CONTRUN_NAME=ubsan_check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $UBSAN $CLANG $DEBUG $SKIP_FORMAT_CHECKS make $PARALLELISM ubsan_check || $CONTRUN_NAME=ubsan_check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             }
@@ -740,7 +747,7 @@ UBSAN_CRASH_TEST_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug ubsan_crash_test',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 ubsan_crash_test || $CONTRUN_NAME=ubsan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $CLANG $SKIP_FORMAT_CHECKS make J=1 ubsan_crash_test || $CONTRUN_NAME=ubsan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -764,7 +771,7 @@ UBSAN_BLACKBOX_CRASH_TEST_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug blackbox ubsan_crash_test',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 blackbox_ubsan_crash_test || $CONTRUN_NAME=blackbox_ubsan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $CLANG $SKIP_FORMAT_CHECKS make J=1 blackbox_ubsan_crash_test || $CONTRUN_NAME=blackbox_ubsan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -788,7 +795,7 @@ UBSAN_WHITEBOX_CRASH_TEST_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug whitebox ubsan_crash_test',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 whitebox_ubsan_crash_test || $CONTRUN_NAME=whitebox_ubsan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $CLANG $SKIP_FORMAT_CHECKS make J=1 whitebox_ubsan_crash_test || $CONTRUN_NAME=whitebox_ubsan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -812,7 +819,7 @@ UBSAN_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug ubsan_crash_test_with_atomic_flush',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 ubsan_crash_test_with_atomic_flush || $CONTRUN_NAME=ubsan_crash_test_with_atomic_flush $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $CLANG $SKIP_FORMAT_CHECKS make J=1 ubsan_crash_test_with_atomic_flush || $CONTRUN_NAME=ubsan_crash_test_with_atomic_flush $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -836,7 +843,7 @@ UBSAN_CRASH_TEST_WITH_TXN_COMMANDS="[
             {
                 'name':'Build and run RocksDB debug ubsan_crash_test_with_txn',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH $CLANG make J=1 ubsan_crash_test_with_txn || $CONTRUN_NAME=ubsan_crash_test_with_txn $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG $NON_TSAN_CRASH $CLANG $SKIP_FORMAT_CHECKS make J=1 ubsan_crash_test_with_txn || $CONTRUN_NAME=ubsan_crash_test_with_txn $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -860,7 +867,7 @@ VALGRIND_TEST_COMMANDS="[
             {
                 'name':'Run RocksDB debug unit tests',
                 'timeout': 86400,
-                'shell':'$SHM $DEBUG make $PARALLELISM valgrind_test || $CONTRUN_NAME=valgrind_check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SHM $DEBUG make $PARALLELISM valgrind_test || $CONTRUN_NAME=valgrind_check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -883,7 +890,7 @@ TSAN_UNIT_TEST_COMMANDS="[
             {
                 'name':'Run RocksDB debug unit test',
                 'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN make $PARALLELISM check || $CONTRUN_NAME=tsan_check $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $DEBUG $TSAN $SKIP_FORMAT_CHECKS make $PARALLELISM check || $CONTRUN_NAME=tsan_check $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -906,7 +913,7 @@ TSAN_CRASH_TEST_COMMANDS="[
             {
                 'name':'Compile and run',
                 'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 crash_test || $CONTRUN_NAME=tsan_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 crash_test || $CONTRUN_NAME=tsan_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -930,7 +937,7 @@ TSAN_BLACKBOX_CRASH_TEST_COMMANDS="[
             {
                 'name':'Compile and run',
                 'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 blackbox_crash_test || $CONTRUN_NAME=tsan_blackbox_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 blackbox_crash_test || $CONTRUN_NAME=tsan_blackbox_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -954,7 +961,7 @@ TSAN_WHITEBOX_CRASH_TEST_COMMANDS="[
             {
                 'name':'Compile and run',
                 'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 whitebox_crash_test || $CONTRUN_NAME=tsan_whitebox_crash_test $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 whitebox_crash_test || $CONTRUN_NAME=tsan_whitebox_crash_test $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -978,7 +985,7 @@ TSAN_CRASH_TEST_WITH_ATOMIC_FLUSH_COMMANDS="[
             {
                 'name':'Compile and run',
                 'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 crash_test_with_atomic_flush || $CONTRUN_NAME=tsan_crash_test_with_atomic_flush $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 crash_test_with_atomic_flush || $CONTRUN_NAME=tsan_crash_test_with_atomic_flush $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -1002,7 +1009,7 @@ TSAN_CRASH_TEST_WITH_TXN_COMMANDS="[
             {
                 'name':'Compile and run',
                 'timeout': 86400,
-                'shell':'set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 crash_test_with_txn || $CONTRUN_NAME=tsan_crash_test_with_txn $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; set -o pipefail && $SHM $DEBUG $TSAN $TSAN_CRASH CRASH_TEST_KILL_ODD=1887 make J=1 crash_test_with_txn || $CONTRUN_NAME=tsan_crash_test_with_txn $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -1036,7 +1043,7 @@ FORMAT_COMPATIBLE_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Run RocksDB debug unit test',
-                'shell':'build_tools/rocksdb-lego-determinator run_format_compatible || $CONTRUN_NAME=run_format_compatible $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; build_tools/rocksdb-lego-determinator run_format_compatible || $CONTRUN_NAME=run_format_compatible $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -1058,6 +1065,7 @@ run_no_compression()
   mv .tmp.fbcode_config.sh build_tools/fbcode_config.sh
   cat Makefile | grep -v tools/ldb_test.py > .tmp.Makefile
   mv .tmp.Makefile Makefile
+  export $SKIP_FORMAT_CHECKS
   make $DEBUG J=1 check
 }
 
@@ -1070,7 +1078,7 @@ NO_COMPRESSION_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Run RocksDB debug unit test',
-                'shell':'build_tools/rocksdb-lego-determinator run_no_compression || $CONTRUN_NAME=run_no_compression $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; build_tools/rocksdb-lego-determinator run_no_compression || $CONTRUN_NAME=run_no_compression $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -1129,7 +1137,7 @@ REGRESSION_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Make and run script',
-                'shell':'build_tools/rocksdb-lego-determinator run_regression || $CONTRUN_NAME=run_regression $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; build_tools/rocksdb-lego-determinator run_regression || $CONTRUN_NAME=run_regression $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },
@@ -1150,7 +1158,7 @@ JAVA_BUILD_TEST_COMMANDS="[
             $CLEANUP_ENV,
             {
                 'name':'Build RocksDB for Java',
-                'shell':'$SETUP_JAVA_ENV; $SHM make rocksdbjava || $CONTRUN_NAME=rocksdbjava $TASK_CREATION_TOOL',
+                'shell':'cd $WORKING_DIR; $SETUP_JAVA_ENV; $SHM make rocksdbjava || $CONTRUN_NAME=rocksdbjava $TASK_CREATION_TOOL',
                 'user':'root',
                 $PARSER
             },

build_tools/run_ci_db_test.ps1

@@ -68,7 +68,7 @@ $BinariesFolder = -Join($RootFolder, "\build\Debug\")
 
 if($WorkFolder -eq "") {
 
-    # If TEST_TMPDIR is set use it    
+    # If TEST_TMPDIR is set use it
     [string]$var = $Env:TEST_TMPDIR
     if($var -eq "") {
         $WorkFolder = -Join($RootFolder, "\db_tests\")
@@ -93,7 +93,7 @@ $ExcludeCasesSet = New-Object System.Collections.Generic.HashSet[string]
 if($ExcludeCases -ne "") {
     Write-Host "ExcludeCases: $ExcludeCases"
     $l = $ExcludeCases -split ' '
-    ForEach($t in $l) { 
+    ForEach($t in $l) {
       $ExcludeCasesSet.Add($t) | Out-Null
     }
 }
@@ -102,7 +102,7 @@ $ExcludeExesSet = New-Object System.Collections.Generic.HashSet[string]
 if($ExcludeExes -ne "") {
     Write-Host "ExcludeExe: $ExcludeExes"
     $l = $ExcludeExes -split ' '
-    ForEach($t in $l) { 
+    ForEach($t in $l) {
       $ExcludeExesSet.Add($t) | Out-Null
     }
 }
@@ -118,6 +118,10 @@ if($ExcludeExes -ne "") {
 #   MultiThreaded/MultiThreadedDBTest.
 #     MultiThreaded/0  # GetParam() = 0
 #     MultiThreaded/1  # GetParam() = 1
+#   RibbonTypeParamTest/0.  # TypeParam = struct DefaultTypesAndSettings
+#     CompactnessAndBacktrackAndFpRate
+#     Extremes
+#     FindOccupancyForSuccessRate
 #
 # into this:
 #
@@ -125,6 +129,9 @@ if($ExcludeExes -ne "") {
 #   DBTest.WriteEmptyBatch
 #   MultiThreaded/MultiThreadedDBTest.MultiThreaded/0
 #   MultiThreaded/MultiThreadedDBTest.MultiThreaded/1
+#   RibbonTypeParamTest/0.CompactnessAndBacktrackAndFpRate
+#   RibbonTypeParamTest/0.Extremes
+#   RibbonTypeParamTest/0.FindOccupancyForSuccessRate
 #
 # Output into the parameter in a form TestName -> Log File Name
 function ExtractTestCases([string]$GTestExe, $HashTable) {
@@ -138,6 +145,8 @@ function ExtractTestCases([string]$GTestExe, $HashTable) {
 
     ForEach( $l in $Tests) {
 
+      # remove trailing comment if any
+      $l = $l -replace '\s+\#.*',''
       # Leading whitespace is fine
       $l = $l -replace '^\s+',''
       # Trailing dot is a test group but no whitespace
@@ -146,8 +155,7 @@ function ExtractTestCases([string]$GTestExe, $HashTable) {
       }  else {
         # Otherwise it is a test name, remove leading space
         $test = $l
-        # remove trailing comment if any and create a log name
-        $test = $test -replace '\s+\#.*',''
+        # create a log name
         $test = "$Group$test"
 
         if($ExcludeCasesSet.Contains($test)) {
@@ -253,7 +261,7 @@ if($Run -ne "") {
 
   $DiscoveredExe = @()
   dir -Path $search_path | ForEach-Object {
-     $DiscoveredExe += ($_.Name)     
+     $DiscoveredExe += ($_.Name)
   }
 
   # Remove exclusions
@@ -293,7 +301,7 @@ if($SuiteRun -ne "") {
 
   $ListOfExe = @()
   dir -Path $search_path | ForEach-Object {
-     $ListOfExe += ($_.Name)     
+     $ListOfExe += ($_.Name)
   }
 
   # Exclude those in RunOnly from running as suites
@@ -348,7 +356,7 @@ function RunJobs($Suites, $TestCmds, [int]$ConcurrencyVal)
 
     # Wait for all to finish and get the results
     while(($JobToLog.Count -gt 0) -or
-          ($TestCmds.Count -gt 0) -or 
+          ($TestCmds.Count -gt 0) -or
            ($Suites.Count -gt 0)) {
 
         # Make sure we have maximum concurrent jobs running if anything
@@ -468,8 +476,8 @@ RunJobs -Suites $CasesToRun -TestCmds $TestExes -ConcurrencyVal $Concurrency
 
 $EndDate = (Get-Date)
 
-New-TimeSpan -Start $StartDate -End $EndDate | 
-  ForEach-Object { 
+New-TimeSpan -Start $StartDate -End $EndDate |
+  ForEach-Object {
     "Elapsed time: {0:g}" -f $_
   }
 
@@ -484,4 +492,4 @@ if(!$script:success) {
 
  exit 0
 
- 
+

cache/cache_bench.cc

@@ -11,15 +11,17 @@ int main() {
 }
 #else
 
-#include <stdio.h>
 #include <sys/types.h>
+
 #include <cinttypes>
+#include <cstdio>
 #include <limits>
 
 #include "port/port.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
+#include "rocksdb/system_clock.h"
 #include "util/coding.h"
 #include "util/gflags_compat.h"
 #include "util/hash.h"
@@ -210,6 +212,7 @@ class CacheBench {
 
   bool Run() {
     ROCKSDB_NAMESPACE::Env* env = ROCKSDB_NAMESPACE::Env::Default();
+    const auto& clock = env->GetSystemClock();
 
     PrintEnv();
     SharedState shared(this);
@@ -224,7 +227,7 @@ class CacheBench {
         shared.GetCondVar()->Wait();
       }
       // Record start time
-      uint64_t start_time = env->NowMicros();
+      uint64_t start_time = clock->NowMicros();
 
       // Start all threads
       shared.SetStart();
@@ -236,7 +239,7 @@ class CacheBench {
       }
 
       // Record end time
-      uint64_t end_time = env->NowMicros();
+      uint64_t end_time = clock->NowMicros();
       double elapsed = static_cast<double>(end_time - start_time) * 1e-6;
       uint32_t qps = static_cast<uint32_t>(
           static_cast<double>(FLAGS_threads * FLAGS_ops_per_thread) / elapsed);

cache/lru_cache.cc

@@ -9,8 +9,8 @@
 
 #include "cache/lru_cache.h"
 
-#include <assert.h>
-#include <stdio.h>
+#include <cassert>
+#include <cstdio>
 #include <string>
 
 #include "util/mutexlock.h"

db/blob/blob_file_builder.cc

@@ -8,6 +8,7 @@
 #include <cassert>
 
 #include "db/blob/blob_file_addition.h"
+#include "db/blob/blob_file_completion_callback.h"
 #include "db/blob/blob_index.h"
 #include "db/blob/blob_log_format.h"
 #include "db/blob/blob_log_writer.h"
@@ -17,39 +18,44 @@
 #include "file/writable_file_writer.h"
 #include "logging/logging.h"
 #include "options/cf_options.h"
+#include "options/options_helper.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/status.h"
 #include "test_util/sync_point.h"
+#include "trace_replay/io_tracer.h"
 #include "util/compression.h"
 
 namespace ROCKSDB_NAMESPACE {
 
 BlobFileBuilder::BlobFileBuilder(
-    VersionSet* versions, Env* env, FileSystem* fs,
+    VersionSet* versions, FileSystem* fs,
     const ImmutableCFOptions* immutable_cf_options,
     const MutableCFOptions* mutable_cf_options, const FileOptions* file_options,
     int job_id, uint32_t column_family_id,
     const std::string& column_family_name, Env::IOPriority io_priority,
     Env::WriteLifeTimeHint write_hint,
+    const std::shared_ptr<IOTracer>& io_tracer,
+    BlobFileCompletionCallback* blob_callback,
     std::vector<std::string>* blob_file_paths,
     std::vector<BlobFileAddition>* blob_file_additions)
-    : BlobFileBuilder([versions]() { return versions->NewFileNumber(); }, env,
-                      fs, immutable_cf_options, mutable_cf_options,
-                      file_options, job_id, column_family_id,
-                      column_family_name, io_priority, write_hint,
-                      blob_file_paths, blob_file_additions) {}
+    : BlobFileBuilder([versions]() { return versions->NewFileNumber(); }, fs,
+                      immutable_cf_options, mutable_cf_options, file_options,
+                      job_id, column_family_id, column_family_name, io_priority,
+                      write_hint, io_tracer, blob_callback, blob_file_paths,
+                      blob_file_additions) {}
 
 BlobFileBuilder::BlobFileBuilder(
-    std::function<uint64_t()> file_number_generator, Env* env, FileSystem* fs,
+    std::function<uint64_t()> file_number_generator, FileSystem* fs,
     const ImmutableCFOptions* immutable_cf_options,
     const MutableCFOptions* mutable_cf_options, const FileOptions* file_options,
     int job_id, uint32_t column_family_id,
     const std::string& column_family_name, Env::IOPriority io_priority,
     Env::WriteLifeTimeHint write_hint,
+    const std::shared_ptr<IOTracer>& io_tracer,
+    BlobFileCompletionCallback* blob_callback,
     std::vector<std::string>* blob_file_paths,
     std::vector<BlobFileAddition>* blob_file_additions)
     : file_number_generator_(std::move(file_number_generator)),
-      env_(env),
       fs_(fs),
       immutable_cf_options_(immutable_cf_options),
       min_blob_size_(mutable_cf_options->min_blob_size),
@@ -61,12 +67,13 @@ BlobFileBuilder::BlobFileBuilder(
       column_family_name_(column_family_name),
       io_priority_(io_priority),
       write_hint_(write_hint),
+      io_tracer_(io_tracer),
+      blob_callback_(blob_callback),
       blob_file_paths_(blob_file_paths),
       blob_file_additions_(blob_file_additions),
       blob_count_(0),
       blob_bytes_(0) {
   assert(file_number_generator_);
-  assert(env_);
   assert(fs_);
   assert(immutable_cf_options_);
   assert(file_options_);
@@ -157,11 +164,12 @@ Status BlobFileBuilder::OpenBlobFileIfNeeded() {
   std::unique_ptr<FSWritableFile> file;
 
   {
-    TEST_SYNC_POINT("BlobFileBuilder::OpenBlobFileIfNeeded:NewWritableFile");
-
     assert(file_options_);
-    const Status s =
-        NewWritableFile(fs_, blob_file_path, &file, *file_options_);
+    Status s = NewWritableFile(fs_, blob_file_path, &file, *file_options_);
+
+    TEST_SYNC_POINT_CALLBACK(
+        "BlobFileBuilder::OpenBlobFileIfNeeded:NewWritableFile", &s);
+
     if (!s.ok()) {
       return s;
     }
@@ -176,17 +184,20 @@ Status BlobFileBuilder::OpenBlobFileIfNeeded() {
   assert(file);
   file->SetIOPriority(io_priority_);
   file->SetWriteLifeTimeHint(write_hint_);
-
+  FileTypeSet tmp_set = immutable_cf_options_->checksum_handoff_file_types;
   Statistics* const statistics = immutable_cf_options_->statistics;
-
   std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-      std::move(file), blob_file_paths_->back(), *file_options_, env_,
-      nullptr /*IOTracer*/, statistics, immutable_cf_options_->listeners,
-      immutable_cf_options_->file_checksum_gen_factory));
+      std::move(file), blob_file_paths_->back(), *file_options_,
+      immutable_cf_options_->clock, io_tracer_, statistics,
+      immutable_cf_options_->listeners,
+      immutable_cf_options_->file_checksum_gen_factory,
+      tmp_set.Contains(FileType::kBlobFile)));
 
-  std::unique_ptr<BlobLogWriter> blob_log_writer(
-      new BlobLogWriter(std::move(file_writer), env_, statistics,
-                        blob_file_number, immutable_cf_options_->use_fsync));
+  constexpr bool do_flush = false;
+
+  std::unique_ptr<BlobLogWriter> blob_log_writer(new BlobLogWriter(
+      std::move(file_writer), immutable_cf_options_->clock, statistics,
+      blob_file_number, immutable_cf_options_->use_fsync, do_flush));
 
   constexpr bool has_ttl = false;
   constexpr ExpirationRange expiration_range;
@@ -195,9 +206,11 @@ Status BlobFileBuilder::OpenBlobFileIfNeeded() {
                        expiration_range);
 
   {
-    TEST_SYNC_POINT("BlobFileBuilder::OpenBlobFileIfNeeded:WriteHeader");
+    Status s = blob_log_writer->WriteHeader(header);
+
+    TEST_SYNC_POINT_CALLBACK(
+        "BlobFileBuilder::OpenBlobFileIfNeeded:WriteHeader", &s);
 
-    const Status s = blob_log_writer->WriteHeader(header);
     if (!s.ok()) {
       return s;
     }
@@ -247,9 +260,10 @@ Status BlobFileBuilder::WriteBlobToFile(const Slice& key, const Slice& blob,
 
   uint64_t key_offset = 0;
 
-  TEST_SYNC_POINT("BlobFileBuilder::WriteBlobToFile:AddRecord");
+  Status s = writer_->AddRecord(key, blob, &key_offset, blob_offset);
+
+  TEST_SYNC_POINT_CALLBACK("BlobFileBuilder::WriteBlobToFile:AddRecord", &s);
 
-  const Status s = writer_->AddRecord(key, blob, &key_offset, blob_offset);
   if (!s.ok()) {
     return s;
   }
@@ -271,10 +285,10 @@ Status BlobFileBuilder::CloseBlobFile() {
   std::string checksum_method;
   std::string checksum_value;
 
-  TEST_SYNC_POINT("BlobFileBuilder::WriteBlobToFile:AppendFooter");
+  Status s = writer_->AppendFooter(footer, &checksum_method, &checksum_value);
+
+  TEST_SYNC_POINT_CALLBACK("BlobFileBuilder::WriteBlobToFile:AppendFooter", &s);
 
-  const Status s =
-      writer_->AppendFooter(footer, &checksum_method, &checksum_value);
   if (!s.ok()) {
     return s;
   }
@@ -293,11 +307,15 @@ Status BlobFileBuilder::CloseBlobFile() {
                  column_family_name_.c_str(), job_id_, blob_file_number,
                  blob_count_, blob_bytes_);
 
+  if (blob_callback_) {
+    s = blob_callback_->OnBlobFileCompleted(blob_file_paths_->back());
+  }
+
   writer_.reset();
   blob_count_ = 0;
   blob_bytes_ = 0;
 
-  return Status::OK();
+  return s;
 }
 
 Status BlobFileBuilder::CloseBlobFileIfNeeded() {
@@ -313,4 +331,20 @@ Status BlobFileBuilder::CloseBlobFileIfNeeded() {
   return CloseBlobFile();
 }
 
+void BlobFileBuilder::Abandon() {
+  if (!IsBlobFileOpen()) {
+    return;
+  }
+
+  if (blob_callback_) {
+    // BlobFileBuilder::Abandon() is called because of error while writing to
+    // Blob files. So we can ignore the below error.
+    blob_callback_->OnBlobFileCompleted(blob_file_paths_->back())
+        .PermitUncheckedError();
+  }
+
+  writer_.reset();
+  blob_count_ = 0;
+  blob_bytes_ = 0;
+}
 }  // namespace ROCKSDB_NAMESPACE

db/blob/blob_file_builder.h

@@ -18,6 +18,7 @@ namespace ROCKSDB_NAMESPACE {
 
 class VersionSet;
 class FileSystem;
+class SystemClock;
 struct ImmutableCFOptions;
 struct MutableCFOptions;
 struct FileOptions;
@@ -25,10 +26,12 @@ class BlobFileAddition;
 class Status;
 class Slice;
 class BlobLogWriter;
+class IOTracer;
+class BlobFileCompletionCallback;
 
 class BlobFileBuilder {
  public:
-  BlobFileBuilder(VersionSet* versions, Env* env, FileSystem* fs,
+  BlobFileBuilder(VersionSet* versions, FileSystem* fs,
                   const ImmutableCFOptions* immutable_cf_options,
                   const MutableCFOptions* mutable_cf_options,
                   const FileOptions* file_options, int job_id,
@@ -36,10 +39,12 @@ class BlobFileBuilder {
                   const std::string& column_family_name,
                   Env::IOPriority io_priority,
                   Env::WriteLifeTimeHint write_hint,
+                  const std::shared_ptr<IOTracer>& io_tracer,
+                  BlobFileCompletionCallback* blob_callback,
                   std::vector<std::string>* blob_file_paths,
                   std::vector<BlobFileAddition>* blob_file_additions);
 
-  BlobFileBuilder(std::function<uint64_t()> file_number_generator, Env* env,
+  BlobFileBuilder(std::function<uint64_t()> file_number_generator,
                   FileSystem* fs,
                   const ImmutableCFOptions* immutable_cf_options,
                   const MutableCFOptions* mutable_cf_options,
@@ -48,6 +53,8 @@ class BlobFileBuilder {
                   const std::string& column_family_name,
                   Env::IOPriority io_priority,
                   Env::WriteLifeTimeHint write_hint,
+                  const std::shared_ptr<IOTracer>& io_tracer,
+                  BlobFileCompletionCallback* blob_callback,
                   std::vector<std::string>* blob_file_paths,
                   std::vector<BlobFileAddition>* blob_file_additions);
 
@@ -58,6 +65,7 @@ class BlobFileBuilder {
 
   Status Add(const Slice& key, const Slice& value, std::string* blob_index);
   Status Finish();
+  void Abandon();
 
  private:
   bool IsBlobFileOpen() const;
@@ -69,7 +77,6 @@ class BlobFileBuilder {
   Status CloseBlobFileIfNeeded();
 
   std::function<uint64_t()> file_number_generator_;
-  Env* env_;
   FileSystem* fs_;
   const ImmutableCFOptions* immutable_cf_options_;
   uint64_t min_blob_size_;
@@ -81,6 +88,8 @@ class BlobFileBuilder {
   std::string column_family_name_;
   Env::IOPriority io_priority_;
   Env::WriteLifeTimeHint write_hint_;
+  std::shared_ptr<IOTracer> io_tracer_;
+  BlobFileCompletionCallback* blob_callback_;
   std::vector<std::string>* blob_file_paths_;
   std::vector<BlobFileAddition>* blob_file_additions_;
   std::unique_ptr<BlobLogWriter> writer_;

db/blob/blob_file_builder_test.cc

@@ -40,7 +40,8 @@ class TestFileNumberGenerator {
 class BlobFileBuilderTest : public testing::Test {
  protected:
   BlobFileBuilderTest() : mock_env_(Env::Default()) {
-    fs_ = mock_env_.GetFileSystem();
+    fs_ = mock_env_.GetFileSystem().get();
+    clock_ = mock_env_.GetSystemClock().get();
   }
 
   void VerifyBlobFile(uint64_t blob_file_number,
@@ -58,11 +59,10 @@ class BlobFileBuilderTest : public testing::Test {
         fs_->NewRandomAccessFile(blob_file_path, file_options_, &file, dbg));
 
     std::unique_ptr<RandomAccessFileReader> file_reader(
-        new RandomAccessFileReader(std::move(file), blob_file_path,
-                                   &mock_env_));
+        new RandomAccessFileReader(std::move(file), blob_file_path, clock_));
 
     constexpr Statistics* statistics = nullptr;
-    BlobLogSequentialReader blob_log_reader(std::move(file_reader), &mock_env_,
+    BlobLogSequentialReader blob_log_reader(std::move(file_reader), clock_,
                                             statistics);
 
     BlobLogHeader header;
@@ -109,7 +109,8 @@ class BlobFileBuilderTest : public testing::Test {
   }
 
   MockEnv mock_env_;
-  std::shared_ptr<FileSystem> fs_;
+  FileSystem* fs_;
+  SystemClock* clock_;
   FileOptions file_options_;
 };
 
@@ -126,6 +127,7 @@ TEST_F(BlobFileBuilderTest, BuildAndCheckOneFile) {
                             "BlobFileBuilderTest_BuildAndCheckOneFile"),
       0);
   options.enable_blob_files = true;
+  options.env = &mock_env_;
 
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
@@ -139,10 +141,11 @@ TEST_F(BlobFileBuilderTest, BuildAndCheckOneFile) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &mock_env_, fs_.get(),
-                          &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
   std::vector<std::pair<std::string, std::string>> expected_key_value_pairs(
@@ -209,6 +212,7 @@ TEST_F(BlobFileBuilderTest, BuildAndCheckMultipleFiles) {
       0);
   options.enable_blob_files = true;
   options.blob_file_size = value_size;
+  options.env = &mock_env_;
 
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
@@ -222,10 +226,11 @@ TEST_F(BlobFileBuilderTest, BuildAndCheckMultipleFiles) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &mock_env_, fs_.get(),
-                          &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
   std::vector<std::pair<std::string, std::string>> expected_key_value_pairs(
@@ -294,6 +299,7 @@ TEST_F(BlobFileBuilderTest, InlinedValues) {
       0);
   options.enable_blob_files = true;
   options.min_blob_size = 1024;
+  options.env = &mock_env_;
 
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
@@ -307,10 +313,11 @@ TEST_F(BlobFileBuilderTest, InlinedValues) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &mock_env_, fs_.get(),
-                          &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
   for (size_t i = 0; i < number_of_blobs; ++i) {
@@ -346,6 +353,7 @@ TEST_F(BlobFileBuilderTest, Compression) {
       test::PerThreadDBPath(&mock_env_, "BlobFileBuilderTest_Compression"), 0);
   options.enable_blob_files = true;
   options.blob_compression_type = kSnappyCompression;
+  options.env = &mock_env_;
 
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
@@ -359,10 +367,11 @@ TEST_F(BlobFileBuilderTest, Compression) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &mock_env_, fs_.get(),
-                          &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
   const std::string key("1");
@@ -428,7 +437,7 @@ TEST_F(BlobFileBuilderTest, CompressionError) {
       0);
   options.enable_blob_files = true;
   options.blob_compression_type = kSnappyCompression;
-
+  options.env = &mock_env_;
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
 
@@ -441,10 +450,11 @@ TEST_F(BlobFileBuilderTest, CompressionError) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &mock_env_, fs_.get(),
-                          &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
   SyncPoint::GetInstance()->SetCallBack("CompressData:TamperWithReturnValue",
@@ -505,6 +515,7 @@ TEST_F(BlobFileBuilderTest, Checksum) {
   options.enable_blob_files = true;
   options.file_checksum_gen_factory =
       std::make_shared<DummyFileChecksumGenFactory>();
+  options.env = &mock_env_;
 
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
@@ -518,10 +529,11 @@ TEST_F(BlobFileBuilderTest, Checksum) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &mock_env_, fs_.get(),
-                          &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
   const std::string key("1");
@@ -571,13 +583,11 @@ class BlobFileBuilderIOErrorTest
  protected:
   BlobFileBuilderIOErrorTest()
       : mock_env_(Env::Default()),
-        fault_injection_env_(&mock_env_),
-        fs_(fault_injection_env_.GetFileSystem()),
+        fs_(mock_env_.GetFileSystem().get()),
         sync_point_(GetParam()) {}
 
   MockEnv mock_env_;
-  FaultInjectionTestEnv fault_injection_env_;
-  std::shared_ptr<FileSystem> fs_;
+  FileSystem* fs_;
   FileOptions file_options_;
   std::string sync_point_;
 };
@@ -598,11 +608,11 @@ TEST_P(BlobFileBuilderIOErrorTest, IOError) {
 
   Options options;
   options.cf_paths.emplace_back(
-      test::PerThreadDBPath(&fault_injection_env_,
-                            "BlobFileBuilderIOErrorTest_IOError"),
+      test::PerThreadDBPath(&mock_env_, "BlobFileBuilderIOErrorTest_IOError"),
       0);
   options.enable_blob_files = true;
   options.blob_file_size = value_size;
+  options.env = &mock_env_;
 
   ImmutableCFOptions immutable_cf_options(options);
   MutableCFOptions mutable_cf_options(options);
@@ -616,15 +626,18 @@ TEST_P(BlobFileBuilderIOErrorTest, IOError) {
   std::vector<std::string> blob_file_paths;
   std::vector<BlobFileAddition> blob_file_additions;
 
-  BlobFileBuilder builder(TestFileNumberGenerator(), &fault_injection_env_,
-                          fs_.get(), &immutable_cf_options, &mutable_cf_options,
-                          &file_options_, job_id, column_family_id,
-                          column_family_name, io_priority, write_hint,
+  BlobFileBuilder builder(TestFileNumberGenerator(), fs_, &immutable_cf_options,
+                          &mutable_cf_options, &file_options_, job_id,
+                          column_family_id, column_family_name, io_priority,
+                          write_hint, nullptr /*IOTracer*/,
+                          nullptr /*BlobFileCompletionCallback*/,
                           &blob_file_paths, &blob_file_additions);
 
-  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
-    fault_injection_env_.SetFilesystemActive(false,
-                                             Status::IOError(sync_point_));
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* arg) {
+    Status* const s = static_cast<Status*>(arg);
+    assert(s);
+
+    (*s) = Status::IOError(sync_point_);
   });
   SyncPoint::GetInstance()->EnableProcessing();
 

db/blob/blob_file_cache.cc

@@ -13,6 +13,7 @@
 #include "rocksdb/cache.h"
 #include "rocksdb/slice.h"
 #include "test_util/sync_point.h"
+#include "trace_replay/io_tracer.h"
 #include "util/hash.h"
 
 namespace ROCKSDB_NAMESPACE {
@@ -21,13 +22,15 @@ BlobFileCache::BlobFileCache(Cache* cache,
                              const ImmutableCFOptions* immutable_cf_options,
                              const FileOptions* file_options,
                              uint32_t column_family_id,
-                             HistogramImpl* blob_file_read_hist)
+                             HistogramImpl* blob_file_read_hist,
+                             const std::shared_ptr<IOTracer>& io_tracer)
     : cache_(cache),
-      mutex_(kNumberOfMutexStripes, GetSliceNPHash64),
+      mutex_(kNumberOfMutexStripes, kGetSliceNPHash64UnseededFnPtr),
       immutable_cf_options_(immutable_cf_options),
       file_options_(file_options),
       column_family_id_(column_family_id),
-      blob_file_read_hist_(blob_file_read_hist) {
+      blob_file_read_hist_(blob_file_read_hist),
+      io_tracer_(io_tracer) {
   assert(cache_);
   assert(immutable_cf_options_);
   assert(file_options_);
@@ -71,7 +74,7 @@ Status BlobFileCache::GetBlobFileReader(
     assert(file_options_);
     const Status s = BlobFileReader::Create(
         *immutable_cf_options_, *file_options_, column_family_id_,
-        blob_file_read_hist_, blob_file_number, &reader);
+        blob_file_read_hist_, blob_file_number, io_tracer_, &reader);
     if (!s.ok()) {
       RecordTick(statistics, NO_FILE_ERRORS);
       return s;

db/blob/blob_file_cache.h

@@ -20,12 +20,14 @@ class HistogramImpl;
 class Status;
 class BlobFileReader;
 class Slice;
+class IOTracer;
 
 class BlobFileCache {
  public:
   BlobFileCache(Cache* cache, const ImmutableCFOptions* immutable_cf_options,
                 const FileOptions* file_options, uint32_t column_family_id,
-                HistogramImpl* blob_file_read_hist);
+                HistogramImpl* blob_file_read_hist,
+                const std::shared_ptr<IOTracer>& io_tracer);
 
   BlobFileCache(const BlobFileCache&) = delete;
   BlobFileCache& operator=(const BlobFileCache&) = delete;
@@ -42,6 +44,7 @@ class BlobFileCache {
   const FileOptions* file_options_;
   uint32_t column_family_id_;
   HistogramImpl* blob_file_read_hist_;
+  std::shared_ptr<IOTracer> io_tracer_;
 
   static constexpr size_t kNumberOfMutexStripes = 1 << 7;
 };

db/blob/blob_file_cache_test.cc

@@ -42,14 +42,15 @@ void WriteBlobFile(uint32_t column_family_id,
 
   std::unique_ptr<WritableFileWriter> file_writer(
       new WritableFileWriter(std::move(file), blob_file_path, FileOptions(),
-                             immutable_cf_options.env));
+                             immutable_cf_options.clock));
 
   constexpr Statistics* statistics = nullptr;
   constexpr bool use_fsync = false;
+  constexpr bool do_flush = false;
 
   BlobLogWriter blob_log_writer(std::move(file_writer),
-                                immutable_cf_options.env, statistics,
-                                blob_file_number, use_fsync);
+                                immutable_cf_options.clock, statistics,
+                                blob_file_number, use_fsync, do_flush);
 
   constexpr bool has_ttl = false;
   constexpr ExpirationRange expiration_range;
@@ -63,7 +64,6 @@ void WriteBlobFile(uint32_t column_family_id,
   constexpr char blob[] = "blob";
 
   std::string compressed_blob;
-  Slice blob_to_write;
 
   uint64_t key_offset = 0;
   uint64_t blob_offset = 0;
@@ -113,7 +113,7 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader) {
 
   BlobFileCache blob_file_cache(backing_cache.get(), &immutable_cf_options,
                                 &file_options, column_family_id,
-                                blob_file_read_hist);
+                                blob_file_read_hist, nullptr /*IOTracer*/);
 
   // First try: reader should be opened and put in cache
   CacheHandleGuard<BlobFileReader> first;
@@ -158,7 +158,7 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_Race) {
 
   BlobFileCache blob_file_cache(backing_cache.get(), &immutable_cf_options,
                                 &file_options, column_family_id,
-                                blob_file_read_hist);
+                                blob_file_read_hist, nullptr /*IOTracer*/);
 
   CacheHandleGuard<BlobFileReader> first;
   CacheHandleGuard<BlobFileReader> second;
@@ -206,7 +206,7 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_IOError) {
 
   BlobFileCache blob_file_cache(backing_cache.get(), &immutable_cf_options,
                                 &file_options, column_family_id,
-                                blob_file_read_hist);
+                                blob_file_read_hist, nullptr /*IOTracer*/);
 
   // Note: there is no blob file with the below number
   constexpr uint64_t blob_file_number = 123;
@@ -247,7 +247,7 @@ TEST_F(BlobFileCacheTest, GetBlobFileReader_CacheFull) {
 
   BlobFileCache blob_file_cache(backing_cache.get(), &immutable_cf_options,
                                 &file_options, column_family_id,
-                                blob_file_read_hist);
+                                blob_file_read_hist, nullptr /*IOTracer*/);
 
   // Insert into cache should fail since it has zero capacity and
   // strict_capacity_limit is set

db/blob/blob_file_completion_callback.h

@@ -0,0 +1,53 @@
+//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under both the GPLv2 (found in the
+//  COPYING file in the root directory) and Apache 2.0 License
+//  (found in the LICENSE.Apache file in the root directory).
+//
+// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+#pragma once
+
+#include "db/error_handler.h"
+#include "file/sst_file_manager_impl.h"
+#include "rocksdb/status.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class BlobFileCompletionCallback {
+ public:
+  BlobFileCompletionCallback(SstFileManager* sst_file_manager,
+                             InstrumentedMutex* mutex,
+                             ErrorHandler* error_handler)
+      : sst_file_manager_(sst_file_manager),
+        mutex_(mutex),
+        error_handler_(error_handler) {}
+
+  Status OnBlobFileCompleted(const std::string& file_name) {
+    Status s;
+
+#ifndef ROCKSDB_LITE
+    auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager_);
+    if (sfm) {
+      // Report new blob files to SstFileManagerImpl
+      s = sfm->OnAddFile(file_name);
+      if (sfm->IsMaxAllowedSpaceReached()) {
+        s = Status::SpaceLimit("Max allowed space was reached");
+        TEST_SYNC_POINT(
+            "BlobFileCompletionCallback::CallBack::MaxAllowedSpaceReached");
+        InstrumentedMutexLock l(mutex_);
+        error_handler_->SetBGError(s, BackgroundErrorReason::kFlush);
+      }
+    }
+#else
+    (void)file_name;
+#endif  // ROCKSDB_LITE
+    return s;
+  }
+
+ private:
+  SstFileManager* sst_file_manager_;
+  InstrumentedMutex* mutex_;
+  ErrorHandler* error_handler_;
+};
+}  // namespace ROCKSDB_NAMESPACE

db/blob/blob_file_reader.cc

@@ -24,6 +24,7 @@ Status BlobFileReader::Create(
     const ImmutableCFOptions& immutable_cf_options,
     const FileOptions& file_options, uint32_t column_family_id,
     HistogramImpl* blob_file_read_hist, uint64_t blob_file_number,
+    const std::shared_ptr<IOTracer>& io_tracer,
     std::unique_ptr<BlobFileReader>* blob_file_reader) {
   assert(blob_file_reader);
   assert(!*blob_file_reader);
@@ -34,7 +35,7 @@ Status BlobFileReader::Create(
   {
     const Status s =
         OpenFile(immutable_cf_options, file_options, blob_file_read_hist,
-                 blob_file_number, &file_size, &file_reader);
+                 blob_file_number, io_tracer, &file_size, &file_reader);
     if (!s.ok()) {
       return s;
     }
@@ -68,8 +69,8 @@ Status BlobFileReader::Create(
 Status BlobFileReader::OpenFile(
     const ImmutableCFOptions& immutable_cf_options,
     const FileOptions& file_opts, HistogramImpl* blob_file_read_hist,
-    uint64_t blob_file_number, uint64_t* file_size,
-    std::unique_ptr<RandomAccessFileReader>* file_reader) {
+    uint64_t blob_file_number, const std::shared_ptr<IOTracer>& io_tracer,
+    uint64_t* file_size, std::unique_ptr<RandomAccessFileReader>* file_reader) {
   assert(file_size);
   assert(file_reader);
 
@@ -117,10 +118,10 @@ Status BlobFileReader::OpenFile(
   }
 
   file_reader->reset(new RandomAccessFileReader(
-      std::move(file), blob_file_path, immutable_cf_options.env,
-      std::shared_ptr<IOTracer>(), immutable_cf_options.statistics,
-      BLOB_DB_BLOB_FILE_READ_MICROS, blob_file_read_hist,
-      immutable_cf_options.rate_limiter, immutable_cf_options.listeners));
+      std::move(file), blob_file_path, immutable_cf_options.clock, io_tracer,
+      immutable_cf_options.statistics, BLOB_DB_BLOB_FILE_READ_MICROS,
+      blob_file_read_hist, immutable_cf_options.rate_limiter,
+      immutable_cf_options.listeners));
 
   return Status::OK();
 }
@@ -269,7 +270,8 @@ Status BlobFileReader::GetBlob(const ReadOptions& read_options,
                                const Slice& user_key, uint64_t offset,
                                uint64_t value_size,
                                CompressionType compression_type,
-                               PinnableSlice* value) const {
+                               PinnableSlice* value,
+                               uint64_t* bytes_read) const {
   assert(value);
 
   const uint64_t key_size = user_key.size();
@@ -292,6 +294,9 @@ Status BlobFileReader::GetBlob(const ReadOptions& read_options,
           : 0;
   assert(offset >= adjustment);
 
+  const uint64_t record_offset = offset - adjustment;
+  const uint64_t record_size = value_size + adjustment;
+
   Slice record_slice;
   Buffer buf;
   AlignedBuf aligned_buf;
@@ -299,9 +304,6 @@ Status BlobFileReader::GetBlob(const ReadOptions& read_options,
   {
     TEST_SYNC_POINT("BlobFileReader::GetBlob:ReadFromFile");
 
-    const uint64_t record_offset = offset - adjustment;
-    const uint64_t record_size = value_size + adjustment;
-
     const Status s = ReadFromFile(file_reader_.get(), record_offset,
                                   static_cast<size_t>(record_size),
                                   &record_slice, &buf, &aligned_buf);
@@ -330,6 +332,10 @@ Status BlobFileReader::GetBlob(const ReadOptions& read_options,
     }
   }
 
+  if (bytes_read) {
+    *bytes_read = record_size;
+  }
+
   return Status::OK();
 }
 

db/blob/blob_file_reader.h

@@ -29,6 +29,7 @@ class BlobFileReader {
                        uint32_t column_family_id,
                        HistogramImpl* blob_file_read_hist,
                        uint64_t blob_file_number,
+                       const std::shared_ptr<IOTracer>& io_tracer,
                        std::unique_ptr<BlobFileReader>* reader);
 
   BlobFileReader(const BlobFileReader&) = delete;
@@ -38,7 +39,8 @@ class BlobFileReader {
 
   Status GetBlob(const ReadOptions& read_options, const Slice& user_key,
                  uint64_t offset, uint64_t value_size,
-                 CompressionType compression_type, PinnableSlice* value) const;
+                 CompressionType compression_type, PinnableSlice* value,
+                 uint64_t* bytes_read) const;
 
  private:
   BlobFileReader(std::unique_ptr<RandomAccessFileReader>&& file_reader,
@@ -47,7 +49,9 @@ class BlobFileReader {
   static Status OpenFile(const ImmutableCFOptions& immutable_cf_options,
                          const FileOptions& file_opts,
                          HistogramImpl* blob_file_read_hist,
-                         uint64_t blob_file_number, uint64_t* file_size,
+                         uint64_t blob_file_number,
+                         const std::shared_ptr<IOTracer>& io_tracer,
+                         uint64_t* file_size,
                          std::unique_ptr<RandomAccessFileReader>* file_reader);
 
   static Status ReadHeader(const RandomAccessFileReader* file_reader,

db/blob/blob_file_reader_test.cc

@@ -50,14 +50,15 @@ void WriteBlobFile(const ImmutableCFOptions& immutable_cf_options,
 
   std::unique_ptr<WritableFileWriter> file_writer(
       new WritableFileWriter(std::move(file), blob_file_path, FileOptions(),
-                             immutable_cf_options.env));
+                             immutable_cf_options.clock));
 
   constexpr Statistics* statistics = nullptr;
   constexpr bool use_fsync = false;
+  constexpr bool do_flush = false;
 
   BlobLogWriter blob_log_writer(std::move(file_writer),
-                                immutable_cf_options.env, statistics,
-                                blob_file_number, use_fsync);
+                                immutable_cf_options.clock, statistics,
+                                blob_file_number, use_fsync, do_flush);
 
   BlobLogHeader header(column_family_id, compression_type, has_ttl,
                        expiration_range_header);
@@ -141,9 +142,9 @@ TEST_F(BlobFileReaderTest, CreateReaderAndGetBlob) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  ASSERT_OK(BlobFileReader::Create(immutable_cf_options, FileOptions(),
-                                   column_family_id, blob_file_read_hist,
-                                   blob_file_number, &reader));
+  ASSERT_OK(BlobFileReader::Create(
+      immutable_cf_options, FileOptions(), column_family_id,
+      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
 
   // Make sure the blob can be retrieved with and without checksum verification
   ReadOptions read_options;
@@ -151,83 +152,103 @@ TEST_F(BlobFileReaderTest, CreateReaderAndGetBlob) {
 
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_OK(reader->GetBlob(read_options, key, blob_offset, blob_size,
-                              kNoCompression, &value));
+                              kNoCompression, &value, &bytes_read));
     ASSERT_EQ(value, blob);
+    ASSERT_EQ(bytes_read, blob_size);
   }
 
   read_options.verify_checksums = true;
 
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_OK(reader->GetBlob(read_options, key, blob_offset, blob_size,
-                              kNoCompression, &value));
+                              kNoCompression, &value, &bytes_read));
     ASSERT_EQ(value, blob);
+
+    constexpr uint64_t key_size = sizeof(key) - 1;
+    ASSERT_EQ(bytes_read,
+              BlobLogRecord::CalculateAdjustmentForRecordHeader(key_size) +
+                  blob_size);
   }
 
   // Invalid offset (too close to start of file)
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(read_options, key, blob_offset - 1, blob_size,
-                              kNoCompression, &value)
+                              kNoCompression, &value, &bytes_read)
                     .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   // Invalid offset (too close to end of file)
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(read_options, key, blob_offset + 1, blob_size,
-                              kNoCompression, &value)
+                              kNoCompression, &value, &bytes_read)
                     .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   // Incorrect compression type
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
-    ASSERT_TRUE(
-        reader
-            ->GetBlob(read_options, key, blob_offset, blob_size, kZSTD, &value)
-            .IsCorruption());
+    ASSERT_TRUE(reader
+                    ->GetBlob(read_options, key, blob_offset, blob_size, kZSTD,
+                              &value, &bytes_read)
+                    .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   // Incorrect key size
   {
     constexpr char shorter_key[] = "k";
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(read_options, shorter_key,
                               blob_offset - (sizeof(key) - sizeof(shorter_key)),
-                              blob_size, kNoCompression, &value)
+                              blob_size, kNoCompression, &value, &bytes_read)
                     .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   // Incorrect key
   {
     constexpr char incorrect_key[] = "foo";
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(read_options, incorrect_key, blob_offset,
-                              blob_size, kNoCompression, &value)
+                              blob_size, kNoCompression, &value, &bytes_read)
                     .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   // Incorrect value size
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(read_options, key, blob_offset, blob_size + 1,
-                              kNoCompression, &value)
+                              kNoCompression, &value, &bytes_read)
                     .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 }
 
@@ -259,14 +280,15 @@ TEST_F(BlobFileReaderTest, Malformed) {
 
     std::unique_ptr<WritableFileWriter> file_writer(
         new WritableFileWriter(std::move(file), blob_file_path, FileOptions(),
-                               immutable_cf_options.env));
+                               immutable_cf_options.clock));
 
     constexpr Statistics* statistics = nullptr;
     constexpr bool use_fsync = false;
+    constexpr bool do_flush = false;
 
     BlobLogWriter blob_log_writer(std::move(file_writer),
-                                  immutable_cf_options.env, statistics,
-                                  blob_file_number, use_fsync);
+                                  immutable_cf_options.clock, statistics,
+                                  blob_file_number, use_fsync, do_flush);
 
     BlobLogHeader header(column_family_id, kNoCompression, has_ttl,
                          expiration_range);
@@ -280,7 +302,8 @@ TEST_F(BlobFileReaderTest, Malformed) {
 
   ASSERT_TRUE(BlobFileReader::Create(immutable_cf_options, FileOptions(),
                                      column_family_id, blob_file_read_hist,
-                                     blob_file_number, &reader)
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -313,7 +336,8 @@ TEST_F(BlobFileReaderTest, TTL) {
 
   ASSERT_TRUE(BlobFileReader::Create(immutable_cf_options, FileOptions(),
                                      column_family_id, blob_file_read_hist,
-                                     blob_file_number, &reader)
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -351,7 +375,8 @@ TEST_F(BlobFileReaderTest, ExpirationRangeInHeader) {
 
   ASSERT_TRUE(BlobFileReader::Create(immutable_cf_options, FileOptions(),
                                      column_family_id, blob_file_read_hist,
-                                     blob_file_number, &reader)
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -389,7 +414,8 @@ TEST_F(BlobFileReaderTest, ExpirationRangeInFooter) {
 
   ASSERT_TRUE(BlobFileReader::Create(immutable_cf_options, FileOptions(),
                                      column_family_id, blob_file_read_hist,
-                                     blob_file_number, &reader)
+                                     blob_file_number, nullptr /*IOTracer*/,
+                                     &reader)
                   .IsCorruption());
 }
 
@@ -427,7 +453,7 @@ TEST_F(BlobFileReaderTest, IncorrectColumnFamily) {
   ASSERT_TRUE(BlobFileReader::Create(immutable_cf_options, FileOptions(),
                                      incorrect_column_family_id,
                                      blob_file_read_hist, blob_file_number,
-                                     &reader)
+                                     nullptr /*IOTracer*/, &reader)
                   .IsCorruption());
 }
 
@@ -458,9 +484,9 @@ TEST_F(BlobFileReaderTest, BlobCRCError) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  ASSERT_OK(BlobFileReader::Create(immutable_cf_options, FileOptions(),
-                                   column_family_id, blob_file_read_hist,
-                                   blob_file_number, &reader));
+  ASSERT_OK(BlobFileReader::Create(
+      immutable_cf_options, FileOptions(), column_family_id,
+      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
 
   SyncPoint::GetInstance()->SetCallBack(
       "BlobFileReader::VerifyBlob:CheckBlobCRC", [](void* arg) {
@@ -473,11 +499,13 @@ TEST_F(BlobFileReaderTest, BlobCRCError) {
   SyncPoint::GetInstance()->EnableProcessing();
 
   PinnableSlice value;
+  uint64_t bytes_read = 0;
 
   ASSERT_TRUE(reader
                   ->GetBlob(ReadOptions(), key, blob_offset, blob_size,
-                            kNoCompression, &value)
+                            kNoCompression, &value, &bytes_read)
                   .IsCorruption());
+  ASSERT_EQ(bytes_read, 0);
 
   SyncPoint::GetInstance()->DisableProcessing();
   SyncPoint::GetInstance()->ClearAllCallBacks();
@@ -514,9 +542,9 @@ TEST_F(BlobFileReaderTest, Compression) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  ASSERT_OK(BlobFileReader::Create(immutable_cf_options, FileOptions(),
-                                   column_family_id, blob_file_read_hist,
-                                   blob_file_number, &reader));
+  ASSERT_OK(BlobFileReader::Create(
+      immutable_cf_options, FileOptions(), column_family_id,
+      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
 
   // Make sure the blob can be retrieved with and without checksum verification
   ReadOptions read_options;
@@ -524,20 +552,28 @@ TEST_F(BlobFileReaderTest, Compression) {
 
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_OK(reader->GetBlob(read_options, key, blob_offset, blob_size,
-                              kSnappyCompression, &value));
+                              kSnappyCompression, &value, &bytes_read));
     ASSERT_EQ(value, blob);
+    ASSERT_EQ(bytes_read, blob_size);
   }
 
   read_options.verify_checksums = true;
 
   {
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_OK(reader->GetBlob(read_options, key, blob_offset, blob_size,
-                              kSnappyCompression, &value));
+                              kSnappyCompression, &value, &bytes_read));
     ASSERT_EQ(value, blob);
+
+    constexpr uint64_t key_size = sizeof(key) - 1;
+    ASSERT_EQ(bytes_read,
+              BlobLogRecord::CalculateAdjustmentForRecordHeader(key_size) +
+                  blob_size);
   }
 }
 
@@ -574,9 +610,9 @@ TEST_F(BlobFileReaderTest, UncompressionError) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  ASSERT_OK(BlobFileReader::Create(immutable_cf_options, FileOptions(),
-                                   column_family_id, blob_file_read_hist,
-                                   blob_file_number, &reader));
+  ASSERT_OK(BlobFileReader::Create(
+      immutable_cf_options, FileOptions(), column_family_id,
+      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader));
 
   SyncPoint::GetInstance()->SetCallBack(
       "BlobFileReader::UncompressBlobIfNeeded:TamperWithResult", [](void* arg) {
@@ -590,11 +626,13 @@ TEST_F(BlobFileReaderTest, UncompressionError) {
   SyncPoint::GetInstance()->EnableProcessing();
 
   PinnableSlice value;
+  uint64_t bytes_read = 0;
 
   ASSERT_TRUE(reader
                   ->GetBlob(ReadOptions(), key, blob_offset, blob_size,
-                            kSnappyCompression, &value)
+                            kSnappyCompression, &value, &bytes_read)
                   .IsCorruption());
+  ASSERT_EQ(bytes_read, 0);
 
   SyncPoint::GetInstance()->DisableProcessing();
   SyncPoint::GetInstance()->ClearAllCallBacks();
@@ -659,9 +697,9 @@ TEST_P(BlobFileReaderIOErrorTest, IOError) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  const Status s = BlobFileReader::Create(immutable_cf_options, FileOptions(),
-                                          column_family_id, blob_file_read_hist,
-                                          blob_file_number, &reader);
+  const Status s = BlobFileReader::Create(
+      immutable_cf_options, FileOptions(), column_family_id,
+      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader);
 
   const bool fail_during_create =
       (sync_point_ != "BlobFileReader::GetBlob:ReadFromFile");
@@ -672,11 +710,13 @@ TEST_P(BlobFileReaderIOErrorTest, IOError) {
     ASSERT_OK(s);
 
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(ReadOptions(), key, blob_offset, blob_size,
-                              kNoCompression, &value)
+                              kNoCompression, &value, &bytes_read)
                     .IsIOError());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   SyncPoint::GetInstance()->DisableProcessing();
@@ -739,9 +779,9 @@ TEST_P(BlobFileReaderDecodingErrorTest, DecodingError) {
 
   std::unique_ptr<BlobFileReader> reader;
 
-  const Status s = BlobFileReader::Create(immutable_cf_options, FileOptions(),
-                                          column_family_id, blob_file_read_hist,
-                                          blob_file_number, &reader);
+  const Status s = BlobFileReader::Create(
+      immutable_cf_options, FileOptions(), column_family_id,
+      blob_file_read_hist, blob_file_number, nullptr /*IOTracer*/, &reader);
 
   const bool fail_during_create =
       sync_point_ != "BlobFileReader::GetBlob:TamperWithResult";
@@ -752,11 +792,13 @@ TEST_P(BlobFileReaderDecodingErrorTest, DecodingError) {
     ASSERT_OK(s);
 
     PinnableSlice value;
+    uint64_t bytes_read = 0;
 
     ASSERT_TRUE(reader
                     ->GetBlob(ReadOptions(), key, blob_offset, blob_size,
-                              kNoCompression, &value)
+                              kNoCompression, &value, &bytes_read)
                     .IsCorruption());
+    ASSERT_EQ(bytes_read, 0);
   }
 
   SyncPoint::GetInstance()->DisableProcessing();

db/blob/blob_log_sequential_reader.cc

@@ -13,10 +13,10 @@
 namespace ROCKSDB_NAMESPACE {
 
 BlobLogSequentialReader::BlobLogSequentialReader(
-    std::unique_ptr<RandomAccessFileReader>&& file_reader, Env* env,
+    std::unique_ptr<RandomAccessFileReader>&& file_reader, SystemClock* clock,
     Statistics* statistics)
     : file_(std::move(file_reader)),
-      env_(env),
+      clock_(clock),
       statistics_(statistics),
       next_byte_(0) {}
 
@@ -27,7 +27,7 @@ Status BlobLogSequentialReader::ReadSlice(uint64_t size, Slice* slice,
   assert(slice);
   assert(file_);
 
-  StopWatch read_sw(env_, statistics_, BLOB_DB_BLOB_FILE_READ_MICROS);
+  StopWatch read_sw(clock_, statistics_, BLOB_DB_BLOB_FILE_READ_MICROS);
   Status s = file_->Read(IOOptions(), next_byte_, static_cast<size_t>(size),
                          slice, buf, nullptr);
   next_byte_ += size;

db/blob/blob_log_sequential_reader.h

@@ -16,6 +16,7 @@ class RandomAccessFileReader;
 class Env;
 class Statistics;
 class Status;
+class SystemClock;
 
 /**
  * BlobLogSequentialReader is a general purpose log stream reader
@@ -35,7 +36,7 @@ class BlobLogSequentialReader {
 
   // Create a reader that will return log records from "*file_reader".
   BlobLogSequentialReader(std::unique_ptr<RandomAccessFileReader>&& file_reader,
-                          Env* env, Statistics* statistics);
+                          SystemClock* clock, Statistics* statistics);
 
   // No copying allowed
   BlobLogSequentialReader(const BlobLogSequentialReader&) = delete;
@@ -63,7 +64,8 @@ class BlobLogSequentialReader {
   Status ReadSlice(uint64_t size, Slice* slice, char* buf);
 
   const std::unique_ptr<RandomAccessFileReader> file_;
-  Env* env_;
+  SystemClock* clock_;
+
   Statistics* statistics_;
 
   Slice buffer_;

db/blob/blob_log_writer.cc

@@ -11,7 +11,7 @@
 #include "db/blob/blob_log_format.h"
 #include "file/writable_file_writer.h"
 #include "monitoring/statistics.h"
-#include "rocksdb/env.h"
+#include "rocksdb/system_clock.h"
 #include "test_util/sync_point.h"
 #include "util/coding.h"
 #include "util/stop_watch.h"
@@ -19,14 +19,16 @@
 namespace ROCKSDB_NAMESPACE {
 
 BlobLogWriter::BlobLogWriter(std::unique_ptr<WritableFileWriter>&& dest,
-                             Env* env, Statistics* statistics,
-                             uint64_t log_number, bool use_fs, uint64_t boffset)
+                             SystemClock* clock, Statistics* statistics,
+                             uint64_t log_number, bool use_fs, bool do_flush,
+                             uint64_t boffset)
     : dest_(std::move(dest)),
-      env_(env),
+      clock_(clock),
       statistics_(statistics),
       log_number_(log_number),
       block_offset_(boffset),
       use_fsync_(use_fs),
+      do_flush_(do_flush),
       last_elem_type_(kEtNone) {}
 
 BlobLogWriter::~BlobLogWriter() = default;
@@ -34,7 +36,7 @@ BlobLogWriter::~BlobLogWriter() = default;
 Status BlobLogWriter::Sync() {
   TEST_SYNC_POINT("BlobLogWriter::Sync");
 
-  StopWatch sync_sw(env_, statistics_, BLOB_DB_BLOB_FILE_SYNC_MICROS);
+  StopWatch sync_sw(clock_, statistics_, BLOB_DB_BLOB_FILE_SYNC_MICROS);
   Status s = dest_->Sync(use_fsync_);
   RecordTick(statistics_, BLOB_DB_BLOB_FILE_SYNCED);
   return s;
@@ -49,7 +51,9 @@ Status BlobLogWriter::WriteHeader(BlobLogHeader& header) {
   Status s = dest_->Append(Slice(str));
   if (s.ok()) {
     block_offset_ += str.size();
-    s = dest_->Flush();
+    if (do_flush_) {
+      s = dest_->Flush();
+    }
   }
   last_elem_type_ = kEtFileHdr;
   RecordTick(statistics_, BLOB_DB_BLOB_FILE_BYTES_WRITTEN,
@@ -144,7 +148,7 @@ Status BlobLogWriter::EmitPhysicalRecord(const std::string& headerbuf,
                                          const Slice& key, const Slice& val,
                                          uint64_t* key_offset,
                                          uint64_t* blob_offset) {
-  StopWatch write_sw(env_, statistics_, BLOB_DB_BLOB_FILE_WRITE_MICROS);
+  StopWatch write_sw(clock_, statistics_, BLOB_DB_BLOB_FILE_WRITE_MICROS);
   Status s = dest_->Append(Slice(headerbuf));
   if (s.ok()) {
     s = dest_->Append(key);
@@ -152,7 +156,7 @@ Status BlobLogWriter::EmitPhysicalRecord(const std::string& headerbuf,
   if (s.ok()) {
     s = dest_->Append(val);
   }
-  if (s.ok()) {
+  if (do_flush_ && s.ok()) {
     s = dest_->Flush();
   }
 

db/blob/blob_log_writer.h

@@ -9,7 +9,6 @@
 #include <string>
 
 #include "db/blob/blob_log_format.h"
-#include "rocksdb/env.h"
 #include "rocksdb/slice.h"
 #include "rocksdb/statistics.h"
 #include "rocksdb/status.h"
@@ -18,7 +17,7 @@
 namespace ROCKSDB_NAMESPACE {
 
 class WritableFileWriter;
-
+class SystemClock;
 /**
  * BlobLogWriter is the blob log stream writer. It provides an append-only
  * abstraction for writing blob data.
@@ -32,9 +31,9 @@ class BlobLogWriter {
   // Create a writer that will append data to "*dest".
   // "*dest" must be initially empty.
   // "*dest" must remain live while this BlobLogWriter is in use.
-  BlobLogWriter(std::unique_ptr<WritableFileWriter>&& dest, Env* env,
+  BlobLogWriter(std::unique_ptr<WritableFileWriter>&& dest, SystemClock* clock,
                 Statistics* statistics, uint64_t log_number, bool use_fsync,
-                uint64_t boffset = 0);
+                bool do_flush, uint64_t boffset = 0);
   // No copying allowed
   BlobLogWriter(const BlobLogWriter&) = delete;
   BlobLogWriter& operator=(const BlobLogWriter&) = delete;
@@ -69,11 +68,12 @@ class BlobLogWriter {
 
  private:
   std::unique_ptr<WritableFileWriter> dest_;
-  Env* env_;
+  SystemClock* clock_;
   Statistics* statistics_;
   uint64_t log_number_;
   uint64_t block_offset_;  // Current offset in block
   bool use_fsync_;
+  bool do_flush_;
 
  public:
   enum ElemType { kEtNone, kEtFileHdr, kEtRecord, kEtFileFooter };

db/blob/db_blob_basic_test.cc

@@ -210,6 +210,63 @@ TEST_F(DBBlobBasicTest, GetBlob_IndexWithInvalidFileNumber) {
                   .IsCorruption());
 }
 
+#ifndef ROCKSDB_LITE
+TEST_F(DBBlobBasicTest, GenerateIOTracing) {
+  Options options = GetDefaultOptions();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  std::string trace_file = dbname_ + "/io_trace_file";
+
+  Reopen(options);
+  {
+    // Create IO trace file
+    std::unique_ptr<TraceWriter> trace_writer;
+    ASSERT_OK(
+        NewFileTraceWriter(env_, EnvOptions(), trace_file, &trace_writer));
+    ASSERT_OK(db_->StartIOTrace(env_, TraceOptions(), std::move(trace_writer)));
+
+    constexpr char key[] = "key";
+    constexpr char blob_value[] = "blob_value";
+
+    ASSERT_OK(Put(key, blob_value));
+    ASSERT_OK(Flush());
+    ASSERT_EQ(Get(key), blob_value);
+
+    ASSERT_OK(db_->EndIOTrace());
+    ASSERT_OK(env_->FileExists(trace_file));
+  }
+  {
+    // Parse trace file to check file opertions related to blob files are
+    // recorded.
+    std::unique_ptr<TraceReader> trace_reader;
+    ASSERT_OK(
+        NewFileTraceReader(env_, EnvOptions(), trace_file, &trace_reader));
+    IOTraceReader reader(std::move(trace_reader));
+
+    IOTraceHeader header;
+    ASSERT_OK(reader.ReadHeader(&header));
+    ASSERT_EQ(kMajorVersion, static_cast<int>(header.rocksdb_major_version));
+    ASSERT_EQ(kMinorVersion, static_cast<int>(header.rocksdb_minor_version));
+
+    // Read records.
+    int blob_files_op_count = 0;
+    Status status;
+    while (true) {
+      IOTraceRecord record;
+      status = reader.ReadIOOp(&record);
+      if (!status.ok()) {
+        break;
+      }
+      if (record.file_name.find("blob") != std::string::npos) {
+        blob_files_op_count++;
+      }
+    }
+    // Assuming blob files will have Append, Close and then Read operations.
+    ASSERT_GT(blob_files_op_count, 2);
+  }
+}
+#endif  // !ROCKSDB_LITE
+
 class DBBlobBasicIOErrorTest : public DBBlobBasicTest,
                                public testing::WithParamInterface<std::string> {
  protected:
@@ -298,6 +355,59 @@ TEST_P(DBBlobBasicIOErrorTest, MultiGetBlobs_IOError) {
   ASSERT_TRUE(statuses[1].IsIOError());
 }
 
+namespace {
+
+class ReadBlobCompactionFilter : public CompactionFilter {
+ public:
+  ReadBlobCompactionFilter() = default;
+  const char* Name() const override {
+    return "rocksdb.compaction.filter.read.blob";
+  }
+  CompactionFilter::Decision FilterV2(
+      int /*level*/, const Slice& /*key*/, ValueType value_type,
+      const Slice& existing_value, std::string* new_value,
+      std::string* /*skip_until*/) const override {
+    if (value_type != CompactionFilter::ValueType::kValue) {
+      return CompactionFilter::Decision::kKeep;
+    }
+    assert(new_value);
+    new_value->assign(existing_value.data(), existing_value.size());
+    return CompactionFilter::Decision::kChangeValue;
+  }
+};
+
+}  // anonymous namespace
+
+TEST_P(DBBlobBasicIOErrorTest, CompactionFilterReadBlob_IOError) {
+  Options options = GetDefaultOptions();
+  options.env = fault_injection_env_.get();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  options.create_if_missing = true;
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new ReadBlobCompactionFilter);
+  options.compaction_filter = compaction_filter_guard.get();
+
+  DestroyAndReopen(options);
+  constexpr char key[] = "foo";
+  constexpr char blob_value[] = "foo_blob_value";
+  ASSERT_OK(Put(key, blob_value));
+  ASSERT_OK(Flush());
+
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
+    fault_injection_env_->SetFilesystemActive(false,
+                                              Status::IOError(sync_point_));
+  });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  ASSERT_TRUE(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                                /*end=*/nullptr)
+                  .IsIOError());
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
 }  // namespace ROCKSDB_NAMESPACE
 
 int main(int argc, char** argv) {

db/blob/db_blob_compaction_test.cc

@@ -0,0 +1,399 @@
+//  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_index.h"
+#include "db/db_test_util.h"
+#include "port/stack_trace.h"
+#include "test_util/sync_point.h"
+#include "utilities/fault_injection_env.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class DBBlobCompactionTest : public DBTestBase {
+ public:
+  explicit DBBlobCompactionTest()
+      : DBTestBase("/db_blob_compaction_test", /*env_do_fsync=*/false) {}
+
+#ifndef ROCKSDB_LITE
+  const std::vector<InternalStats::CompactionStats>& GetCompactionStats() {
+    VersionSet* const versions = dbfull()->TEST_GetVersionSet();
+    assert(versions);
+    assert(versions->GetColumnFamilySet());
+
+    ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
+    assert(cfd);
+
+    const InternalStats* const internal_stats = cfd->internal_stats();
+    assert(internal_stats);
+
+    return internal_stats->TEST_GetCompactionStats();
+  }
+#endif  // ROCKSDB_LITE
+};
+
+namespace {
+
+class FilterByKeyLength : public CompactionFilter {
+ public:
+  explicit FilterByKeyLength(size_t len) : length_threshold_(len) {}
+  const char* Name() const override {
+    return "rocksdb.compaction.filter.by.key.length";
+  }
+  CompactionFilter::Decision FilterBlobByKey(
+      int /*level*/, const Slice& key, std::string* /*new_value*/,
+      std::string* /*skip_until*/) const override {
+    if (key.size() < length_threshold_) {
+      return CompactionFilter::Decision::kRemove;
+    }
+    return CompactionFilter::Decision::kKeep;
+  }
+
+ private:
+  size_t length_threshold_;
+};
+
+class BadBlobCompactionFilter : public CompactionFilter {
+ public:
+  explicit BadBlobCompactionFilter(std::string prefix,
+                                   CompactionFilter::Decision filter_by_key,
+                                   CompactionFilter::Decision filter_v2)
+      : prefix_(std::move(prefix)),
+        filter_blob_by_key_(filter_by_key),
+        filter_v2_(filter_v2) {}
+  const char* Name() const override { return "rocksdb.compaction.filter.bad"; }
+  CompactionFilter::Decision FilterBlobByKey(
+      int /*level*/, const Slice& key, std::string* /*new_value*/,
+      std::string* /*skip_until*/) const override {
+    if (key.size() >= prefix_.size() &&
+        0 == strncmp(prefix_.data(), key.data(), prefix_.size())) {
+      return CompactionFilter::Decision::kUndetermined;
+    }
+    return filter_blob_by_key_;
+  }
+  CompactionFilter::Decision FilterV2(
+      int /*level*/, const Slice& /*key*/, ValueType /*value_type*/,
+      const Slice& /*existing_value*/, std::string* /*new_value*/,
+      std::string* /*skip_until*/) const override {
+    return filter_v2_;
+  }
+
+ private:
+  const std::string prefix_;
+  const CompactionFilter::Decision filter_blob_by_key_;
+  const CompactionFilter::Decision filter_v2_;
+};
+
+class ValueBlindWriteFilter : public CompactionFilter {
+ public:
+  explicit ValueBlindWriteFilter(std::string new_val)
+      : new_value_(std::move(new_val)) {}
+  const char* Name() const override {
+    return "rocksdb.compaction.filter.blind.write";
+  }
+  CompactionFilter::Decision FilterBlobByKey(
+      int level, const Slice& key, std::string* new_value,
+      std::string* skip_until) const override;
+
+ private:
+  const std::string new_value_;
+};
+
+CompactionFilter::Decision ValueBlindWriteFilter::FilterBlobByKey(
+    int /*level*/, const Slice& /*key*/, std::string* new_value,
+    std::string* /*skip_until*/) const {
+  assert(new_value);
+  new_value->assign(new_value_);
+  return CompactionFilter::Decision::kChangeValue;
+}
+
+class ValueMutationFilter : public CompactionFilter {
+ public:
+  explicit ValueMutationFilter(std::string padding)
+      : padding_(std::move(padding)) {}
+  const char* Name() const override {
+    return "rocksdb.compaction.filter.value.mutation";
+  }
+  CompactionFilter::Decision FilterV2(int level, const Slice& key,
+                                      ValueType value_type,
+                                      const Slice& existing_value,
+                                      std::string* new_value,
+                                      std::string* skip_until) const override;
+
+ private:
+  const std::string padding_;
+};
+
+CompactionFilter::Decision ValueMutationFilter::FilterV2(
+    int /*level*/, const Slice& /*key*/, ValueType value_type,
+    const Slice& existing_value, std::string* new_value,
+    std::string* /*skip_until*/) const {
+  assert(CompactionFilter::ValueType::kBlobIndex != value_type);
+  if (CompactionFilter::ValueType::kValue != value_type) {
+    return CompactionFilter::Decision::kKeep;
+  }
+  assert(new_value);
+  new_value->assign(existing_value.data(), existing_value.size());
+  new_value->append(padding_);
+  return CompactionFilter::Decision::kChangeValue;
+}
+
+class AlwaysKeepFilter : public CompactionFilter {
+ public:
+  explicit AlwaysKeepFilter() = default;
+  const char* Name() const override {
+    return "rocksdb.compaction.filter.always.keep";
+  }
+  CompactionFilter::Decision FilterV2(
+      int /*level*/, const Slice& /*key*/, ValueType /*value_type*/,
+      const Slice& /*existing_value*/, std::string* /*new_value*/,
+      std::string* /*skip_until*/) const override {
+    return CompactionFilter::Decision::kKeep;
+  }
+};
+}  // anonymous namespace
+
+class DBBlobBadCompactionFilterTest
+    : public DBBlobCompactionTest,
+      public testing::WithParamInterface<
+          std::tuple<std::string, CompactionFilter::Decision,
+                     CompactionFilter::Decision>> {
+ public:
+  explicit DBBlobBadCompactionFilterTest()
+      : compaction_filter_guard_(new BadBlobCompactionFilter(
+            std::get<0>(GetParam()), std::get<1>(GetParam()),
+            std::get<2>(GetParam()))) {}
+
+ protected:
+  std::unique_ptr<CompactionFilter> compaction_filter_guard_;
+};
+
+INSTANTIATE_TEST_CASE_P(
+    BadCompactionFilter, DBBlobBadCompactionFilterTest,
+    testing::Combine(
+        testing::Values("a"),
+        testing::Values(CompactionFilter::Decision::kChangeBlobIndex,
+                        CompactionFilter::Decision::kIOError),
+        testing::Values(CompactionFilter::Decision::kUndetermined,
+                        CompactionFilter::Decision::kChangeBlobIndex,
+                        CompactionFilter::Decision::kIOError)));
+
+TEST_F(DBBlobCompactionTest, FilterByKeyLength) {
+  Options options = GetDefaultOptions();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  options.create_if_missing = true;
+  constexpr size_t kKeyLength = 2;
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new FilterByKeyLength(kKeyLength));
+  options.compaction_filter = compaction_filter_guard.get();
+
+  constexpr char short_key[] = "a";
+  constexpr char long_key[] = "abc";
+  constexpr char blob_value[] = "value";
+
+  DestroyAndReopen(options);
+  ASSERT_OK(Put(short_key, blob_value));
+  ASSERT_OK(Put(long_key, blob_value));
+  ASSERT_OK(Flush());
+  CompactRangeOptions cro;
+  ASSERT_OK(db_->CompactRange(cro, /*begin=*/nullptr, /*end=*/nullptr));
+  std::string value;
+  ASSERT_TRUE(db_->Get(ReadOptions(), short_key, &value).IsNotFound());
+  value.clear();
+  ASSERT_OK(db_->Get(ReadOptions(), long_key, &value));
+  ASSERT_EQ("value", value);
+
+#ifndef ROCKSDB_LITE
+  const auto& compaction_stats = GetCompactionStats();
+  ASSERT_GE(compaction_stats.size(), 2);
+
+  // Filter decides between kKeep and kRemove solely based on key;
+  // this involves neither reading nor writing blobs
+  ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
+  ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
+
+  Close();
+}
+
+TEST_F(DBBlobCompactionTest, BlindWriteFilter) {
+  Options options = GetDefaultOptions();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  options.create_if_missing = true;
+  constexpr char new_blob_value[] = "new_blob_value";
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new ValueBlindWriteFilter(new_blob_value));
+  options.compaction_filter = compaction_filter_guard.get();
+  DestroyAndReopen(options);
+  const std::vector<std::string> keys = {"a", "b", "c"};
+  const std::vector<std::string> values = {"a_value", "b_value", "c_value"};
+  assert(keys.size() == values.size());
+  for (size_t i = 0; i < keys.size(); ++i) {
+    ASSERT_OK(Put(keys[i], values[i]));
+  }
+  ASSERT_OK(Flush());
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                              /*end=*/nullptr));
+  for (const auto& key : keys) {
+    ASSERT_EQ(new_blob_value, Get(key));
+  }
+
+#ifndef ROCKSDB_LITE
+  const auto& compaction_stats = GetCompactionStats();
+  ASSERT_GE(compaction_stats.size(), 2);
+
+  // Filter unconditionally changes value in FilterBlobByKey;
+  // this involves writing but not reading blobs
+  ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
+  ASSERT_GT(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
+
+  Close();
+}
+
+TEST_P(DBBlobBadCompactionFilterTest, BadDecisionFromCompactionFilter) {
+  Options options = GetDefaultOptions();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  options.create_if_missing = true;
+  options.compaction_filter = compaction_filter_guard_.get();
+  DestroyAndReopen(options);
+  ASSERT_OK(Put("b", "value"));
+  ASSERT_OK(Flush());
+  ASSERT_TRUE(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                                /*end=*/nullptr)
+                  .IsNotSupported());
+  Close();
+
+  DestroyAndReopen(options);
+  std::string key(std::get<0>(GetParam()));
+  ASSERT_OK(Put(key, "value"));
+  ASSERT_OK(Flush());
+  ASSERT_TRUE(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                                /*end=*/nullptr)
+                  .IsNotSupported());
+  Close();
+}
+
+TEST_F(DBBlobCompactionTest, CompactionFilter_InlinedTTLIndex) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new ValueMutationFilter(""));
+  options.compaction_filter = compaction_filter_guard.get();
+  DestroyAndReopen(options);
+  constexpr char key[] = "key";
+  constexpr char blob[] = "blob";
+  // Fake an inlined TTL blob index.
+  std::string blob_index;
+  constexpr uint64_t expiration = 1234567890;
+  BlobIndex::EncodeInlinedTTL(&blob_index, expiration, blob);
+  WriteBatch batch;
+  ASSERT_OK(WriteBatchInternal::PutBlobIndex(&batch, 0, key, blob_index));
+  ASSERT_OK(db_->Write(WriteOptions(), &batch));
+  ASSERT_OK(Flush());
+  ASSERT_TRUE(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                                /*end=*/nullptr)
+                  .IsCorruption());
+  Close();
+}
+
+TEST_F(DBBlobCompactionTest, CompactionFilter) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  constexpr char padding[] = "_delta";
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new ValueMutationFilter(padding));
+  options.compaction_filter = compaction_filter_guard.get();
+  DestroyAndReopen(options);
+  const std::vector<std::pair<std::string, std::string>> kvs = {
+      {"a", "a_value"}, {"b", "b_value"}, {"c", "c_value"}};
+  for (const auto& kv : kvs) {
+    ASSERT_OK(Put(kv.first, kv.second));
+  }
+  ASSERT_OK(Flush());
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                              /*end=*/nullptr));
+  for (const auto& kv : kvs) {
+    ASSERT_EQ(kv.second + std::string(padding), Get(kv.first));
+  }
+
+#ifndef ROCKSDB_LITE
+  const auto& compaction_stats = GetCompactionStats();
+  ASSERT_GE(compaction_stats.size(), 2);
+
+  // Filter changes the value using the previous value in FilterV2;
+  // this involves reading and writing blobs
+  ASSERT_GT(compaction_stats[1].bytes_read_blob, 0);
+  ASSERT_GT(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
+
+  Close();
+}
+
+TEST_F(DBBlobCompactionTest, CorruptedBlobIndex) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new ValueMutationFilter(""));
+  options.compaction_filter = compaction_filter_guard.get();
+  DestroyAndReopen(options);
+  // Mock a corrupted blob index
+  constexpr char key[] = "key";
+  std::string blob_idx("blob_idx");
+  WriteBatch write_batch;
+  ASSERT_OK(WriteBatchInternal::PutBlobIndex(&write_batch, 0, key, blob_idx));
+  ASSERT_OK(db_->Write(WriteOptions(), &write_batch));
+  ASSERT_OK(Flush());
+  ASSERT_TRUE(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                                /*end=*/nullptr)
+                  .IsCorruption());
+  Close();
+}
+
+TEST_F(DBBlobCompactionTest, CompactionFilterReadBlobAndKeep) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  std::unique_ptr<CompactionFilter> compaction_filter_guard(
+      new AlwaysKeepFilter());
+  options.compaction_filter = compaction_filter_guard.get();
+  DestroyAndReopen(options);
+  ASSERT_OK(Put("foo", "foo_value"));
+  ASSERT_OK(Flush());
+  std::vector<uint64_t> blob_files = GetBlobFileNumbers();
+  ASSERT_EQ(1, blob_files.size());
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                              /*end=*/nullptr));
+  ASSERT_EQ(blob_files, GetBlobFileNumbers());
+
+#ifndef ROCKSDB_LITE
+  const auto& compaction_stats = GetCompactionStats();
+  ASSERT_GE(compaction_stats.size(), 2);
+
+  // Filter decides to keep the existing value in FilterV2;
+  // this involves reading but not writing blobs
+  ASSERT_GT(compaction_stats[1].bytes_read_blob, 0);
+  ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+#endif  // ROCKSDB_LITE
+
+  Close();
+}
+
+}  // namespace ROCKSDB_NAMESPACE
+
+int main(int argc, char** argv) {
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
+  ::testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}

db/blob/db_blob_corruption_test.cc

@@ -0,0 +1,81 @@
+//  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/db_test_util.h"
+#include "port/stack_trace.h"
+#include "test_util/sync_point.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+class DBBlobCorruptionTest : public DBTestBase {
+ protected:
+  DBBlobCorruptionTest()
+      : DBTestBase("/db_blob_corruption_test", /* env_do_fsync */ false) {}
+
+  void Corrupt(FileType filetype, int offset, int bytes_to_corrupt) {
+    // Pick file to corrupt
+    std::vector<std::string> filenames;
+    ASSERT_OK(env_->GetChildren(dbname_, &filenames));
+    uint64_t number;
+    FileType type;
+    std::string fname;
+    uint64_t picked_number = kInvalidBlobFileNumber;
+    for (size_t i = 0; i < filenames.size(); i++) {
+      if (ParseFileName(filenames[i], &number, &type) && type == filetype &&
+          number > picked_number) {  // Pick latest file
+        fname = dbname_ + "/" + filenames[i];
+        picked_number = number;
+      }
+    }
+    ASSERT_TRUE(!fname.empty()) << filetype;
+    ASSERT_OK(test::CorruptFile(env_, fname, offset, bytes_to_corrupt));
+  }
+};
+
+#ifndef ROCKSDB_LITE
+TEST_F(DBBlobCorruptionTest, VerifyWholeBlobFileChecksum) {
+  Options options = GetDefaultOptions();
+  options.enable_blob_files = true;
+  options.min_blob_size = 0;
+  options.create_if_missing = true;
+  options.file_checksum_gen_factory =
+      ROCKSDB_NAMESPACE::GetFileChecksumGenCrc32cFactory();
+  Reopen(options);
+
+  ASSERT_OK(Put(Slice("key_1"), Slice("blob_value_1")));
+  ASSERT_OK(Flush());
+  ASSERT_OK(Put(Slice("key_2"), Slice("blob_value_2")));
+  ASSERT_OK(Flush());
+  ASSERT_OK(db_->VerifyFileChecksums(ReadOptions()));
+  Close();
+
+  Corrupt(kBlobFile, 0, 2);
+
+  ASSERT_OK(TryReopen(options));
+
+  int count{0};
+  SyncPoint::GetInstance()->SetCallBack(
+      "DBImpl::VerifyFullFileChecksum:mismatch", [&](void* arg) {
+        const Status* s = static_cast<Status*>(arg);
+        ASSERT_NE(s, nullptr);
+        ++count;
+        ASSERT_NOK(*s);
+      });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  ASSERT_TRUE(db_->VerifyFileChecksums(ReadOptions()).IsCorruption());
+  ASSERT_EQ(1, count);
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+#endif  // !ROCKSDB_LITE
+}  // namespace ROCKSDB_NAMESPACE
+
+int main(int argc, char** argv) {
+  ROCKSDB_NAMESPACE::port::InstallStackTraceHandler();
+  ::testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}

db/builder.cc

@@ -23,11 +23,13 @@
 #include "db/range_del_aggregator.h"
 #include "db/table_cache.h"
 #include "db/version_edit.h"
+#include "file/file_util.h"
 #include "file/filename.h"
 #include "file/read_write_util.h"
 #include "file/writable_file_writer.h"
 #include "monitoring/iostats_context_imp.h"
 #include "monitoring/thread_status_util.h"
+#include "options/options_helper.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
 #include "rocksdb/iterator.h"
@@ -91,7 +93,8 @@ Status BuildTable(
     TableProperties* table_properties, int level, const uint64_t creation_time,
     const uint64_t oldest_key_time, Env::WriteLifeTimeHint write_hint,
     const uint64_t file_creation_time, const std::string& db_id,
-    const std::string& db_session_id, const std::string* full_history_ts_low) {
+    const std::string& db_session_id, const std::string* full_history_ts_low,
+    BlobFileCompletionCallback* blob_callback) {
   assert((column_family_id ==
           TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
          column_family_name.empty());
@@ -124,6 +127,7 @@ Status BuildTable(
   assert(env);
   FileSystem* fs = db_options.fs.get();
   assert(fs);
+
   TableProperties tp;
   if (iter->Valid() || !range_del_agg->IsEmpty()) {
     TableBuilder* builder;
@@ -147,13 +151,14 @@ Status BuildTable(
             file_checksum, file_checksum_func_name);
         return s;
       }
+      FileTypeSet tmp_set = ioptions.checksum_handoff_file_types;
       file->SetIOPriority(io_priority);
       file->SetWriteLifeTimeHint(write_hint);
-
       file_writer.reset(new WritableFileWriter(
-          std::move(file), fname, file_options, env, io_tracer,
+          std::move(file), fname, file_options, ioptions.clock, io_tracer,
           ioptions.statistics, ioptions.listeners,
-          ioptions.file_checksum_gen_factory));
+          ioptions.file_checksum_gen_factory,
+          tmp_set.Contains(FileType::kTableFile)));
 
       builder = NewTableBuilder(
           ioptions, mutable_cf_options, internal_comparator,
@@ -172,10 +177,10 @@ Status BuildTable(
 
     std::unique_ptr<BlobFileBuilder> blob_file_builder(
         (mutable_cf_options.enable_blob_files && blob_file_additions)
-            ? new BlobFileBuilder(versions, env, fs, &ioptions,
-                                  &mutable_cf_options, &file_options, job_id,
-                                  column_family_id, column_family_name,
-                                  io_priority, write_hint, &blob_file_paths,
+            ? new BlobFileBuilder(versions, fs, &ioptions, &mutable_cf_options,
+                                  &file_options, job_id, column_family_id,
+                                  column_family_name, io_priority, write_hint,
+                                  io_tracer, blob_callback, &blob_file_paths,
                                   blob_file_additions)
             : nullptr);
 
@@ -254,7 +259,7 @@ Status BuildTable(
     // Finish and check for file errors
     TEST_SYNC_POINT("BuildTable:BeforeSyncTable");
     if (s.ok() && !empty) {
-      StopWatch sw(env, ioptions.statistics, TABLE_SYNC_MICROS);
+      StopWatch sw(ioptions.clock, ioptions.statistics, TABLE_SYNC_MICROS);
       *io_status = file_writer->Sync(ioptions.use_fsync);
     }
     TEST_SYNC_POINT("BuildTable:BeforeCloseTableFile");
@@ -276,8 +281,9 @@ Status BuildTable(
     if (blob_file_builder) {
       if (s.ok()) {
         s = blob_file_builder->Finish();
+      } else {
+        blob_file_builder->Abandon();
       }
-
       blob_file_builder.reset();
     }
 
@@ -337,8 +343,10 @@ Status BuildTable(
 
     if (blob_file_additions) {
       for (const std::string& blob_file_path : blob_file_paths) {
-        ignored = fs->DeleteFile(blob_file_path, IOOptions(), dbg);
+        ignored = DeleteDBFile(&db_options, blob_file_path, dbname,
+                               /*force_bg=*/false, /*force_fg=*/false);
         ignored.PermitUncheckedError();
+        TEST_SYNC_POINT("BuildTable::AfterDeleteFile");
       }
     }
   }

db/builder.h

@@ -38,6 +38,7 @@ class VersionEdit;
 class TableBuilder;
 class WritableFileWriter;
 class InternalStats;
+class BlobFileCompletionCallback;
 
 // @param column_family_name Name of the column family that is also identified
 //    by column_family_id, or empty string if unknown. It must outlive the
@@ -90,6 +91,7 @@ extern Status BuildTable(
     Env::WriteLifeTimeHint write_hint = Env::WLTH_NOT_SET,
     const uint64_t file_creation_time = 0, const std::string& db_id = "",
     const std::string& db_session_id = "",
-    const std::string* full_history_ts_low = nullptr);
+    const std::string* full_history_ts_low = nullptr,
+    BlobFileCompletionCallback* blob_callback = nullptr);
 
 }  // namespace ROCKSDB_NAMESPACE

db/c.cc

@@ -11,7 +11,11 @@
 
 #include "rocksdb/c.h"
 
-#include <stdlib.h>
+#include <cstdlib>
+#include <map>
+#include <unordered_set>
+#include <vector>
+
 #include "port/port.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/compaction_filter.h"
@@ -24,6 +28,7 @@
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/merge_operator.h"
 #include "rocksdb/options.h"
+#include "rocksdb/perf_context.h"
 #include "rocksdb/rate_limiter.h"
 #include "rocksdb/slice_transform.h"
 #include "rocksdb/statistics.h"
@@ -39,13 +44,8 @@
 #include "rocksdb/utilities/transaction_db.h"
 #include "rocksdb/utilities/write_batch_with_index.h"
 #include "rocksdb/write_batch.h"
-#include "rocksdb/perf_context.h"
 #include "utilities/merge_operators.h"
 
-#include <vector>
-#include <unordered_set>
-#include <map>
-
 using ROCKSDB_NAMESPACE::BackupableDBOptions;
 using ROCKSDB_NAMESPACE::BackupEngine;
 using ROCKSDB_NAMESPACE::BackupID;
@@ -2774,6 +2774,14 @@ void rocksdb_options_set_bottommost_compression_options_zstd_max_train_bytes(
   opt->rep.bottommost_compression_opts.enabled = enabled;
 }
 
+void rocksdb_options_set_bottommost_compression_options_max_dict_buffer_bytes(
+    rocksdb_options_t* opt, uint64_t max_dict_buffer_bytes,
+    unsigned char enabled) {
+  opt->rep.bottommost_compression_opts.max_dict_buffer_bytes =
+      max_dict_buffer_bytes;
+  opt->rep.bottommost_compression_opts.enabled = enabled;
+}
+
 void rocksdb_options_set_compression_options(rocksdb_options_t* opt, int w_bits,
                                              int level, int strategy,
                                              int max_dict_bytes) {
@@ -2788,6 +2796,11 @@ void rocksdb_options_set_compression_options_zstd_max_train_bytes(
   opt->rep.compression_opts.zstd_max_train_bytes = zstd_max_train_bytes;
 }
 
+void rocksdb_options_set_compression_options_max_dict_buffer_bytes(
+    rocksdb_options_t* opt, uint64_t max_dict_buffer_bytes) {
+  opt->rep.compression_opts.max_dict_buffer_bytes = max_dict_buffer_bytes;
+}
+
 void rocksdb_options_set_prefix_extractor(
     rocksdb_options_t* opt, rocksdb_slicetransform_t* prefix_extractor) {
   opt->rep.prefix_extractor.reset(prefix_extractor);
@@ -2946,6 +2959,8 @@ void rocksdb_options_set_access_hint_on_compaction_start(
       opt->rep.access_hint_on_compaction_start =
           ROCKSDB_NAMESPACE::Options::WILLNEED;
       break;
+    default:
+      assert(0);
   }
 }
 
@@ -3936,6 +3951,25 @@ unsigned char rocksdb_readoptions_get_ignore_range_deletions(
   return opt->rep.ignore_range_deletions;
 }
 
+void rocksdb_readoptions_set_deadline(rocksdb_readoptions_t* opt,
+                                      uint64_t microseconds) {
+  opt->rep.deadline = std::chrono::microseconds(microseconds);
+}
+
+uint64_t rocksdb_readoptions_get_deadline(rocksdb_readoptions_t* opt) {
+  return opt->rep.deadline.count();
+}
+
+void rocksdb_readoptions_set_io_timeout(rocksdb_readoptions_t* opt,
+                                        uint64_t microseconds) {
+  opt->rep.io_timeout = std::chrono::microseconds(microseconds);
+}
+
+extern ROCKSDB_LIBRARY_API uint64_t
+rocksdb_readoptions_get_io_timeout(rocksdb_readoptions_t* opt) {
+  return opt->rep.io_timeout.count();
+}
+
 rocksdb_writeoptions_t* rocksdb_writeoptions_create() {
   return new rocksdb_writeoptions_t;
 }

db/c_test.c

@@ -1119,9 +1119,8 @@ int main(int argc, char** argv) {
         // Essentially a fingerprint of the block-based Bloom schema
         CheckCondition(hits == 241);
       } else {
-        // Essentially a fingerprint of the full Bloom schema(s),
-        // format_version < 5, which vary for three different CACHE_LINE_SIZEs
-        CheckCondition(hits == 224 || hits == 180 || hits == 125);
+        // Essentially a fingerprint of full Bloom schema, format_version=5
+        CheckCondition(hits == 188);
       }
       CheckCondition(
           (keys_to_query - hits) ==
@@ -2281,6 +2280,12 @@ int main(int argc, char** argv) {
     rocksdb_readoptions_set_ignore_range_deletions(ro, 1);
     CheckCondition(1 == rocksdb_readoptions_get_ignore_range_deletions(ro));
 
+    rocksdb_readoptions_set_deadline(ro, 300);
+    CheckCondition(300 == rocksdb_readoptions_get_deadline(ro));
+
+    rocksdb_readoptions_set_io_timeout(ro, 400);
+    CheckCondition(400 == rocksdb_readoptions_get_io_timeout(ro));
+
     rocksdb_readoptions_destroy(ro);
   }
 
@@ -2875,7 +2880,7 @@ int main(int argc, char** argv) {
 
 #else
 
-int main() {
+int main(void) {
   fprintf(stderr, "SKIPPED\n");
   return 0;
 }

db/column_family.cc

@@ -33,6 +33,7 @@
 #include "monitoring/thread_status_util.h"
 #include "options/options_helper.h"
 #include "port/port.h"
+#include "rocksdb/convenience.h"
 #include "rocksdb/table.h"
 #include "table/merging_iterator.h"
 #include "util/autovector.h"
@@ -557,12 +558,12 @@ ColumnFamilyData::ColumnFamilyData(
   // if _dummy_versions is nullptr, then this is a dummy column family.
   if (_dummy_versions != nullptr) {
     internal_stats_.reset(
-        new InternalStats(ioptions_.num_levels, db_options.env, this));
+        new InternalStats(ioptions_.num_levels, ioptions_.clock, this));
     table_cache_.reset(new TableCache(ioptions_, file_options, _table_cache,
                                       block_cache_tracer, io_tracer));
     blob_file_cache_.reset(
         new BlobFileCache(_table_cache, ioptions(), soptions(), id_,
-                          internal_stats_->GetBlobFileReadHist()));
+                          internal_stats_->GetBlobFileReadHist(), io_tracer));
 
     if (ioptions_.compaction_style == kCompactionStyleLevel) {
       compaction_picker_.reset(
@@ -632,7 +633,7 @@ ColumnFamilyData::~ColumnFamilyData() {
 
   if (dummy_versions_ != nullptr) {
     // List must be empty
-    assert(dummy_versions_->TEST_Next() == dummy_versions_);
+    assert(dummy_versions_->Next() == dummy_versions_);
     bool deleted __attribute__((__unused__));
     deleted = dummy_versions_->Unref();
     assert(deleted);
@@ -834,7 +835,8 @@ std::pair<WriteStallCondition, ColumnFamilyData::WriteStallCause>
 ColumnFamilyData::GetWriteStallConditionAndCause(
     int num_unflushed_memtables, int num_l0_files,
     uint64_t num_compaction_needed_bytes,
-    const MutableCFOptions& mutable_cf_options) {
+    const MutableCFOptions& mutable_cf_options,
+    const ImmutableCFOptions& immutable_cf_options) {
   if (num_unflushed_memtables >= mutable_cf_options.max_write_buffer_number) {
     return {WriteStallCondition::kStopped, WriteStallCause::kMemtableLimit};
   } else if (!mutable_cf_options.disable_auto_compactions &&
@@ -848,7 +850,9 @@ ColumnFamilyData::GetWriteStallConditionAndCause(
             WriteStallCause::kPendingCompactionBytes};
   } else if (mutable_cf_options.max_write_buffer_number > 3 &&
              num_unflushed_memtables >=
-                 mutable_cf_options.max_write_buffer_number - 1) {
+                 mutable_cf_options.max_write_buffer_number - 1 &&
+             num_unflushed_memtables - 1 >=
+                 immutable_cf_options.min_write_buffer_number_to_merge) {
     return {WriteStallCondition::kDelayed, WriteStallCause::kMemtableLimit};
   } else if (!mutable_cf_options.disable_auto_compactions &&
              mutable_cf_options.level0_slowdown_writes_trigger >= 0 &&
@@ -876,7 +880,8 @@ WriteStallCondition ColumnFamilyData::RecalculateWriteStallConditions(
 
     auto write_stall_condition_and_cause = GetWriteStallConditionAndCause(
         imm()->NumNotFlushed(), vstorage->l0_delay_trigger_count(),
-        vstorage->estimated_compaction_needed_bytes(), mutable_cf_options);
+        vstorage->estimated_compaction_needed_bytes(), mutable_cf_options,
+        *ioptions());
     write_stall_condition = write_stall_condition_and_cause.first;
     auto write_stall_cause = write_stall_condition_and_cause.second;
 
@@ -1355,19 +1360,19 @@ Status ColumnFamilyData::ValidateOptions(
 
 #ifndef ROCKSDB_LITE
 Status ColumnFamilyData::SetOptions(
-    const DBOptions& db_options,
+    const DBOptions& db_opts,
     const std::unordered_map<std::string, std::string>& options_map) {
-  MutableCFOptions new_mutable_cf_options;
-  Status s =
-      GetMutableOptionsFromStrings(mutable_cf_options_, options_map,
-                                   ioptions_.info_log, &new_mutable_cf_options);
+  ColumnFamilyOptions cf_opts =
+      BuildColumnFamilyOptions(initial_cf_options_, mutable_cf_options_);
+  ConfigOptions config_opts;
+  config_opts.mutable_options_only = true;
+  Status s = GetColumnFamilyOptionsFromMap(config_opts, cf_opts, options_map,
+                                           &cf_opts);
   if (s.ok()) {
-    ColumnFamilyOptions cf_options =
-        BuildColumnFamilyOptions(initial_cf_options_, new_mutable_cf_options);
-    s = ValidateOptions(db_options, cf_options);
+    s = ValidateOptions(db_opts, cf_opts);
   }
   if (s.ok()) {
-    mutable_cf_options_ = new_mutable_cf_options;
+    mutable_cf_options_ = MutableCFOptions(cf_opts);
     mutable_cf_options_.RefreshDerivedOptions(ioptions_);
   }
   return s;

db/column_family.h

@@ -475,9 +475,11 @@ class ColumnFamilyData {
     kPendingCompactionBytes,
   };
   static std::pair<WriteStallCondition, WriteStallCause>
-  GetWriteStallConditionAndCause(int num_unflushed_memtables, int num_l0_files,
-                                 uint64_t num_compaction_needed_bytes,
-                                 const MutableCFOptions& mutable_cf_options);
+  GetWriteStallConditionAndCause(
+      int num_unflushed_memtables, int num_l0_files,
+      uint64_t num_compaction_needed_bytes,
+      const MutableCFOptions& mutable_cf_options,
+      const ImmutableCFOptions& immutable_cf_options);
 
   // Recalculate some small conditions, which are changed only during
   // compaction, adding new memtable and/or

db/compaction/compaction_iteration_stats.h

@@ -34,4 +34,8 @@ struct CompactionIterationStats {
   // Single-Delete diagnostics for exceptional situations
   uint64_t num_single_del_fallthru = 0;
   uint64_t num_single_del_mismatch = 0;
+
+  // Blob related statistics
+  uint64_t num_blobs_read = 0;
+  uint64_t total_blob_bytes_read = 0;
 };

db/compaction/compaction_iterator.cc

@@ -80,6 +80,7 @@ CompactionIterator::CompactionIterator(
       earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
       snapshot_checker_(snapshot_checker),
       env_(env),
+      clock_(env_->GetSystemClock().get()),
       report_detailed_time_(report_detailed_time),
       expect_valid_internal_key_(expect_valid_internal_key),
       range_del_agg_(range_del_agg),
@@ -203,72 +204,146 @@ void CompactionIterator::Next() {
 
 bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
                                               Slice* skip_until) {
-  if (compaction_filter_ != nullptr &&
-      (ikey_.type == kTypeValue || ikey_.type == kTypeBlobIndex)) {
-    // If the user has specified a compaction filter and the sequence
-    // number is greater than any external snapshot, then invoke the
-    // filter. If the return value of the compaction filter is true,
-    // replace the entry with a deletion marker.
-    CompactionFilter::Decision filter;
-    compaction_filter_value_.clear();
-    compaction_filter_skip_until_.Clear();
-    CompactionFilter::ValueType value_type =
-        ikey_.type == kTypeValue ? CompactionFilter::ValueType::kValue
-                                 : CompactionFilter::ValueType::kBlobIndex;
-    // Hack: pass internal key to BlobIndexCompactionFilter since it needs
-    // to get sequence number.
-    Slice& filter_key = ikey_.type == kTypeValue ? ikey_.user_key : key_;
-    {
-      StopWatchNano timer(env_, report_detailed_time_);
+  if (!compaction_filter_ ||
+      (ikey_.type != kTypeValue && ikey_.type != kTypeBlobIndex)) {
+    return true;
+  }
+  bool error = false;
+  // If the user has specified a compaction filter and the sequence
+  // number is greater than any external snapshot, then invoke the
+  // filter. If the return value of the compaction filter is true,
+  // replace the entry with a deletion marker.
+  CompactionFilter::Decision filter = CompactionFilter::Decision::kUndetermined;
+  compaction_filter_value_.clear();
+  compaction_filter_skip_until_.Clear();
+  CompactionFilter::ValueType value_type =
+      ikey_.type == kTypeValue ? CompactionFilter::ValueType::kValue
+                               : CompactionFilter::ValueType::kBlobIndex;
+  // Hack: pass internal key to BlobIndexCompactionFilter since it needs
+  // to get sequence number.
+  assert(compaction_filter_);
+  Slice& filter_key =
+      (ikey_.type == kTypeValue ||
+       !compaction_filter_->IsStackedBlobDbInternalCompactionFilter())
+          ? ikey_.user_key
+          : key_;
+  {
+    StopWatchNano timer(clock_, report_detailed_time_);
+    if (kTypeBlobIndex == ikey_.type) {
+      blob_value_.Reset();
+      filter = compaction_filter_->FilterBlobByKey(
+          compaction_->level(), filter_key, &compaction_filter_value_,
+          compaction_filter_skip_until_.rep());
+      if (CompactionFilter::Decision::kUndetermined == filter &&
+          !compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
+        // For integrated BlobDB impl, CompactionIterator reads blob value.
+        // For Stacked BlobDB impl, the corresponding CompactionFilter's
+        // FilterV2 method should read the blob value.
+        BlobIndex blob_index;
+        Status s = blob_index.DecodeFrom(value_);
+        if (!s.ok()) {
+          status_ = s;
+          valid_ = false;
+          return false;
+        }
+        if (blob_index.HasTTL() || blob_index.IsInlined()) {
+          status_ = Status::Corruption("Unexpected TTL/inlined blob index");
+          valid_ = false;
+          return false;
+        }
+        const Version* const version = compaction_->input_version();
+        assert(version);
+
+        uint64_t bytes_read = 0;
+        s = version->GetBlob(ReadOptions(), ikey_.user_key, blob_index,
+                             &blob_value_, &bytes_read);
+        if (!s.ok()) {
+          status_ = s;
+          valid_ = false;
+          return false;
+        }
+
+        ++iter_stats_.num_blobs_read;
+        iter_stats_.total_blob_bytes_read += bytes_read;
+
+        value_type = CompactionFilter::ValueType::kValue;
+      }
+    }
+    if (CompactionFilter::Decision::kUndetermined == filter) {
       filter = compaction_filter_->FilterV2(
-          compaction_->level(), filter_key, value_type, value_,
-          &compaction_filter_value_, compaction_filter_skip_until_.rep());
-      iter_stats_.total_filter_time +=
-          env_ != nullptr && report_detailed_time_ ? timer.ElapsedNanos() : 0;
+          compaction_->level(), filter_key, value_type,
+          blob_value_.empty() ? value_ : blob_value_, &compaction_filter_value_,
+          compaction_filter_skip_until_.rep());
     }
+    iter_stats_.total_filter_time +=
+        env_ != nullptr && report_detailed_time_ ? timer.ElapsedNanos() : 0;
+  }
 
-    if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil &&
-        cmp_->Compare(*compaction_filter_skip_until_.rep(), ikey_.user_key) <=
-            0) {
-      // Can't skip to a key smaller than the current one.
-      // Keep the key as per FilterV2 documentation.
-      filter = CompactionFilter::Decision::kKeep;
+  if (CompactionFilter::Decision::kUndetermined == filter) {
+    // Should not reach here, since FilterV2 should never return kUndetermined.
+    status_ =
+        Status::NotSupported("FilterV2() should never return kUndetermined");
+    valid_ = false;
+    return false;
+  }
+
+  if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil &&
+      cmp_->Compare(*compaction_filter_skip_until_.rep(), ikey_.user_key) <=
+          0) {
+    // Can't skip to a key smaller than the current one.
+    // Keep the key as per FilterV2 documentation.
+    filter = CompactionFilter::Decision::kKeep;
+  }
+
+  if (filter == CompactionFilter::Decision::kRemove) {
+    // convert the current key to a delete; key_ is pointing into
+    // current_key_ at this point, so updating current_key_ updates key()
+    ikey_.type = kTypeDeletion;
+    current_key_.UpdateInternalKey(ikey_.sequence, kTypeDeletion);
+    // no value associated with delete
+    value_.clear();
+    iter_stats_.num_record_drop_user++;
+  } else if (filter == CompactionFilter::Decision::kChangeValue) {
+    if (ikey_.type == kTypeBlobIndex) {
+      // value transfer from blob file to inlined data
+      ikey_.type = kTypeValue;
+      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
     }
-
-    if (filter == CompactionFilter::Decision::kRemove) {
-      // convert the current key to a delete; key_ is pointing into
-      // current_key_ at this point, so updating current_key_ updates key()
-      ikey_.type = kTypeDeletion;
-      current_key_.UpdateInternalKey(ikey_.sequence, kTypeDeletion);
-      // no value associated with delete
-      value_.clear();
-      iter_stats_.num_record_drop_user++;
-    } else if (filter == CompactionFilter::Decision::kChangeValue) {
-      if (ikey_.type == kTypeBlobIndex) {
-        // value transfer from blob file to inlined data
-        ikey_.type = kTypeValue;
-        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-      }
-      value_ = compaction_filter_value_;
-    } else if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil) {
-      *need_skip = true;
-      compaction_filter_skip_until_.ConvertFromUserKey(kMaxSequenceNumber,
-                                                       kValueTypeForSeek);
-      *skip_until = compaction_filter_skip_until_.Encode();
-    } else if (filter == CompactionFilter::Decision::kChangeBlobIndex) {
-      if (ikey_.type == kTypeValue) {
-        // value transfer from inlined data to blob file
-        ikey_.type = kTypeBlobIndex;
-        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
-      }
-      value_ = compaction_filter_value_;
-    } else if (filter == CompactionFilter::Decision::kIOError) {
-      status_ =
-          Status::IOError("Failed to access blob during compaction filter");
+    value_ = compaction_filter_value_;
+  } else if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil) {
+    *need_skip = true;
+    compaction_filter_skip_until_.ConvertFromUserKey(kMaxSequenceNumber,
+                                                     kValueTypeForSeek);
+    *skip_until = compaction_filter_skip_until_.Encode();
+  } else if (filter == CompactionFilter::Decision::kChangeBlobIndex) {
+    // Only the StackableDB-based BlobDB impl's compaction filter should return
+    // kChangeBlobIndex. Decision about rewriting blob and changing blob index
+    // in the integrated BlobDB impl is made in subsequent call to
+    // PrepareOutput() and its callees.
+    if (!compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
+      status_ = Status::NotSupported(
+          "Only stacked BlobDB's internal compaction filter can return "
+          "kChangeBlobIndex.");
+      valid_ = false;
       return false;
     }
+    if (ikey_.type == kTypeValue) {
+      // value transfer from inlined data to blob file
+      ikey_.type = kTypeBlobIndex;
+      current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
+    }
+    value_ = compaction_filter_value_;
+  } else if (filter == CompactionFilter::Decision::kIOError) {
+    if (!compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
+      status_ = Status::NotSupported(
+          "CompactionFilter for integrated BlobDB should not return kIOError");
+      valid_ = false;
+      return false;
+    }
+    status_ = Status::IOError("Failed to access blob during compaction filter");
+    error = true;
   }
-  return true;
+  return !error;
 }
 
 void CompactionIterator::NextFromInput() {
@@ -315,13 +390,11 @@ void CompactionIterator::NextFromInput() {
     // merge_helper_->compaction_filter_skip_until_.
     Slice skip_until;
 
-    int cmp_user_key_without_ts = 0;
+    bool user_key_equal_without_ts = false;
     int cmp_ts = 0;
     if (has_current_user_key_) {
-      cmp_user_key_without_ts =
-          timestamp_size_
-              ? cmp_->CompareWithoutTimestamp(ikey_.user_key, current_user_key_)
-              : cmp_->Compare(ikey_.user_key, current_user_key_);
+      user_key_equal_without_ts =
+          cmp_->EqualWithoutTimestamp(ikey_.user_key, current_user_key_);
       // if timestamp_size_ > 0, then curr_ts_ has been initialized by a
       // previous key.
       cmp_ts = timestamp_size_ ? cmp_->CompareTimestamp(
@@ -334,7 +407,7 @@ void CompactionIterator::NextFromInput() {
     // Check whether the user key changed. After this if statement current_key_
     // is a copy of the current input key (maybe converted to a delete by the
     // compaction filter). ikey_.user_key is pointing to the copy.
-    if (!has_current_user_key_ || cmp_user_key_without_ts != 0 || cmp_ts != 0) {
+    if (!has_current_user_key_ || !user_key_equal_without_ts || cmp_ts != 0) {
       // First occurrence of this user key
       // Copy key for output
       key_ = current_key_.SetInternalKey(key_, &ikey_);
@@ -355,7 +428,7 @@ void CompactionIterator::NextFromInput() {
       // consider this key for GC, e.g. it may be dropped if certain conditions
       // match.
       if (!has_current_user_key_ || !timestamp_size_ || !full_history_ts_low_ ||
-          0 != cmp_user_key_without_ts || cmp_with_history_ts_low_ >= 0) {
+          !user_key_equal_without_ts || cmp_with_history_ts_low_ >= 0) {
         // Initialize for future comparison for rule (A) and etc.
         current_user_key_sequence_ = kMaxSequenceNumber;
         current_user_key_snapshot_ = 0;
@@ -422,8 +495,8 @@ void CompactionIterator::NextFromInput() {
       // In the previous iteration we encountered a single delete that we could
       // not compact out.  We will keep this Put, but can drop it's data.
       // (See Optimization 3, below.)
-      assert(ikey_.type == kTypeValue);
-      if (ikey_.type != kTypeValue) {
+      assert(ikey_.type == kTypeValue || ikey_.type == kTypeBlobIndex);
+      if (ikey_.type != kTypeValue && ikey_.type != kTypeBlobIndex) {
         ROCKS_LOG_FATAL(info_log_,
                         "Unexpected key type %d for compaction output",
                         ikey_.type);
@@ -436,6 +509,11 @@ void CompactionIterator::NextFromInput() {
                         current_user_key_snapshot_, last_snapshot);
       }
 
+      if (ikey_.type == kTypeBlobIndex) {
+        ikey_.type = kTypeValue;
+        current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
+      }
+
       value_.clear();
       valid_ = true;
       clear_and_output_next_key_ = false;
@@ -644,8 +722,7 @@ void CompactionIterator::NextFromInput() {
              input_->Valid() &&
              (ParseInternalKey(input_->key(), &next_ikey, allow_data_in_errors_)
                   .ok()) &&
-             0 == cmp_->CompareWithoutTimestamp(ikey_.user_key,
-                                                next_ikey.user_key) &&
+             cmp_->EqualWithoutTimestamp(ikey_.user_key, next_ikey.user_key) &&
              (prev_snapshot == 0 ||
               DEFINITELY_NOT_IN_SNAPSHOT(next_ikey.sequence, prev_snapshot))) {
         input_->Next();
@@ -655,8 +732,7 @@ void CompactionIterator::NextFromInput() {
       if (input_->Valid() &&
           (ParseInternalKey(input_->key(), &next_ikey, allow_data_in_errors_)
                .ok()) &&
-          0 == cmp_->CompareWithoutTimestamp(ikey_.user_key,
-                                             next_ikey.user_key)) {
+          cmp_->EqualWithoutTimestamp(ikey_.user_key, next_ikey.user_key)) {
         valid_ = true;
         at_next_ = true;
       }
@@ -809,9 +885,11 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
     const Version* const version = compaction_->input_version();
     assert(version);
 
+    uint64_t bytes_read = 0;
+
     {
-      const Status s =
-          version->GetBlob(ReadOptions(), user_key(), blob_index, &blob_value_);
+      const Status s = version->GetBlob(ReadOptions(), user_key(), blob_index,
+                                        &blob_value_, &bytes_read);
 
       if (!s.ok()) {
         status_ = s;
@@ -821,6 +899,9 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
       }
     }
 
+    ++iter_stats_.num_blobs_read;
+    iter_stats_.total_blob_bytes_read += bytes_read;
+
     value_ = blob_value_;
 
     if (ExtractLargeValueIfNeededImpl()) {
@@ -834,7 +915,8 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
   }
 
   // GC for stacked BlobDB
-  if (compaction_filter_) {
+  if (compaction_filter_ &&
+      compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
     const auto blob_decision = compaction_filter_->PrepareBlobOutput(
         user_key(), value_, &compaction_filter_value_);
 

db/compaction/compaction_iterator.h

@@ -248,6 +248,7 @@ class CompactionIterator {
   const SequenceNumber earliest_write_conflict_snapshot_;
   const SnapshotChecker* const snapshot_checker_;
   Env* env_;
+  SystemClock* clock_;
   bool report_detailed_time_;
   bool expect_valid_internal_key_;
   CompactionRangeDelAggregator* range_del_agg_;

db/compaction/compaction_iterator_test.cc

@@ -970,6 +970,21 @@ TEST_F(CompactionIteratorWithSnapshotCheckerTest,
           2 /*earliest_write_conflict_snapshot*/);
 }
 
+// Same as above but with a blob index. In addition to the value getting
+// trimmed, the type of the KV is changed to kTypeValue.
+TEST_F(CompactionIteratorWithSnapshotCheckerTest,
+       KeepSingleDeletionForWriteConflictChecking_BlobIndex) {
+  AddSnapshot(2, 0);
+  RunTest({test::KeyStr("a", 2, kTypeSingleDeletion),
+           test::KeyStr("a", 1, kTypeBlobIndex)},
+          {"", "fake_blob_index"},
+          {test::KeyStr("a", 2, kTypeSingleDeletion),
+           test::KeyStr("a", 1, kTypeValue)},
+          {"", ""}, 2 /*last_committed_seq*/, nullptr /*merge_operator*/,
+          nullptr /*compaction_filter*/, false /*bottommost_level*/,
+          2 /*earliest_write_conflict_snapshot*/);
+}
+
 // Compaction filter should keep uncommitted key as-is, and
 //   * Convert the latest velue to deletion, and/or
 //   * if latest value is a merge, apply filter to all suequent merges.

db/compaction/compaction_job.cc

@@ -46,6 +46,7 @@
 #include "monitoring/iostats_context_imp.h"
 #include "monitoring/perf_context_imp.h"
 #include "monitoring/thread_status_util.h"
+#include "options/options_helper.h"
 #include "port/port.h"
 #include "rocksdb/db.h"
 #include "rocksdb/env.h"
@@ -309,7 +310,8 @@ CompactionJob::CompactionJob(
     const std::string& dbname, CompactionJobStats* compaction_job_stats,
     Env::Priority thread_pri, const std::shared_ptr<IOTracer>& io_tracer,
     const std::atomic<int>* manual_compaction_paused, const std::string& db_id,
-    const std::string& db_session_id, std::string full_history_ts_low)
+    const std::string& db_session_id, std::string full_history_ts_low,
+    BlobFileCompletionCallback* blob_callback)
     : job_id_(job_id),
       compact_(new CompactionState(compaction)),
       compaction_job_stats_(compaction_job_stats),
@@ -345,7 +347,8 @@ CompactionJob::CompactionJob(
       measure_io_stats_(measure_io_stats),
       write_hint_(Env::WLTH_NOT_SET),
       thread_pri_(thread_pri),
-      full_history_ts_low_(std::move(full_history_ts_low)) {
+      full_history_ts_low_(std::move(full_history_ts_low)),
+      blob_callback_(blob_callback) {
   assert(compaction_job_stats_ != nullptr);
   assert(log_buffer_ != nullptr);
   const auto* cfd = compact_->compaction->column_family_data();
@@ -419,7 +422,7 @@ void CompactionJob::Prepare() {
 
   if (c->ShouldFormSubcompactions()) {
     {
-      StopWatch sw(env_, stats_, SUBCOMPACTION_SETUP_TIME);
+      StopWatch sw(db_options_.clock, stats_, SUBCOMPACTION_SETUP_TIME);
       GenSubcompactionBoundaries();
     }
     assert(sizes_.size() == boundaries_.size() + 1);
@@ -585,7 +588,7 @@ Status CompactionJob::Run() {
 
   const size_t num_threads = compact_->sub_compact_states.size();
   assert(num_threads > 0);
-  const uint64_t start_micros = env_->NowMicros();
+  const uint64_t start_micros = db_options_.clock->NowMicros();
 
   // Launch a thread for each of subcompactions 1...num_threads-1
   std::vector<port::Thread> thread_pool;
@@ -604,7 +607,7 @@ Status CompactionJob::Run() {
     thread.join();
   }
 
-  compaction_stats_.micros = env_->NowMicros() - start_micros;
+  compaction_stats_.micros = db_options_.clock->NowMicros() - start_micros;
   compaction_stats_.cpu_micros = 0;
   for (size_t i = 0; i < compact_->sub_compact_states.size(); i++) {
     compaction_stats_.cpu_micros +=
@@ -792,39 +795,45 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
   double bytes_read_per_sec = 0;
   double bytes_written_per_sec = 0;
 
-  if (stats.bytes_read_non_output_levels > 0) {
-    read_write_amp = (stats.bytes_written + stats.bytes_read_output_level +
-                      stats.bytes_read_non_output_levels) /
-                     static_cast<double>(stats.bytes_read_non_output_levels);
-    write_amp = stats.bytes_written /
-                static_cast<double>(stats.bytes_read_non_output_levels);
+  const uint64_t bytes_read_non_output_and_blob =
+      stats.bytes_read_non_output_levels + stats.bytes_read_blob;
+  const uint64_t bytes_read_all =
+      stats.bytes_read_output_level + bytes_read_non_output_and_blob;
+  const uint64_t bytes_written_all =
+      stats.bytes_written + stats.bytes_written_blob;
+
+  if (bytes_read_non_output_and_blob > 0) {
+    read_write_amp = (bytes_written_all + bytes_read_all) /
+                     static_cast<double>(bytes_read_non_output_and_blob);
+    write_amp =
+        bytes_written_all / static_cast<double>(bytes_read_non_output_and_blob);
   }
   if (stats.micros > 0) {
-    bytes_read_per_sec =
-        (stats.bytes_read_non_output_levels + stats.bytes_read_output_level) /
-        static_cast<double>(stats.micros);
+    bytes_read_per_sec = bytes_read_all / static_cast<double>(stats.micros);
     bytes_written_per_sec =
-        stats.bytes_written / static_cast<double>(stats.micros);
+        bytes_written_all / static_cast<double>(stats.micros);
   }
 
   const std::string& column_family_name = cfd->GetName();
 
+  constexpr double kMB = 1048576.0;
+
   ROCKS_LOG_BUFFER(
       log_buffer_,
       "[%s] compacted to: %s, MB/sec: %.1f rd, %.1f wr, level %d, "
-      "files in(%d, %d) out(%d) "
-      "MB in(%.1f, %.1f) out(%.1f), read-write-amplify(%.1f) "
-      "write-amplify(%.1f) %s, records in: %" PRIu64
+      "files in(%d, %d) out(%d +%d blob) "
+      "MB in(%.1f, %.1f +%.1f blob) out(%.1f +%.1f blob), "
+      "read-write-amplify(%.1f) write-amplify(%.1f) %s, records in: %" PRIu64
       ", records dropped: %" PRIu64 " output_compression: %s\n",
       column_family_name.c_str(), vstorage->LevelSummary(&tmp),
       bytes_read_per_sec, bytes_written_per_sec,
       compact_->compaction->output_level(),
       stats.num_input_files_in_non_output_levels,
       stats.num_input_files_in_output_level, stats.num_output_files,
-      stats.bytes_read_non_output_levels / 1048576.0,
-      stats.bytes_read_output_level / 1048576.0,
-      stats.bytes_written / 1048576.0, read_write_amp, write_amp,
-      status.ToString().c_str(), stats.num_input_records,
+      stats.num_output_files_blob, stats.bytes_read_non_output_levels / kMB,
+      stats.bytes_read_output_level / kMB, stats.bytes_read_blob / kMB,
+      stats.bytes_written / kMB, stats.bytes_written_blob / kMB, read_write_amp,
+      write_amp, status.ToString().c_str(), stats.num_input_records,
       stats.num_dropped_records,
       CompressionTypeToString(compact_->compaction->output_compression())
           .c_str());
@@ -894,7 +903,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
   assert(sub_compact);
   assert(sub_compact->compaction);
 
-  uint64_t prev_cpu_micros = env_->NowCPUNanos() / 1000;
+  uint64_t prev_cpu_micros = db_options_.clock->CPUNanos() / 1000;
 
   ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
 
@@ -971,12 +980,13 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
 
   std::unique_ptr<BlobFileBuilder> blob_file_builder(
       mutable_cf_options->enable_blob_files
-          ? new BlobFileBuilder(
-                versions_, env_, fs_.get(),
-                sub_compact->compaction->immutable_cf_options(),
-                mutable_cf_options, &file_options_, job_id_, cfd->GetID(),
-                cfd->GetName(), Env::IOPriority::IO_LOW, write_hint_,
-                &blob_file_paths, &sub_compact->blob_file_additions)
+          ? new BlobFileBuilder(versions_, fs_.get(),
+                                sub_compact->compaction->immutable_cf_options(),
+                                mutable_cf_options, &file_options_, job_id_,
+                                cfd->GetID(), cfd->GetName(),
+                                Env::IOPriority::IO_LOW, write_hint_,
+                                io_tracer_, blob_callback_, &blob_file_paths,
+                                &sub_compact->blob_file_additions)
           : nullptr);
 
   TEST_SYNC_POINT("CompactionJob::Run():Inprogress");
@@ -1119,6 +1129,10 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
     }
   }
 
+  sub_compact->compaction_job_stats.num_blobs_read =
+      c_iter_stats.num_blobs_read;
+  sub_compact->compaction_job_stats.total_blob_bytes_read =
+      c_iter_stats.total_blob_bytes_read;
   sub_compact->compaction_job_stats.num_input_deletion_records =
       c_iter_stats.num_input_deletion_records;
   sub_compact->compaction_job_stats.num_corrupt_keys =
@@ -1178,13 +1192,14 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
   if (blob_file_builder) {
     if (status.ok()) {
       status = blob_file_builder->Finish();
+    } else {
+      blob_file_builder->Abandon();
     }
-
     blob_file_builder.reset();
   }
 
   sub_compact->compaction_job_stats.cpu_micros =
-      env_->NowCPUNanos() / 1000 - prev_cpu_micros;
+      db_options_.clock->CPUNanos() / 1000 - prev_cpu_micros;
 
   if (measure_io_stats_) {
     sub_compact->compaction_job_stats.file_write_nanos +=
@@ -1463,7 +1478,7 @@ Status CompactionJob::FinishCompactionOutputFile(
 
   // Finish and check for file errors
   if (s.ok()) {
-    StopWatch sw(env_, stats_, COMPACTION_OUTFILE_SYNC_MICROS);
+    StopWatch sw(db_options_.clock, stats_, COMPACTION_OUTFILE_SYNC_MICROS);
     io_s = sub_compact->outfile->Sync(db_options_.use_fsync);
   }
   if (s.ok() && io_s.ok()) {
@@ -1702,7 +1717,7 @@ Status CompactionJob::OpenCompactionOutputFile(
 
   // Try to figure out the output file's oldest ancester time.
   int64_t temp_current_time = 0;
-  auto get_time_status = env_->GetCurrentTime(&temp_current_time);
+  auto get_time_status = db_options_.clock->GetCurrentTime(&temp_current_time);
   // Safe to proceed even if GetCurrentTime fails. So, log and proceed.
   if (!get_time_status.ok()) {
     ROCKS_LOG_WARN(db_options_.info_log,
@@ -1733,14 +1748,16 @@ Status CompactionJob::OpenCompactionOutputFile(
 
   writable_file->SetIOPriority(Env::IOPriority::IO_LOW);
   writable_file->SetWriteLifeTimeHint(write_hint_);
+  FileTypeSet tmp_set = db_options_.checksum_handoff_file_types;
   writable_file->SetPreallocationBlockSize(static_cast<size_t>(
       sub_compact->compaction->OutputFilePreallocationSize()));
   const auto& listeners =
       sub_compact->compaction->immutable_cf_options()->listeners;
   sub_compact->outfile.reset(new WritableFileWriter(
-      std::move(writable_file), fname, file_options_, env_, io_tracer_,
-      db_options_.statistics.get(), listeners,
-      db_options_.file_checksum_gen_factory.get()));
+      std::move(writable_file), fname, file_options_, db_options_.clock,
+      io_tracer_, db_options_.statistics.get(), listeners,
+      db_options_.file_checksum_gen_factory.get(),
+      tmp_set.Contains(FileType::kTableFile)));
 
   // If the Column family flag is to only optimize filters for hits,
   // we can skip creating filters if this is the bottommost_level where
@@ -1820,11 +1837,16 @@ void CompactionJob::UpdateCompactionStats() {
     }
   }
 
+  assert(compaction_job_stats_);
+  compaction_stats_.bytes_read_blob =
+      compaction_job_stats_->total_blob_bytes_read;
+
   compaction_stats_.num_output_files =
-      static_cast<int>(compact_->num_output_files) +
+      static_cast<int>(compact_->num_output_files);
+  compaction_stats_.num_output_files_blob =
       static_cast<int>(compact_->num_blob_output_files);
-  compaction_stats_.bytes_written =
-      compact_->total_bytes + compact_->total_blob_bytes;
+  compaction_stats_.bytes_written = compact_->total_bytes;
+  compaction_stats_.bytes_written_blob = compact_->total_blob_bytes;
 
   if (compaction_stats_.num_input_records > compact_->num_output_records) {
     compaction_stats_.num_dropped_records =
@@ -1864,8 +1886,10 @@ void CompactionJob::UpdateCompactionJobStats(
 
   // output information
   compaction_job_stats_->total_output_bytes = stats.bytes_written;
+  compaction_job_stats_->total_output_bytes_blob = stats.bytes_written_blob;
   compaction_job_stats_->num_output_records = compact_->num_output_records;
   compaction_job_stats_->num_output_files = stats.num_output_files;
+  compaction_job_stats_->num_output_files_blob = stats.num_output_files_blob;
 
   if (stats.num_output_files > 0) {
     CopyPrefix(compact_->SmallestUserKey(),

db/compaction/compaction_job.h

@@ -17,6 +17,7 @@
 #include <utility>
 #include <vector>
 
+#include "db/blob/blob_file_completion_callback.h"
 #include "db/column_family.h"
 #include "db/compaction/compaction_iterator.h"
 #include "db/dbformat.h"
@@ -50,6 +51,7 @@ class Arena;
 class ErrorHandler;
 class MemTable;
 class SnapshotChecker;
+class SystemClock;
 class TableCache;
 class Version;
 class VersionEdit;
@@ -79,7 +81,8 @@ class CompactionJob {
       Env::Priority thread_pri, const std::shared_ptr<IOTracer>& io_tracer,
       const std::atomic<int>* manual_compaction_paused = nullptr,
       const std::string& db_id = "", const std::string& db_session_id = "",
-      std::string full_history_ts_low = "");
+      std::string full_history_ts_low = "",
+      BlobFileCompletionCallback* blob_callback = nullptr);
 
   ~CompactionJob();
 
@@ -203,6 +206,7 @@ class CompactionJob {
   Env::Priority thread_pri_;
   IOStatus io_status_;
   std::string full_history_ts_low_;
+  BlobFileCompletionCallback* blob_callback_;
 };
 
 }  // namespace ROCKSDB_NAMESPACE

db/compaction/compaction_job_test.cc

@@ -5,6 +5,8 @@
 
 #ifndef ROCKSDB_LITE
 
+#include "db/compaction/compaction_job.h"
+
 #include <algorithm>
 #include <array>
 #include <cinttypes>
@@ -14,13 +16,13 @@
 
 #include "db/blob/blob_index.h"
 #include "db/column_family.h"
-#include "db/compaction/compaction_job.h"
 #include "db/db_impl/db_impl.h"
 #include "db/error_handler.h"
 #include "db/version_set.h"
 #include "file/writable_file_writer.h"
 #include "rocksdb/cache.h"
 #include "rocksdb/db.h"
+#include "rocksdb/file_system.h"
 #include "rocksdb/options.h"
 #include "rocksdb/write_buffer_manager.h"
 #include "table/mock_table.h"
@@ -277,12 +279,13 @@ class CompactionJobTestBase : public testing::Test {
     new_db.SetLastSequence(0);
 
     const std::string manifest = DescriptorFileName(dbname_, 1);
-    std::unique_ptr<WritableFile> file;
-    Status s = env_->NewWritableFile(
-        manifest, &file, env_->OptimizeForManifestWrite(env_options_));
+    std::unique_ptr<WritableFileWriter> file_writer;
+    const auto& fs = env_->GetFileSystem();
+    Status s = WritableFileWriter::Create(
+        fs, manifest, fs->OptimizeForManifestWrite(env_options_), &file_writer,
+        nullptr);
+
     ASSERT_OK(s);
-    std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-        NewLegacyWritableFileWrapper(std::move(file)), manifest, env_options_));
     {
       log::Writer log(std::move(file_writer), 0, false);
       std::string record;

db/compaction/compaction_picker_fifo.cc

@@ -45,7 +45,7 @@ Compaction* FIFOCompactionPicker::PickTTLCompaction(
   uint64_t total_size = GetTotalFilesSize(level_files);
 
   int64_t _current_time;
-  auto status = ioptions_.env->GetCurrentTime(&_current_time);
+  auto status = ioptions_.clock->GetCurrentTime(&_current_time);
   if (!status.ok()) {
     ROCKS_LOG_BUFFER(log_buffer,
                      "[%s] FIFO compaction: Couldn't get current time: %s. "

db/corruption_test.cc

@@ -19,7 +19,6 @@
 #include "db/db_test_util.h"
 #include "db/log_format.h"
 #include "db/version_set.h"
-#include "env/composite_env_wrapper.h"
 #include "file/filename.h"
 #include "port/stack_trace.h"
 #include "rocksdb/cache.h"
@@ -539,14 +538,15 @@ TEST_F(CorruptionTest, RangeDeletionCorrupted) {
   ASSERT_EQ(static_cast<size_t>(1), metadata.size());
   std::string filename = dbname_ + metadata[0].name;
 
-  std::unique_ptr<RandomAccessFile> file;
-  ASSERT_OK(options_.env->NewRandomAccessFile(filename, &file, EnvOptions()));
-  std::unique_ptr<RandomAccessFileReader> file_reader(
-      new RandomAccessFileReader(NewLegacyRandomAccessFileWrapper(file),
-                                 filename));
+  FileOptions file_opts;
+  const auto& fs = options_.env->GetFileSystem();
+  std::unique_ptr<RandomAccessFileReader> file_reader;
+  ASSERT_OK(RandomAccessFileReader::Create(fs, filename, file_opts,
+                                           &file_reader, nullptr));
 
   uint64_t file_size;
-  ASSERT_OK(options_.env->GetFileSize(filename, &file_size));
+  ASSERT_OK(
+      fs->GetFileSize(filename, file_opts.io_options, &file_size, nullptr));
 
   BlockHandle range_del_handle;
   ASSERT_OK(FindMetaBlock(
@@ -882,7 +882,7 @@ TEST_F(CorruptionTest, VerifyWholeTableChecksum) {
   SyncPoint::GetInstance()->ClearAllCallBacks();
   int count{0};
   SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::VerifySstFileChecksum:mismatch", [&](void* arg) {
+      "DBImpl::VerifyFullFileChecksum:mismatch", [&](void* arg) {
         auto* s = reinterpret_cast<Status*>(arg);
         ASSERT_NE(s, nullptr);
         ++count;

db/db_bloom_filter_test.cc

@@ -7,6 +7,9 @@
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 
+#include <iomanip>
+#include <sstream>
+
 #include "db/db_test_util.h"
 #include "options/options_helper.h"
 #include "port/stack_trace.h"
@@ -2117,6 +2120,54 @@ TEST_F(DBBloomFilterTest, DynamicBloomFilterOptions) {
   }
 }
 
+TEST_F(DBBloomFilterTest, SeekForPrevWithPartitionedFilters) {
+  Options options = CurrentOptions();
+  constexpr size_t kNumKeys = 10000;
+  static_assert(kNumKeys <= 10000, "kNumKeys have to be <= 10000");
+  options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeys + 10));
+  options.create_if_missing = true;
+  constexpr size_t kPrefixLength = 4;
+  options.prefix_extractor.reset(NewFixedPrefixTransform(kPrefixLength));
+  options.compression = kNoCompression;
+  BlockBasedTableOptions bbto;
+  bbto.filter_policy.reset(NewBloomFilterPolicy(50));
+  bbto.index_shortening =
+      BlockBasedTableOptions::IndexShorteningMode::kNoShortening;
+  bbto.block_size = 128;
+  bbto.metadata_block_size = 128;
+  bbto.partition_filters = true;
+  bbto.index_type = BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch;
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  DestroyAndReopen(options);
+
+  const std::string value(64, '\0');
+
+  WriteOptions write_opts;
+  write_opts.disableWAL = true;
+  for (size_t i = 0; i < kNumKeys; ++i) {
+    std::ostringstream oss;
+    oss << std::setfill('0') << std::setw(4) << std::fixed << i;
+    ASSERT_OK(db_->Put(write_opts, oss.str(), value));
+  }
+  ASSERT_OK(Flush());
+
+  ReadOptions read_opts;
+  // Use legacy, implicit prefix seek
+  read_opts.total_order_seek = false;
+  read_opts.auto_prefix_mode = false;
+  std::unique_ptr<Iterator> it(db_->NewIterator(read_opts));
+  for (size_t i = 0; i < kNumKeys; ++i) {
+    // Seek with a key after each one added but with same prefix. One will
+    // surely cross a partition boundary.
+    std::ostringstream oss;
+    oss << std::setfill('0') << std::setw(4) << std::fixed << i << "a";
+    it->SeekForPrev(oss.str());
+    ASSERT_OK(it->status());
+    ASSERT_TRUE(it->Valid());
+  }
+  it.reset();
+}
+
 #endif  // ROCKSDB_LITE
 
 }  // namespace ROCKSDB_NAMESPACE

db/db_compaction_test.cc

@@ -11,6 +11,7 @@
 
 #include "db/blob/blob_index.h"
 #include "db/db_test_util.h"
+#include "env/mock_env.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
 #include "rocksdb/concurrent_task_limiter.h"
@@ -21,6 +22,7 @@
 #include "util/concurrent_task_limiter_impl.h"
 #include "util/random.h"
 #include "utilities/fault_injection_env.h"
+#include "utilities/fault_injection_fs.h"
 
 namespace ROCKSDB_NAMESPACE {
 
@@ -31,31 +33,6 @@ class DBCompactionTest : public DBTestBase {
  public:
   DBCompactionTest()
       : DBTestBase("/db_compaction_test", /*env_do_fsync=*/true) {}
-
-  std::vector<uint64_t> GetBlobFileNumbers() {
-    VersionSet* const versions = dbfull()->TEST_GetVersionSet();
-    assert(versions);
-
-    ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
-    assert(cfd);
-
-    Version* const current = cfd->current();
-    assert(current);
-
-    const VersionStorageInfo* const storage_info = current->storage_info();
-    assert(storage_info);
-
-    const auto& blob_files = storage_info->GetBlobFiles();
-
-    std::vector<uint64_t> result;
-    result.reserve(blob_files.size());
-
-    for (const auto& blob_file : blob_files) {
-      result.emplace_back(blob_file.first);
-    }
-
-    return result;
-  }
 };
 
 class DBCompactionTestWithParam
@@ -204,6 +181,7 @@ Options DeletionTriggerOptions(Options options) {
       options.target_file_size_base * options.target_file_size_multiplier;
   options.max_bytes_for_level_multiplier = 2;
   options.disable_auto_compactions = false;
+  options.compaction_options_universal.max_size_amplification_percent = 100;
   return options;
 }
 
@@ -358,12 +336,34 @@ TEST_P(DBCompactionTestWithParam, CompactionDeletionTrigger) {
     for (int k = 0; k < kTestSize; ++k) {
       ASSERT_OK(Delete(Key(k)));
     }
-    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+    ASSERT_OK(Flush());
     ASSERT_OK(dbfull()->TEST_WaitForCompact());
     ASSERT_OK(Size(Key(0), Key(kTestSize - 1), &db_size[1]));
 
-    // must have much smaller db size.
-    ASSERT_GT(db_size[0] / 3, db_size[1]);
+    if (options.compaction_style == kCompactionStyleUniversal) {
+      // Claim: in universal compaction none of the original data will remain
+      // once compactions settle.
+      //
+      // Proof: The compensated size of the file containing the most tombstones
+      // is enough on its own to trigger size amp compaction. Size amp
+      // compaction is a full compaction, so all tombstones meet the obsolete
+      // keys they cover.
+      ASSERT_EQ(0, db_size[1]);
+    } else {
+      // Claim: in level compaction at most `db_size[0] / 2` of the original
+      // data will remain once compactions settle.
+      //
+      // Proof: Assume the original data is all in the bottom level. If it were
+      // not, it would meet its tombstone sooner. The original data size is
+      // large enough to require fanout to bottom level to be greater than
+      // `max_bytes_for_level_multiplier == 2`. In the level just above,
+      // tombstones must cover less than `db_size[0] / 4` bytes since fanout >=
+      // 2 and file size is compensated by doubling the size of values we expect
+      // are covered (`kDeletionWeightOnCompaction == 2`). The tombstones in
+      // levels above must cover less than `db_size[0] / 8` bytes of original
+      // data, `db_size[0] / 16`, and so on.
+      ASSERT_GT(db_size[0] / 2, db_size[1]);
+    }
   }
 }
 #endif  // ROCKSDB_VALGRIND_RUN
@@ -630,9 +630,8 @@ TEST_P(DBCompactionTestWithParam, CompactionDeletionTriggerReopen) {
     }
     ASSERT_OK(Size(Key(0), Key(kTestSize - 1), &db_size[1]));
     Close();
-    // as auto_compaction is off, we shouldn't see too much reduce
-    // in db size.
-    ASSERT_LT(db_size[0] / 3, db_size[1]);
+    // as auto_compaction is off, we shouldn't see any reduction in db size.
+    ASSERT_LE(db_size[0], db_size[1]);
 
     // round 3 --- reopen db with auto_compaction on and see if
     // deletion compensation still work.
@@ -646,7 +645,13 @@ TEST_P(DBCompactionTestWithParam, CompactionDeletionTriggerReopen) {
     ASSERT_OK(dbfull()->TEST_WaitForCompact());
     ASSERT_OK(Size(Key(0), Key(kTestSize - 1), &db_size[2]));
     // this time we're expecting significant drop in size.
-    ASSERT_GT(db_size[0] / 3, db_size[2]);
+    //
+    // See "CompactionDeletionTrigger" test for proof that at most
+    // `db_size[0] / 2` of the original data remains. In addition to that, this
+    // test inserts `db_size[0] / 10` to push the tombstones into SST files and
+    // then through automatic compactions. So in total `3 * db_size[0] / 5` of
+    // the original data may remain.
+    ASSERT_GT(3 * db_size[0] / 5, db_size[2]);
   }
 }
 
@@ -733,8 +738,15 @@ TEST_F(DBCompactionTest, DisableStatsUpdateReopen) {
       values.push_back(rnd.RandomString(kCDTValueSize));
       ASSERT_OK(Put(Key(k), values[k]));
     }
-    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+    ASSERT_OK(Flush());
     ASSERT_OK(dbfull()->TEST_WaitForCompact());
+    // L1 and L2 can fit deletions iff size compensation does not take effect,
+    // i.e., when `skip_stats_update_on_db_open == true`. Move any remaining
+    // files at or above L2 down to L3 to ensure obsolete data does not
+    // accidentally meet its tombstone above L3. This makes the final size more
+    // deterministic and easy to see whether size compensation for deletions
+    // took effect.
+    MoveFilesToLevel(3 /* level */);
     ASSERT_OK(Size(Key(0), Key(kTestSize - 1), &db_size[0]));
     Close();
 
@@ -750,9 +762,8 @@ TEST_F(DBCompactionTest, DisableStatsUpdateReopen) {
     }
     ASSERT_OK(Size(Key(0), Key(kTestSize - 1), &db_size[1]));
     Close();
-    // as auto_compaction is off, we shouldn't see too much reduce
-    // in db size.
-    ASSERT_LT(db_size[0] / 3, db_size[1]);
+    // as auto_compaction is off, we shouldn't see any reduction in db size.
+    ASSERT_LE(db_size[0], db_size[1]);
 
     // round 3 --- reopen db with auto_compaction on and see if
     // deletion compensation still work.
@@ -765,10 +776,17 @@ TEST_F(DBCompactionTest, DisableStatsUpdateReopen) {
     if (options.skip_stats_update_on_db_open) {
       // If update stats on DB::Open is disable, we don't expect
       // deletion entries taking effect.
-      ASSERT_LT(db_size[0] / 3, db_size[2]);
+      //
+      // The deletions are small enough to fit in L1 and L2, and obsolete keys
+      // were moved to L3+, so none of the original data should have been
+      // dropped.
+      ASSERT_LE(db_size[0], db_size[2]);
     } else {
       // Otherwise, we should see a significant drop in db size.
-      ASSERT_GT(db_size[0] / 3, db_size[2]);
+      //
+      // See "CompactionDeletionTrigger" test for proof that at most
+      // `db_size[0] / 2` of the original data remains.
+      ASSERT_GT(db_size[0] / 2, db_size[2]);
     }
   }
 }
@@ -5944,24 +5962,22 @@ TEST_F(DBCompactionTest, CompactionWithBlob) {
   const InternalStats* const internal_stats = cfd->internal_stats();
   ASSERT_NE(internal_stats, nullptr);
 
-  const uint64_t expected_bytes =
-      table_file->fd.GetFileSize() + blob_file->GetTotalBlobBytes();
-
   const auto& compaction_stats = internal_stats->TEST_GetCompactionStats();
   ASSERT_GE(compaction_stats.size(), 2);
-  ASSERT_EQ(compaction_stats[1].bytes_written, expected_bytes);
-  ASSERT_EQ(compaction_stats[1].num_output_files, 2);
+  ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
+  ASSERT_EQ(compaction_stats[1].bytes_written, table_file->fd.GetFileSize());
+  ASSERT_EQ(compaction_stats[1].bytes_written_blob,
+            blob_file->GetTotalBlobBytes());
+  ASSERT_EQ(compaction_stats[1].num_output_files, 1);
+  ASSERT_EQ(compaction_stats[1].num_output_files_blob, 1);
 }
 
 class DBCompactionTestBlobError
     : public DBCompactionTest,
       public testing::WithParamInterface<std::string> {
  public:
-  DBCompactionTestBlobError()
-      : fault_injection_env_(env_), sync_point_(GetParam()) {}
-  ~DBCompactionTestBlobError() { Close(); }
+  DBCompactionTestBlobError() : sync_point_(GetParam()) {}
 
-  FaultInjectionTestEnv fault_injection_env_;
   std::string sync_point_;
 };
 
@@ -5996,13 +6012,14 @@ TEST_P(DBCompactionTestBlobError, CompactionError) {
   ASSERT_OK(Flush());
 
   options.enable_blob_files = true;
-  options.env = &fault_injection_env_;
 
   Reopen(options);
 
-  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
-    fault_injection_env_.SetFilesystemActive(false,
-                                             Status::IOError(sync_point_));
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* arg) {
+    Status* const s = static_cast<Status*>(arg);
+    assert(s);
+
+    (*s) = Status::IOError(sync_point_);
   });
   SyncPoint::GetInstance()->EnableProcessing();
 
@@ -6039,12 +6056,18 @@ TEST_P(DBCompactionTestBlobError, CompactionError) {
   ASSERT_GE(compaction_stats.size(), 2);
 
   if (sync_point_ == "BlobFileBuilder::WriteBlobToFile:AddRecord") {
+    ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
     ASSERT_EQ(compaction_stats[1].bytes_written, 0);
+    ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
     ASSERT_EQ(compaction_stats[1].num_output_files, 0);
+    ASSERT_EQ(compaction_stats[1].num_output_files_blob, 0);
   } else {
     // SST file writing succeeded; blob file writing failed (during Finish)
+    ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
     ASSERT_GT(compaction_stats[1].bytes_written, 0);
+    ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
     ASSERT_EQ(compaction_stats[1].num_output_files, 1);
+    ASSERT_EQ(compaction_stats[1].num_output_files_blob, 0);
   }
 }
 
@@ -6129,6 +6152,36 @@ TEST_P(DBCompactionTestBlobGC, CompactionWithBlobGC) {
   for (size_t i = cutoff_index; i < original_blob_files.size(); ++i) {
     ASSERT_EQ(new_blob_files[i - cutoff_index], original_blob_files[i]);
   }
+
+  VersionSet* const versions = dbfull()->TEST_GetVersionSet();
+  assert(versions);
+  assert(versions->GetColumnFamilySet());
+
+  ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
+  assert(cfd);
+
+  const InternalStats* const internal_stats = cfd->internal_stats();
+  assert(internal_stats);
+
+  const auto& compaction_stats = internal_stats->TEST_GetCompactionStats();
+  ASSERT_GE(compaction_stats.size(), 2);
+
+  if (blob_gc_age_cutoff_ > 0.0) {
+    ASSERT_GT(compaction_stats[1].bytes_read_blob, 0);
+
+    if (updated_enable_blob_files_) {
+      // GC relocated some blobs to new blob files
+      ASSERT_GT(compaction_stats[1].bytes_written_blob, 0);
+      ASSERT_EQ(compaction_stats[1].bytes_read_blob,
+                compaction_stats[1].bytes_written_blob);
+    } else {
+      // GC moved some blobs back to the LSM, no new blob files
+      ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+    }
+  } else {
+    ASSERT_EQ(compaction_stats[1].bytes_read_blob, 0);
+    ASSERT_EQ(compaction_stats[1].bytes_written_blob, 0);
+  }
 }
 
 TEST_F(DBCompactionTest, CompactionWithBlobGCError_CorruptIndex) {
@@ -6275,6 +6328,297 @@ TEST_F(DBCompactionTest, CompactionWithBlobGCError_IndexWithInvalidFileNumber) {
       db_->CompactRange(CompactRangeOptions(), begin, end).IsCorruption());
 }
 
+TEST_F(DBCompactionTest, CompactionWithChecksumHandoff1) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.level0_file_num_compaction_trigger = 2;
+  options.num_levels = 3;
+  options.env = fault_fs_env.get();
+  options.create_if_missing = true;
+  options.checksum_handoff_file_types.Add(FileType::kTableFile);
+  Status s;
+  Reopen(options);
+
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+  Destroy(options);
+  Reopen(options);
+
+  // The hash does not match, compaction write fails
+  // fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+  // Since the file system returns IOStatus::Corruption, it is an
+  // unrecoverable error.
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
+        "BackgroundCallCompaction:0"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BackgroundCallCompaction:0", [&](void*) {
+        fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s.severity(),
+            ROCKSDB_NAMESPACE::Status::Severity::kUnrecoverableError);
+  SyncPoint::GetInstance()->DisableProcessing();
+  Destroy(options);
+  Reopen(options);
+
+  // The file system does not support checksum handoff. The check
+  // will be ignored.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+
+  // Each write will be similated as corrupted.
+  // Since the file system returns IOStatus::Corruption, it is an
+  // unrecoverable error.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
+        "BackgroundCallCompaction:0"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BackgroundCallCompaction:0",
+      [&](void*) { fault_fs->IngestDataCorruptionBeforeWrite(); });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s.severity(),
+            ROCKSDB_NAMESPACE::Status::Severity::kUnrecoverableError);
+  SyncPoint::GetInstance()->DisableProcessing();
+
+  Destroy(options);
+}
+
+TEST_F(DBCompactionTest, CompactionWithChecksumHandoff2) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.level0_file_num_compaction_trigger = 2;
+  options.num_levels = 3;
+  options.env = fault_fs_env.get();
+  options.create_if_missing = true;
+  Status s;
+  Reopen(options);
+
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+  Destroy(options);
+  Reopen(options);
+
+  // options is not set, the checksum handoff will not be triggered
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
+        "BackgroundCallCompaction:0"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BackgroundCallCompaction:0", [&](void*) {
+        fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+  SyncPoint::GetInstance()->DisableProcessing();
+  Destroy(options);
+  Reopen(options);
+
+  // The file system does not support checksum handoff. The check
+  // will be ignored.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+
+  // options is not set, the checksum handoff will not be triggered
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
+        "BackgroundCallCompaction:0"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BackgroundCallCompaction:0",
+      [&](void*) { fault_fs->IngestDataCorruptionBeforeWrite(); });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+
+  Destroy(options);
+}
+
+TEST_F(DBCompactionTest, CompactionWithChecksumHandoffManifest1) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.level0_file_num_compaction_trigger = 2;
+  options.num_levels = 3;
+  options.env = fault_fs_env.get();
+  options.create_if_missing = true;
+  options.checksum_handoff_file_types.Add(FileType::kDescriptorFile);
+  Status s;
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  Reopen(options);
+
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+  Destroy(options);
+  Reopen(options);
+
+  // The hash does not match, compaction write fails
+  // fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+  // Since the file system returns IOStatus::Corruption, it is mapped to
+  // kFatalError error.
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
+        "BackgroundCallCompaction:0"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BackgroundCallCompaction:0", [&](void*) {
+        fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+      });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kFatalError);
+  SyncPoint::GetInstance()->DisableProcessing();
+  Destroy(options);
+}
+
+TEST_F(DBCompactionTest, CompactionWithChecksumHandoffManifest2) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.level0_file_num_compaction_trigger = 2;
+  options.num_levels = 3;
+  options.env = fault_fs_env.get();
+  options.create_if_missing = true;
+  options.checksum_handoff_file_types.Add(FileType::kDescriptorFile);
+  Status s;
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+  Reopen(options);
+
+  // The file system does not support checksum handoff. The check
+  // will be ignored.
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s, Status::OK());
+
+  // Each write will be similated as corrupted.
+  // Since the file system returns IOStatus::Corruption, it is mapped to
+  // kFatalError error.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put(Key(0), "value1"));
+  ASSERT_OK(Put(Key(2), "value2"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::FlushMemTable:FlushMemTableFinished",
+        "BackgroundCallCompaction:0"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BackgroundCallCompaction:0",
+      [&](void*) { fault_fs->IngestDataCorruptionBeforeWrite(); });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Put(Key(1), "value3"));
+  s = Flush();
+  ASSERT_EQ(s, Status::OK());
+  s = dbfull()->TEST_WaitForCompact();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kFatalError);
+  SyncPoint::GetInstance()->DisableProcessing();
+
+  Destroy(options);
+}
+
 #endif  // !defined(ROCKSDB_LITE)
 
 }  // namespace ROCKSDB_NAMESPACE

db/db_dynamic_level_test.cc

@@ -13,6 +13,7 @@
 #if !defined(ROCKSDB_LITE)
 
 #include "db/db_test_util.h"
+#include "env/mock_env.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
 #include "util/random.h"

db/db_flush_test.cc

@@ -11,6 +11,7 @@
 
 #include "db/db_impl/db_impl.h"
 #include "db/db_test_util.h"
+#include "env/mock_env.h"
 #include "file/filename.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
@@ -19,6 +20,7 @@
 #include "util/cast_util.h"
 #include "util/mutexlock.h"
 #include "utilities/fault_injection_env.h"
+#include "utilities/fault_injection_fs.h"
 
 namespace ROCKSDB_NAMESPACE {
 
@@ -80,29 +82,16 @@ TEST_F(DBFlushTest, SyncFail) {
   options.env = fault_injection_env.get();
 
   SyncPoint::GetInstance()->LoadDependency(
-      {{"DBFlushTest::SyncFail:GetVersionRefCount:1",
-        "DBImpl::FlushMemTableToOutputFile:BeforePickMemtables"},
-       {"DBImpl::FlushMemTableToOutputFile:AfterPickMemtables",
-        "DBFlushTest::SyncFail:GetVersionRefCount:2"},
-       {"DBFlushTest::SyncFail:1", "DBImpl::SyncClosedLogs:Start"},
+      {{"DBFlushTest::SyncFail:1", "DBImpl::SyncClosedLogs:Start"},
        {"DBImpl::SyncClosedLogs:Failed", "DBFlushTest::SyncFail:2"}});
   SyncPoint::GetInstance()->EnableProcessing();
 
   CreateAndReopenWithCF({"pikachu"}, options);
   ASSERT_OK(Put("key", "value"));
-  auto* cfd =
-      static_cast_with_check<ColumnFamilyHandleImpl>(db_->DefaultColumnFamily())
-          ->cfd();
   FlushOptions flush_options;
   flush_options.wait = false;
   ASSERT_OK(dbfull()->Flush(flush_options));
   // Flush installs a new super-version. Get the ref count after that.
-  auto current_before = cfd->current();
-  int refs_before = cfd->current()->TEST_refs();
-  TEST_SYNC_POINT("DBFlushTest::SyncFail:GetVersionRefCount:1");
-  TEST_SYNC_POINT("DBFlushTest::SyncFail:GetVersionRefCount:2");
-  int refs_after_picking_memtables = cfd->current()->TEST_refs();
-  ASSERT_EQ(refs_before + 1, refs_after_picking_memtables);
   fault_injection_env->SetFilesystemActive(false);
   TEST_SYNC_POINT("DBFlushTest::SyncFail:1");
   TEST_SYNC_POINT("DBFlushTest::SyncFail:2");
@@ -113,9 +102,6 @@ TEST_F(DBFlushTest, SyncFail) {
 #ifndef ROCKSDB_LITE
   ASSERT_EQ("", FilesPerLevel());  // flush failed.
 #endif                             // ROCKSDB_LITE
-  // Backgroun flush job should release ref count to current version.
-  ASSERT_EQ(current_before, cfd->current());
-  ASSERT_EQ(refs_before, cfd->current()->TEST_refs());
   Destroy(options);
 }
 
@@ -451,7 +437,6 @@ TEST_F(DBFlushTest, FlushWithBlob) {
   constexpr uint64_t min_blob_size = 10;
 
   Options options;
-  options.env = CurrentOptions().env;
   options.enable_blob_files = true;
   options.min_blob_size = min_blob_size;
   options.disable_auto_compactions = true;
@@ -512,27 +497,224 @@ TEST_F(DBFlushTest, FlushWithBlob) {
   const InternalStats* const internal_stats = cfd->internal_stats();
   assert(internal_stats);
 
-  const uint64_t expected_bytes =
-      table_file->fd.GetFileSize() + blob_file->GetTotalBlobBytes();
-
   const auto& compaction_stats = internal_stats->TEST_GetCompactionStats();
   ASSERT_FALSE(compaction_stats.empty());
-  ASSERT_EQ(compaction_stats[0].bytes_written, expected_bytes);
-  ASSERT_EQ(compaction_stats[0].num_output_files, 2);
+  ASSERT_EQ(compaction_stats[0].bytes_written, table_file->fd.GetFileSize());
+  ASSERT_EQ(compaction_stats[0].bytes_written_blob,
+            blob_file->GetTotalBlobBytes());
+  ASSERT_EQ(compaction_stats[0].num_output_files, 1);
+  ASSERT_EQ(compaction_stats[0].num_output_files_blob, 1);
 
   const uint64_t* const cf_stats_value = internal_stats->TEST_GetCFStatsValue();
-  ASSERT_EQ(cf_stats_value[InternalStats::BYTES_FLUSHED], expected_bytes);
+  ASSERT_EQ(cf_stats_value[InternalStats::BYTES_FLUSHED],
+            compaction_stats[0].bytes_written +
+                compaction_stats[0].bytes_written_blob);
 #endif  // ROCKSDB_LITE
 }
 
+TEST_F(DBFlushTest, FlushWithChecksumHandoff1) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.write_buffer_size = 100;
+  options.max_write_buffer_number = 4;
+  options.min_write_buffer_number_to_merge = 3;
+  options.disable_auto_compactions = true;
+  options.env = fault_fs_env.get();
+  options.checksum_handoff_file_types.Add(FileType::kTableFile);
+  Reopen(options);
+
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put("key1", "value1"));
+  ASSERT_OK(Put("key2", "value2"));
+  ASSERT_OK(dbfull()->TEST_SwitchMemtable());
+
+  // The hash does not match, write fails
+  // fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+  // Since the file system returns IOStatus::Corruption, it is an
+  // unrecoverable error.
+  SyncPoint::GetInstance()->SetCallBack("FlushJob::Start", [&](void*) {
+    fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+  });
+  ASSERT_OK(Put("key3", "value3"));
+  ASSERT_OK(Put("key4", "value4"));
+  SyncPoint::GetInstance()->EnableProcessing();
+  Status s = Flush();
+  ASSERT_EQ(s.severity(),
+            ROCKSDB_NAMESPACE::Status::Severity::kUnrecoverableError);
+  SyncPoint::GetInstance()->DisableProcessing();
+  Destroy(options);
+  Reopen(options);
+
+  // The file system does not support checksum handoff. The check
+  // will be ignored.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+  ASSERT_OK(Put("key5", "value5"));
+  ASSERT_OK(Put("key6", "value6"));
+  ASSERT_OK(dbfull()->TEST_SwitchMemtable());
+
+  // Each write will be similated as corrupted.
+  // Since the file system returns IOStatus::Corruption, it is an
+  // unrecoverable error.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  SyncPoint::GetInstance()->SetCallBack("FlushJob::Start", [&](void*) {
+    fault_fs->IngestDataCorruptionBeforeWrite();
+  });
+  ASSERT_OK(Put("key7", "value7"));
+  ASSERT_OK(Put("key8", "value8"));
+  SyncPoint::GetInstance()->EnableProcessing();
+  s = Flush();
+  ASSERT_EQ(s.severity(),
+            ROCKSDB_NAMESPACE::Status::Severity::kUnrecoverableError);
+  SyncPoint::GetInstance()->DisableProcessing();
+
+  Destroy(options);
+}
+
+TEST_F(DBFlushTest, FlushWithChecksumHandoff2) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.write_buffer_size = 100;
+  options.max_write_buffer_number = 4;
+  options.min_write_buffer_number_to_merge = 3;
+  options.disable_auto_compactions = true;
+  options.env = fault_fs_env.get();
+  Reopen(options);
+
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  ASSERT_OK(Put("key1", "value1"));
+  ASSERT_OK(Put("key2", "value2"));
+  ASSERT_OK(Flush());
+
+  // options is not set, the checksum handoff will not be triggered
+  SyncPoint::GetInstance()->SetCallBack("FlushJob::Start", [&](void*) {
+    fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+  });
+  ASSERT_OK(Put("key3", "value3"));
+  ASSERT_OK(Put("key4", "value4"));
+  SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Flush());
+  SyncPoint::GetInstance()->DisableProcessing();
+  Destroy(options);
+  Reopen(options);
+
+  // The file system does not support checksum handoff. The check
+  // will be ignored.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+  ASSERT_OK(Put("key5", "value5"));
+  ASSERT_OK(Put("key6", "value6"));
+  ASSERT_OK(Flush());
+
+  // options is not set, the checksum handoff will not be triggered
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  SyncPoint::GetInstance()->SetCallBack("FlushJob::Start", [&](void*) {
+    fault_fs->IngestDataCorruptionBeforeWrite();
+  });
+  ASSERT_OK(Put("key7", "value7"));
+  ASSERT_OK(Put("key8", "value8"));
+  SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_OK(Flush());
+  SyncPoint::GetInstance()->DisableProcessing();
+
+  Destroy(options);
+}
+
+TEST_F(DBFlushTest, FlushWithChecksumHandoffManifest1) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.write_buffer_size = 100;
+  options.max_write_buffer_number = 4;
+  options.min_write_buffer_number_to_merge = 3;
+  options.disable_auto_compactions = true;
+  options.env = fault_fs_env.get();
+  options.checksum_handoff_file_types.Add(FileType::kDescriptorFile);
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  Reopen(options);
+
+  ASSERT_OK(Put("key1", "value1"));
+  ASSERT_OK(Put("key2", "value2"));
+  ASSERT_OK(Flush());
+
+  // The hash does not match, write fails
+  // fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+  // Since the file system returns IOStatus::Corruption, it is mapped to
+  // kFatalError error.
+  ASSERT_OK(Put("key3", "value3"));
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionSet::LogAndApply:WriteManifest", [&](void*) {
+        fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+      });
+  ASSERT_OK(Put("key3", "value3"));
+  ASSERT_OK(Put("key4", "value4"));
+  SyncPoint::GetInstance()->EnableProcessing();
+  Status s = Flush();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kFatalError);
+  SyncPoint::GetInstance()->DisableProcessing();
+  Destroy(options);
+}
+
+TEST_F(DBFlushTest, FlushWithChecksumHandoffManifest2) {
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  Options options = CurrentOptions();
+  options.write_buffer_size = 100;
+  options.max_write_buffer_number = 4;
+  options.min_write_buffer_number_to_merge = 3;
+  options.disable_auto_compactions = true;
+  options.env = fault_fs_env.get();
+  options.checksum_handoff_file_types.Add(FileType::kDescriptorFile);
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+  Reopen(options);
+  // The file system does not support checksum handoff. The check
+  // will be ignored.
+  ASSERT_OK(Put("key5", "value5"));
+  ASSERT_OK(Put("key6", "value6"));
+  ASSERT_OK(Flush());
+
+  // Each write will be similated as corrupted.
+  // Since the file system returns IOStatus::Corruption, it is mapped to
+  // kFatalError error.
+  fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+  SyncPoint::GetInstance()->SetCallBack(
+      "VersionSet::LogAndApply:WriteManifest",
+      [&](void*) { fault_fs->IngestDataCorruptionBeforeWrite(); });
+  ASSERT_OK(Put("key7", "value7"));
+  ASSERT_OK(Put("key8", "value8"));
+  SyncPoint::GetInstance()->EnableProcessing();
+  Status s = Flush();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kFatalError);
+  SyncPoint::GetInstance()->DisableProcessing();
+
+  Destroy(options);
+}
+
 class DBFlushTestBlobError : public DBFlushTest,
                              public testing::WithParamInterface<std::string> {
  public:
-  DBFlushTestBlobError()
-      : fault_injection_env_(env_), sync_point_(GetParam()) {}
-  ~DBFlushTestBlobError() { Close(); }
+  DBFlushTestBlobError() : sync_point_(GetParam()) {}
 
-  FaultInjectionTestEnv fault_injection_env_;
   std::string sync_point_;
 };
 
@@ -545,20 +727,18 @@ TEST_P(DBFlushTestBlobError, FlushError) {
   Options options;
   options.enable_blob_files = true;
   options.disable_auto_compactions = true;
-  options.env = &fault_injection_env_;
+  options.env = env_;
 
   Reopen(options);
 
   ASSERT_OK(Put("key", "blob"));
 
-  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
-    fault_injection_env_.SetFilesystemActive(false,
-                                             Status::IOError(sync_point_));
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* arg) {
+    Status* const s = static_cast<Status*>(arg);
+    assert(s);
+
+    (*s) = Status::IOError(sync_point_);
   });
-  SyncPoint::GetInstance()->SetCallBack(
-      "BuildTable:BeforeDeleteFile", [this](void* /* arg */) {
-        fault_injection_env_.SetFilesystemActive(true);
-      });
   SyncPoint::GetInstance()->EnableProcessing();
 
   ASSERT_NOK(Flush());
@@ -608,16 +788,21 @@ TEST_P(DBFlushTestBlobError, FlushError) {
 
   if (sync_point_ == "BlobFileBuilder::WriteBlobToFile:AddRecord") {
     ASSERT_EQ(compaction_stats[0].bytes_written, 0);
+    ASSERT_EQ(compaction_stats[0].bytes_written_blob, 0);
     ASSERT_EQ(compaction_stats[0].num_output_files, 0);
+    ASSERT_EQ(compaction_stats[0].num_output_files_blob, 0);
   } else {
     // SST file writing succeeded; blob file writing failed (during Finish)
     ASSERT_GT(compaction_stats[0].bytes_written, 0);
+    ASSERT_EQ(compaction_stats[0].bytes_written_blob, 0);
     ASSERT_EQ(compaction_stats[0].num_output_files, 1);
+    ASSERT_EQ(compaction_stats[0].num_output_files_blob, 0);
   }
 
   const uint64_t* const cf_stats_value = internal_stats->TEST_GetCFStatsValue();
   ASSERT_EQ(cf_stats_value[InternalStats::BYTES_FLUSHED],
-            compaction_stats[0].bytes_written);
+            compaction_stats[0].bytes_written +
+                compaction_stats[0].bytes_written_blob);
 #endif  // ROCKSDB_LITE
 }
 

db/db_impl/db_impl.cc

@@ -82,6 +82,7 @@
 #include "rocksdb/stats_history.h"
 #include "rocksdb/status.h"
 #include "rocksdb/table.h"
+#include "rocksdb/version.h"
 #include "rocksdb/write_buffer_manager.h"
 #include "table/block_based/block.h"
 #include "table/block_based/block_based_table_factory.h"
@@ -93,7 +94,6 @@
 #include "table/two_level_iterator.h"
 #include "test_util/sync_point.h"
 #include "util/autovector.h"
-#include "util/build_version.h"
 #include "util/cast_util.h"
 #include "util/coding.h"
 #include "util/compression.h"
@@ -156,7 +156,7 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
       fs_(immutable_db_options_.fs, io_tracer_),
       mutable_db_options_(initial_db_options_),
       stats_(immutable_db_options_.statistics.get()),
-      mutex_(stats_, env_, DB_MUTEX_WAIT_MICROS,
+      mutex_(stats_, immutable_db_options_.clock, DB_MUTEX_WAIT_MICROS,
              immutable_db_options_.use_adaptive_mutex),
       default_cf_handle_(nullptr),
       max_total_in_memory_state_(0),
@@ -191,7 +191,7 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
       bg_purge_scheduled_(0),
       disable_delete_obsolete_files_(0),
       pending_purge_obsolete_files_(0),
-      delete_obsolete_files_last_run_(env_->NowMicros()),
+      delete_obsolete_files_last_run_(immutable_db_options_.clock->NowMicros()),
       last_stats_dump_time_microsec_(0),
       next_job_id_(1),
       has_unpersisted_data_(false),
@@ -231,7 +231,9 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
       preserve_deletes_(options.preserve_deletes),
       closed_(false),
       error_handler_(this, immutable_db_options_, &mutex_),
-      atomic_flush_install_cv_(&mutex_) {
+      atomic_flush_install_cv_(&mutex_),
+      blob_callback_(immutable_db_options_.sst_file_manager.get(), &mutex_,
+                     &error_handler_) {
   // !batch_per_trx_ implies seq_per_batch_ because it is only unset for
   // WriteUnprepared, which should use seq_per_batch_.
   assert(batch_per_txn_ || seq_per_batch_);
@@ -511,6 +513,11 @@ Status DBImpl::CloseHelper() {
   }
   mutex_.Unlock();
 
+  // Below check is added as recovery_error_ is not checked and it causes crash
+  // in DBSSTTest.DBWithMaxSpaceAllowedWithBlobFiles when space limit is
+  // reached.
+  error_handler_.GetRecoveryError().PermitUncheckedError();
+
   // CancelAllBackgroundWork called with false means we just set the shutdown
   // marker. After this we do a variant of the waiting and unschedule work
   // (to consider: moving all the waiting into CancelAllBackgroundWork(true))
@@ -752,7 +759,8 @@ void DBImpl::PersistStats() {
     return;
   }
   TEST_SYNC_POINT("DBImpl::PersistStats:StartRunning");
-  uint64_t now_seconds = env_->NowMicros() / kMicrosInSecond;
+  uint64_t now_seconds =
+      immutable_db_options_.clock->NowMicros() / kMicrosInSecond;
 
   Statistics* statistics = immutable_db_options_.statistics.get();
   if (!statistics) {
@@ -1653,8 +1661,8 @@ Status DBImpl::GetImpl(const ReadOptions& read_options, const Slice& key,
   }
 #endif  // NDEBUG
 
-  PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
-  StopWatch sw(env_, stats_, DB_GET);
+  PERF_CPU_TIMER_GUARD(get_cpu_nanos, immutable_db_options_.clock);
+  StopWatch sw(immutable_db_options_.clock, stats_, DB_GET);
   PERF_TIMER_GUARD(get_snapshot_time);
 
   auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(
@@ -1844,8 +1852,8 @@ std::vector<Status> DBImpl::MultiGet(
     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_CPU_TIMER_GUARD(get_cpu_nanos, immutable_db_options_.clock);
+  StopWatch sw(immutable_db_options_.clock, stats_, DB_MULTIGET);
   PERF_TIMER_GUARD(get_snapshot_time);
 
 #ifndef NDEBUG
@@ -1974,9 +1982,8 @@ std::vector<Status> DBImpl::MultiGet(
         break;
       }
     }
-
     if (read_options.deadline.count() &&
-        env_->NowMicros() >
+        immutable_db_options_.clock->NowMicros() >
             static_cast<uint64_t>(read_options.deadline.count())) {
       break;
     }
@@ -1985,8 +1992,8 @@ std::vector<Status> DBImpl::MultiGet(
   if (keys_read < num_keys) {
     // The only reason to break out of the loop is when the deadline is
     // exceeded
-    assert(env_->NowMicros() >
-        static_cast<uint64_t>(read_options.deadline.count()));
+    assert(immutable_db_options_.clock->NowMicros() >
+           static_cast<uint64_t>(read_options.deadline.count()));
     for (++keys_read; keys_read < num_keys; ++keys_read) {
       stat_list[keys_read] = Status::TimedOut();
     }
@@ -2425,8 +2432,8 @@ Status DBImpl::MultiGetImpl(
     autovector<KeyContext*, MultiGetContext::MAX_BATCH_SIZE>* sorted_keys,
     SuperVersion* super_version, SequenceNumber snapshot,
     ReadCallback* callback) {
-  PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
-  StopWatch sw(env_, stats_, DB_MULTIGET);
+  PERF_CPU_TIMER_GUARD(get_cpu_nanos, immutable_db_options_.clock);
+  StopWatch sw(immutable_db_options_.clock, stats_, DB_MULTIGET);
 
   // For each of the given keys, apply the entire "get" process as follows:
   // First look in the memtable, then in the immutable memtable (if any).
@@ -2437,7 +2444,7 @@ Status DBImpl::MultiGetImpl(
   uint64_t curr_value_size = 0;
   while (keys_left) {
     if (read_options.deadline.count() &&
-        env_->NowMicros() >
+        immutable_db_options_.clock->NowMicros() >
             static_cast<uint64_t>(read_options.deadline.count())) {
       s = Status::TimedOut();
       break;
@@ -2990,7 +2997,8 @@ const Snapshot* DBImpl::GetSnapshotForWriteConflictBoundary() {
 SnapshotImpl* DBImpl::GetSnapshotImpl(bool is_write_conflict_boundary,
                                       bool lock) {
   int64_t unix_time = 0;
-  env_->GetCurrentTime(&unix_time).PermitUncheckedError();  // Ignore error
+  immutable_db_options_.clock->GetCurrentTime(&unix_time)
+      .PermitUncheckedError();  // Ignore error
   SnapshotImpl* s = new SnapshotImpl;
 
   if (lock) {
@@ -3135,12 +3143,17 @@ FileSystem* DBImpl::GetFileSystem() const {
   return immutable_db_options_.fs.get();
 }
 
+SystemClock* DBImpl::GetSystemClock() const {
+  return immutable_db_options_.clock;
+}
+
 #ifndef ROCKSDB_LITE
 
-Status DBImpl::StartIOTrace(Env* env, const TraceOptions& trace_options,
+Status DBImpl::StartIOTrace(Env* /*env*/, const TraceOptions& trace_options,
                             std::unique_ptr<TraceWriter>&& trace_writer) {
   assert(trace_writer != nullptr);
-  return io_tracer_->StartIOTrace(env, trace_options, std::move(trace_writer));
+  return io_tracer_->StartIOTrace(GetSystemClock(), trace_options,
+                                  std::move(trace_writer));
 }
 
 Status DBImpl::EndIOTrace() {
@@ -3937,7 +3950,8 @@ Status DestroyDB(const std::string& dbname, const Options& options,
         std::string path_to_delete = dbname + "/" + fname;
         if (type == kMetaDatabase) {
           del = DestroyDB(path_to_delete, options);
-        } else if (type == kTableFile || type == kWalFile) {
+        } else if (type == kTableFile || type == kWalFile ||
+                   type == kBlobFile) {
           del = DeleteDBFile(&soptions, path_to_delete, dbname,
                              /*force_bg=*/false, /*force_fg=*/!wal_in_db_path);
         } else {
@@ -3962,9 +3976,10 @@ Status DestroyDB(const std::string& dbname, const Options& options,
       if (env->GetChildren(path, &filenames).ok()) {
         for (const auto& fname : filenames) {
           if (ParseFileName(fname, &number, &type) &&
-              type == kTableFile) {  // Lock file will be deleted at end
-            std::string table_path = path + "/" + fname;
-            Status del = DeleteDBFile(&soptions, table_path, dbname,
+              (type == kTableFile ||
+               type == kBlobFile)) {  // Lock file will be deleted at end
+            std::string file_path = path + "/" + fname;
+            Status del = DeleteDBFile(&soptions, file_path, dbname,
                                       /*force_bg=*/false, /*force_fg=*/false);
             if (!del.ok() && result.ok()) {
               result = del;
@@ -4208,16 +4223,17 @@ void DBImpl::EraseThreadStatusDbInfo() const {}
 //
 // A global method that can dump out the build version
 void DumpRocksDBBuildVersion(Logger* log) {
-#if !defined(IOS_CROSS_COMPILE)
-  // if we compile with Xcode, we don't run build_detect_version, so we don't
-  // generate util/build_version.cc
-  ROCKS_LOG_HEADER(log, "RocksDB version: %d.%d.%d\n", ROCKSDB_MAJOR,
-                   ROCKSDB_MINOR, ROCKSDB_PATCH);
-  ROCKS_LOG_HEADER(log, "Git sha %s", rocksdb_build_git_sha);
-  ROCKS_LOG_HEADER(log, "Compile date %s", rocksdb_build_compile_date);
-#else
-  (void)log;  // ignore "-Wunused-parameter"
-#endif
+  ROCKS_LOG_HEADER(log, "RocksDB version: %s\n",
+                   GetRocksVersionAsString().c_str());
+  const auto& props = GetRocksBuildProperties();
+  const auto& sha = props.find("rocksdb_build_git_sha");
+  if (sha != props.end()) {
+    ROCKS_LOG_HEADER(log, "Git sha %s", sha->second.c_str());
+  }
+  const auto date = props.find("rocksdb_build_date");
+  if (date != props.end()) {
+    ROCKS_LOG_HEADER(log, "Compile date %s", date->second.c_str());
+  }
 }
 
 #ifndef ROCKSDB_LITE
@@ -4419,9 +4435,9 @@ Status DBImpl::IngestExternalFiles(
   std::vector<ExternalSstFileIngestionJob> ingestion_jobs;
   for (const auto& arg : args) {
     auto* cfd = static_cast<ColumnFamilyHandleImpl*>(arg.column_family)->cfd();
-    ingestion_jobs.emplace_back(
-        env_, versions_.get(), cfd, immutable_db_options_, file_options_,
-        &snapshots_, arg.options, &directories_, &event_logger_, io_tracer_);
+    ingestion_jobs.emplace_back(versions_.get(), cfd, immutable_db_options_,
+                                file_options_, &snapshots_, arg.options,
+                                &directories_, &event_logger_, io_tracer_);
   }
 
   // TODO(yanqin) maybe make jobs run in parallel
@@ -4688,9 +4704,9 @@ Status DBImpl::CreateColumnFamilyWithImport(
   // Import sst files from metadata.
   auto cfh = static_cast_with_check<ColumnFamilyHandleImpl>(*handle);
   auto cfd = cfh->cfd();
-  ImportColumnFamilyJob import_job(env_, versions_.get(), cfd,
-                                   immutable_db_options_, file_options_,
-                                   import_options, metadata.files, io_tracer_);
+  ImportColumnFamilyJob import_job(versions_.get(), cfd, immutable_db_options_,
+                                   file_options_, import_options,
+                                   metadata.files, io_tracer_);
 
   SuperVersionContext dummy_sv_ctx(/* create_superversion */ true);
   VersionEdit dummy_edit;
@@ -4852,17 +4868,37 @@ Status DBImpl::VerifyChecksumInternal(const ReadOptions& read_options,
         std::string fname = TableFileName(cfd->ioptions()->cf_paths,
                                           fd.GetNumber(), fd.GetPathId());
         if (use_file_checksum) {
-          s = VerifySstFileChecksum(*fmeta, fname, read_options);
+          s = VerifyFullFileChecksum(fmeta->file_checksum,
+                                     fmeta->file_checksum_func_name, fname,
+                                     read_options);
         } else {
           s = ROCKSDB_NAMESPACE::VerifySstFileChecksum(opts, file_options_,
                                                        read_options, fname);
         }
       }
     }
+
+    if (s.ok() && use_file_checksum) {
+      const auto& blob_files = vstorage->GetBlobFiles();
+      for (const auto& pair : blob_files) {
+        const uint64_t blob_file_number = pair.first;
+        const auto& meta = pair.second;
+        assert(meta);
+        const std::string blob_file_name = BlobFileName(
+            cfd->ioptions()->cf_paths.front().path, blob_file_number);
+        s = VerifyFullFileChecksum(meta->GetChecksumValue(),
+                                   meta->GetChecksumMethod(), blob_file_name,
+                                   read_options);
+        if (!s.ok()) {
+          break;
+        }
+      }
+    }
     if (!s.ok()) {
       break;
     }
   }
+
   bool defer_purge =
           immutable_db_options().avoid_unnecessary_blocking_io;
   {
@@ -4887,29 +4923,31 @@ Status DBImpl::VerifyChecksumInternal(const ReadOptions& read_options,
   return s;
 }
 
-Status DBImpl::VerifySstFileChecksum(const FileMetaData& fmeta,
-                                     const std::string& fname,
-                                     const ReadOptions& read_options) {
+Status DBImpl::VerifyFullFileChecksum(const std::string& file_checksum_expected,
+                                      const std::string& func_name_expected,
+                                      const std::string& fname,
+                                      const ReadOptions& read_options) {
   Status s;
-  if (fmeta.file_checksum == kUnknownFileChecksum) {
+  if (file_checksum_expected == kUnknownFileChecksum) {
     return s;
   }
   std::string file_checksum;
   std::string func_name;
   s = ROCKSDB_NAMESPACE::GenerateOneFileChecksum(
       fs_.get(), fname, immutable_db_options_.file_checksum_gen_factory.get(),
-      fmeta.file_checksum_func_name, &file_checksum, &func_name,
+      func_name_expected, &file_checksum, &func_name,
       read_options.readahead_size, immutable_db_options_.allow_mmap_reads,
       io_tracer_, immutable_db_options_.rate_limiter.get());
   if (s.ok()) {
-    assert(fmeta.file_checksum_func_name == func_name);
-    if (file_checksum != fmeta.file_checksum) {
+    assert(func_name_expected == func_name);
+    if (file_checksum != file_checksum_expected) {
       std::ostringstream oss;
       oss << fname << " file checksum mismatch, ";
-      oss << "expecting " << Slice(fmeta.file_checksum).ToString(/*hex=*/true);
+      oss << "expecting "
+          << Slice(file_checksum_expected).ToString(/*hex=*/true);
       oss << ", but actual " << Slice(file_checksum).ToString(/*hex=*/true);
       s = Status::Corruption(oss.str());
-      TEST_SYNC_POINT_CALLBACK("DBImpl::VerifySstFileChecksum:mismatch", &s);
+      TEST_SYNC_POINT_CALLBACK("DBImpl::VerifyFullFileChecksum:mismatch", &s);
     }
   }
   return s;
@@ -4944,7 +4982,8 @@ void DBImpl::WaitForIngestFile() {
 Status DBImpl::StartTrace(const TraceOptions& trace_options,
                           std::unique_ptr<TraceWriter>&& trace_writer) {
   InstrumentedMutexLock lock(&trace_mutex_);
-  tracer_.reset(new Tracer(env_, trace_options, std::move(trace_writer)));
+  tracer_.reset(new Tracer(immutable_db_options_.clock, trace_options,
+                           std::move(trace_writer)));
   return Status::OK();
 }
 
@@ -4963,8 +5002,8 @@ Status DBImpl::EndTrace() {
 Status DBImpl::StartBlockCacheTrace(
     const TraceOptions& trace_options,
     std::unique_ptr<TraceWriter>&& trace_writer) {
-  return block_cache_tracer_.StartTrace(env_, trace_options,
-                                        std::move(trace_writer));
+  return block_cache_tracer_.StartTrace(immutable_db_options_.clock,
+                                        trace_options, std::move(trace_writer));
 }
 
 Status DBImpl::EndBlockCacheTrace() {
@@ -4972,24 +5011,27 @@ Status DBImpl::EndBlockCacheTrace() {
   return Status::OK();
 }
 
-Status DBImpl::TraceIteratorSeek(const uint32_t& cf_id, const Slice& key) {
+Status DBImpl::TraceIteratorSeek(const uint32_t& cf_id, const Slice& key,
+                                 const Slice& lower_bound,
+                                 const Slice upper_bound) {
   Status s;
   if (tracer_) {
     InstrumentedMutexLock lock(&trace_mutex_);
     if (tracer_) {
-      s = tracer_->IteratorSeek(cf_id, key);
+      s = tracer_->IteratorSeek(cf_id, key, lower_bound, upper_bound);
     }
   }
   return s;
 }
 
-Status DBImpl::TraceIteratorSeekForPrev(const uint32_t& cf_id,
-                                        const Slice& key) {
+Status DBImpl::TraceIteratorSeekForPrev(const uint32_t& cf_id, const Slice& key,
+                                        const Slice& lower_bound,
+                                        const Slice upper_bound) {
   Status s;
   if (tracer_) {
     InstrumentedMutexLock lock(&trace_mutex_);
     if (tracer_) {
-      s = tracer_->IteratorSeekForPrev(cf_id, key);
+      s = tracer_->IteratorSeekForPrev(cf_id, key, lower_bound, upper_bound);
     }
   }
   return s;

db/db_impl/db_impl.h

@@ -54,9 +54,6 @@
 #include "rocksdb/transaction_log.h"
 #include "rocksdb/write_buffer_manager.h"
 #include "table/scoped_arena_iterator.h"
-#include "trace_replay/block_cache_tracer.h"
-#include "trace_replay/io_tracer.h"
-#include "trace_replay/trace_replay.h"
 #include "util/autovector.h"
 #include "util/hash.h"
 #include "util/repeatable_thread.h"
@@ -450,9 +447,10 @@ class DBImpl : public DB {
   Status VerifyChecksumInternal(const ReadOptions& read_options,
                                 bool use_file_checksum);
 
-  Status VerifySstFileChecksum(const FileMetaData& fmeta,
-                               const std::string& fpath,
-                               const ReadOptions& read_options);
+  Status VerifyFullFileChecksum(const std::string& file_checksum_expected,
+                                const std::string& func_name_expected,
+                                const std::string& fpath,
+                                const ReadOptions& read_options);
 
   using DB::StartTrace;
   virtual Status StartTrace(
@@ -488,6 +486,7 @@ class DBImpl : public DB {
 #endif  // ROCKSDB_LITE
 
   // ---- End of implementations of the DB interface ----
+  SystemClock* GetSystemClock() const;
 
   struct GetImplOptions {
     ColumnFamilyHandle* column_family = nullptr;
@@ -595,8 +594,11 @@ class DBImpl : public DB {
                                  bool* found_record_for_key,
                                  bool* is_blob_index = nullptr);
 
-  Status TraceIteratorSeek(const uint32_t& cf_id, const Slice& key);
-  Status TraceIteratorSeekForPrev(const uint32_t& cf_id, const Slice& key);
+  Status TraceIteratorSeek(const uint32_t& cf_id, const Slice& key,
+                           const Slice& lower_bound, const Slice upper_bound);
+  Status TraceIteratorSeekForPrev(const uint32_t& cf_id, const Slice& key,
+                                  const Slice& lower_bound,
+                                  const Slice upper_bound);
 #endif  // ROCKSDB_LITE
 
   // Similar to GetSnapshot(), but also lets the db know that this snapshot
@@ -1873,6 +1875,8 @@ class DBImpl : public DB {
 
   Status DisableFileDeletionsWithLock();
 
+  Status IncreaseFullHistoryTsLow(ColumnFamilyData* cfd, std::string ts_low);
+
   // table_cache_ provides its own synchronization
   std::shared_ptr<Cache> table_cache_;
 
@@ -2215,6 +2219,8 @@ class DBImpl : public DB {
   InstrumentedCondVar atomic_flush_install_cv_;
 
   bool wal_in_db_path_;
+
+  BlobFileCompletionCallback blob_callback_;
 };
 
 extern Options SanitizeOptions(const std::string& db, const Options& src);

db/db_impl/db_impl_compaction_flush.cc

@@ -131,16 +131,11 @@ IOStatus DBImpl::SyncClosedLogs(JobContext* job_context) {
       MarkLogsNotSynced(current_log_number - 1);
     }
     if (!io_s.ok()) {
-      if (total_log_size_ > 0) {
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
-      } else {
-        // If the WAL is empty, we use different error reason
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
-      }
       TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Failed");
       return io_s;
     }
   }
+  TEST_SYNC_POINT("DBImpl::SyncClosedLogs:end");
   return io_s;
 }
 
@@ -166,20 +161,18 @@ 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,
-      io_tracer_, db_id_, db_session_id_, cfd->GetFullHistoryTsLow());
+      io_tracer_, db_id_, db_session_id_, cfd->GetFullHistoryTsLow(),
+      &blob_callback_);
   FileMetaData file_meta;
 
-  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:BeforePickMemtables");
-  flush_job.PickMemTable();
-  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:AfterPickMemtables");
-
 #ifndef ROCKSDB_LITE
   // may temporarily unlock and lock the mutex.
   NotifyOnFlushBegin(cfd, &file_meta, mutable_cf_options, job_context->job_id);
 #endif  // ROCKSDB_LITE
 
   Status s;
-  IOStatus io_s = IOStatus::OK();
+  bool need_cancel = false;
+  IOStatus log_io_s = IOStatus::OK();
   if (logfile_number_ > 0 &&
       versions_->GetColumnFamilySet()->NumberOfColumnFamilies() > 1) {
     // If there are more than one column families, we need to make sure that
@@ -188,15 +181,24 @@ 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.
-    io_s = SyncClosedLogs(job_context);
-    s = io_s;
-    if (!io_s.ok() && !io_s.IsShutdownInProgress() &&
-        !io_s.IsColumnFamilyDropped()) {
-      error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+    log_io_s = SyncClosedLogs(job_context);
+    if (!log_io_s.ok() && !log_io_s.IsShutdownInProgress() &&
+        !log_io_s.IsColumnFamilyDropped()) {
+      error_handler_.SetBGError(log_io_s, BackgroundErrorReason::kFlush);
     }
   } else {
     TEST_SYNC_POINT("DBImpl::SyncClosedLogs:Skip");
   }
+  s = log_io_s;
+
+  // If the log sync failed, we do not need to pick memtable. Otherwise,
+  // num_flush_not_started_ needs to be rollback.
+  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:BeforePickMemtables");
+  if (s.ok()) {
+    flush_job.PickMemTable();
+    need_cancel = true;
+  }
+  TEST_SYNC_POINT("DBImpl::FlushMemTableToOutputFile:AfterPickMemtables");
 
   // Within flush_job.Run, rocksdb may call event listener to notify
   // file creation and deletion.
@@ -206,11 +208,16 @@ Status DBImpl::FlushMemTableToOutputFile(
   // is unlocked by the current thread.
   if (s.ok()) {
     s = flush_job.Run(&logs_with_prep_tracker_, &file_meta);
-  } else {
+    need_cancel = false;
+  }
+
+  if (!s.ok() && need_cancel) {
     flush_job.Cancel();
   }
-  if (io_s.ok()) {
-    io_s = flush_job.io_status();
+  IOStatus io_s = IOStatus::OK();
+  io_s = flush_job.io_status();
+  if (s.ok()) {
+    s = io_s;
   }
 
   if (s.ok()) {
@@ -246,29 +253,30 @@ Status DBImpl::FlushMemTableToOutputFile(
   if (!s.ok() && !s.IsShutdownInProgress() && !s.IsColumnFamilyDropped()) {
     if (!io_s.ok() && !io_s.IsShutdownInProgress() &&
         !io_s.IsColumnFamilyDropped()) {
+      assert(log_io_s.ok());
       // Error while writing to MANIFEST.
       // In fact, versions_->io_status() can also be the result of renaming
       // CURRENT file. With current code, it's just difficult to tell. So just
       // be pessimistic and try write to a new MANIFEST.
       // TODO: distinguish between MANIFEST write and CURRENT renaming
       if (!versions_->io_status().ok()) {
-        if (total_log_size_ > 0) {
-          // If the WAL is empty, we use different error reason
-          error_handler_.SetBGError(io_s,
-                                    BackgroundErrorReason::kManifestWrite);
-        } else {
-          error_handler_.SetBGError(io_s,
-                                    BackgroundErrorReason::kManifestWriteNoWAL);
-        }
-      } else if (total_log_size_ > 0) {
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+        // If WAL sync is successful (either WAL size is 0 or there is no IO
+        // error), all the Manifest write will be map to soft error.
+        // TODO: kManifestWriteNoWAL and kFlushNoWAL are misleading. Refactor is
+        // needed.
+        error_handler_.SetBGError(io_s,
+                                  BackgroundErrorReason::kManifestWriteNoWAL);
       } else {
-        // If the WAL is empty, we use different error reason
+        // If WAL sync is successful (either WAL size is 0 or there is no IO
+        // error), all the other SST file write errors will be set as
+        // kFlushNoWAL.
         error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
       }
     } else {
-      Status new_bg_error = s;
-      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
+      if (log_io_s.ok()) {
+        Status new_bg_error = s;
+        error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
+      }
     }
   } else {
     // If we got here, then we decided not to care about the i_os status (either
@@ -401,12 +409,11 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
         false /* sync_output_directory */, false /* write_manifest */,
         thread_pri, io_tracer_, db_id_, db_session_id_,
         cfd->GetFullHistoryTsLow()));
-    jobs.back()->PickMemTable();
   }
 
   std::vector<FileMetaData> file_meta(num_cfs);
   Status s;
-  IOStatus io_s;
+  IOStatus log_io_s = IOStatus::OK();
   assert(num_cfs == static_cast<int>(jobs.size()));
 
 #ifndef ROCKSDB_LITE
@@ -421,18 +428,36 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
   if (logfile_number_ > 0) {
     // TODO (yanqin) investigate whether we should sync the closed logs for
     // single column family case.
-    io_s = SyncClosedLogs(job_context);
-    s = io_s;
+    log_io_s = SyncClosedLogs(job_context);
+    if (!log_io_s.ok() && !log_io_s.IsShutdownInProgress() &&
+        !log_io_s.IsColumnFamilyDropped()) {
+      if (total_log_size_ > 0) {
+        error_handler_.SetBGError(log_io_s, BackgroundErrorReason::kFlush);
+      } else {
+        // If the WAL is empty, we use different error reason
+        error_handler_.SetBGError(log_io_s, BackgroundErrorReason::kFlushNoWAL);
+      }
+    }
   }
+  s = log_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;
+  std::vector<bool> pick_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());
+    pick_status.push_back(false);
+  }
+
+  if (s.ok()) {
+    for (int i = 0; i != num_cfs; ++i) {
+      jobs[i]->PickMemTable();
+      pick_status[i] = true;
+    }
   }
 
   if (s.ok()) {
@@ -473,6 +498,7 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
     s = error_status.ok() ? s : error_status;
   }
 
+  IOStatus io_s = IOStatus::OK();
   if (io_s.ok()) {
     IOStatus io_error = IOStatus::OK();
     for (int i = 0; i != static_cast<int>(io_status.size()); i++) {
@@ -508,12 +534,12 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
     // 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) {
+      if (pick_status[i] && !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()) {
+      if (exec_status[i].second.ok() && exec_status[i].first) {
         auto& mems = jobs[i]->GetMemTables();
         cfds[i]->imm()->RollbackMemtableFlush(mems,
                                               file_meta[i].fd.GetNumber());
@@ -652,29 +678,30 @@ Status DBImpl::AtomicFlushMemTablesToOutputFiles(
   // it is not because of CF drop.
   if (!s.ok() && !s.IsColumnFamilyDropped()) {
     if (!io_s.ok() && !io_s.IsColumnFamilyDropped()) {
+      assert(log_io_s.ok());
       // Error while writing to MANIFEST.
       // In fact, versions_->io_status() can also be the result of renaming
       // CURRENT file. With current code, it's just difficult to tell. So just
       // be pessimistic and try write to a new MANIFEST.
       // TODO: distinguish between MANIFEST write and CURRENT renaming
       if (!versions_->io_status().ok()) {
-        if (total_log_size_ > 0) {
-          // If the WAL is empty, we use different error reason
-          error_handler_.SetBGError(io_s,
-                                    BackgroundErrorReason::kManifestWrite);
-        } else {
-          error_handler_.SetBGError(io_s,
-                                    BackgroundErrorReason::kManifestWriteNoWAL);
-        }
-      } else if (total_log_size_ > 0) {
-        error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlush);
+        // If WAL sync is successful (either WAL size is 0 or there is no IO
+        // error), all the Manifest write will be map to soft error.
+        // TODO: kManifestWriteNoWAL and kFlushNoWAL are misleading. Refactor
+        // is needed.
+        error_handler_.SetBGError(io_s,
+                                  BackgroundErrorReason::kManifestWriteNoWAL);
       } else {
-        // If the WAL is empty, we use different error reason
+        // If WAL sync is successful (either WAL size is 0 or there is no IO
+        // error), all the other SST file write errors will be set as
+        // kFlushNoWAL.
         error_handler_.SetBGError(io_s, BackgroundErrorReason::kFlushNoWAL);
       }
     } else {
-      Status new_bg_error = s;
-      error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
+      if (log_io_s.ok()) {
+        Status new_bg_error = s;
+        error_handler_.SetBGError(new_bg_error, BackgroundErrorReason::kFlush);
+      }
     }
   }
 
@@ -806,6 +833,25 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
                               end_with_ts);
 }
 
+Status DBImpl::IncreaseFullHistoryTsLow(ColumnFamilyData* cfd,
+                                        std::string ts_low) {
+  VersionEdit edit;
+  edit.SetColumnFamily(cfd->GetID());
+  edit.SetFullHistoryTsLow(ts_low);
+
+  InstrumentedMutexLock l(&mutex_);
+  std::string current_ts_low = cfd->GetFullHistoryTsLow();
+  const Comparator* ucmp = cfd->user_comparator();
+  if (!current_ts_low.empty() &&
+      ucmp->CompareTimestamp(ts_low, current_ts_low) < 0) {
+    return Status::InvalidArgument(
+        "Cannot decrease full_history_timestamp_low");
+  }
+
+  return versions_->LogAndApply(cfd, *cfd->GetLatestMutableCFOptions(), &edit,
+                                &mutex_);
+}
+
 Status DBImpl::CompactRangeInternal(const CompactRangeOptions& options,
                                     ColumnFamilyHandle* column_family,
                                     const Slice* begin, const Slice* end) {
@@ -817,6 +863,22 @@ Status DBImpl::CompactRangeInternal(const CompactRangeOptions& options,
   }
 
   bool flush_needed = true;
+
+  // Update full_history_ts_low if it's set
+  if (options.full_history_ts_low != nullptr &&
+      !options.full_history_ts_low->empty()) {
+    std::string ts_low = options.full_history_ts_low->ToString();
+    if (begin != nullptr || end != nullptr) {
+      return Status::InvalidArgument(
+          "Cannot specify compaction range with full_history_ts_low");
+    }
+    Status s = IncreaseFullHistoryTsLow(cfd, ts_low);
+    if (!s.ok()) {
+      LogFlush(immutable_db_options_.info_log);
+      return s;
+    }
+  }
+
   Status s;
   if (begin != nullptr && end != nullptr) {
     // TODO(ajkr): We could also optimize away the flush in certain cases where
@@ -2026,12 +2088,12 @@ Status DBImpl::WaitUntilFlushWouldNotStallWrites(ColumnFamilyData* cfd,
       // check whether one extra immutable memtable or an extra L0 file would
       // cause write stalling mode to be entered. It could still enter stall
       // mode due to pending compaction bytes, but that's less common
-      write_stall_condition =
-          ColumnFamilyData::GetWriteStallConditionAndCause(
-              cfd->imm()->NumNotFlushed() + 1,
-              vstorage->l0_delay_trigger_count() + 1,
-              vstorage->estimated_compaction_needed_bytes(), mutable_cf_options)
-              .first;
+      write_stall_condition = ColumnFamilyData::GetWriteStallConditionAndCause(
+                                  cfd->imm()->NumNotFlushed() + 1,
+                                  vstorage->l0_delay_trigger_count() + 1,
+                                  vstorage->estimated_compaction_needed_bytes(),
+                                  mutable_cf_options, *cfd->ioptions())
+                                  .first;
     } while (write_stall_condition != WriteStallCondition::kNormal);
   }
   return Status::OK();
@@ -2501,6 +2563,8 @@ void DBImpl::BackgroundCallFlush(Env::Priority thread_pri) {
 
   LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL,
                        immutable_db_options_.info_log.get());
+  TEST_SYNC_POINT("DBImpl::BackgroundCallFlush:Start:1");
+  TEST_SYNC_POINT("DBImpl::BackgroundCallFlush:Start:2");
   {
     InstrumentedMutexLock l(&mutex_);
     assert(bg_flush_scheduled_);
@@ -2529,7 +2593,7 @@ void DBImpl::BackgroundCallFlush(Env::Priority thread_pri) {
                       s.ToString().c_str(), error_cnt);
       log_buffer.FlushBufferToLog();
       LogFlush(immutable_db_options_.info_log);
-      env_->SleepForMicroseconds(1000000);
+      immutable_db_options_.clock->SleepForMicroseconds(1000000);
       mutex_.Lock();
     }
 
@@ -2602,7 +2666,8 @@ void DBImpl::BackgroundCallCompaction(PrepickedCompaction* prepicked_compaction,
     if (s.IsBusy()) {
       bg_cv_.SignalAll();  // In case a waiter can proceed despite the error
       mutex_.Unlock();
-      env_->SleepForMicroseconds(10000);  // prevent hot loop
+      immutable_db_options_.clock->SleepForMicroseconds(
+          10000);  // prevent hot loop
       mutex_.Lock();
     } else if (!s.ok() && !s.IsShutdownInProgress() &&
                !s.IsManualCompactionPaused() && !s.IsColumnFamilyDropped()) {
@@ -2620,7 +2685,7 @@ void DBImpl::BackgroundCallCompaction(PrepickedCompaction* prepicked_compaction,
                       "Accumulated background error counts: %" PRIu64,
                       s.ToString().c_str(), error_cnt);
       LogFlush(immutable_db_options_.info_log);
-      env_->SleepForMicroseconds(1000000);
+      immutable_db_options_.clock->SleepForMicroseconds(1000000);
       mutex_.Lock();
     } else if (s.IsManualCompactionPaused()) {
       ManualCompactionState* m = prepicked_compaction->manual_compaction_state;

db/db_impl/db_impl_files.cc

@@ -120,7 +120,7 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
              mutable_db_options_.delete_obsolete_files_period_micros == 0) {
     doing_the_full_scan = true;
   } else {
-    const uint64_t now_micros = env_->NowMicros();
+    const uint64_t now_micros = immutable_db_options_.clock->NowMicros();
     if ((delete_obsolete_files_last_run_ +
          mutable_db_options_.delete_obsolete_files_period_micros) <
         now_micros) {
@@ -937,7 +937,7 @@ Status DBImpl::DeleteUnreferencedSstFiles() {
     return s;
   }
 
-  if (largest_file_number > next_file_number) {
+  if (largest_file_number >= next_file_number) {
     versions_->next_file_number_.store(largest_file_number + 1);
   }
 

db/db_impl/db_impl_open.cc

@@ -175,7 +175,7 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
         NewSstFileManager(result.env, result.info_log));
     result.sst_file_manager = sst_file_manager;
   }
-#endif
+#endif  // !ROCKSDB_LITE
 
   if (!result.paranoid_checks) {
     result.skip_checking_sst_file_sizes_on_db_open = true;
@@ -283,23 +283,28 @@ Status DBImpl::NewDB(std::vector<std::string>* new_filenames) {
   ROCKS_LOG_INFO(immutable_db_options_.info_log, "Creating manifest 1 \n");
   const std::string manifest = DescriptorFileName(dbname_, 1);
   {
+    if (fs_->FileExists(manifest, IOOptions(), nullptr).ok()) {
+      fs_->DeleteFile(manifest, IOOptions(), nullptr).PermitUncheckedError();
+    }
     std::unique_ptr<FSWritableFile> file;
     FileOptions file_options = fs_->OptimizeForManifestWrite(file_options_);
     s = NewWritableFile(fs_.get(), manifest, &file, file_options);
     if (!s.ok()) {
       return s;
     }
+    FileTypeSet tmp_set = immutable_db_options_.checksum_handoff_file_types;
     file->SetPreallocationBlockSize(
         immutable_db_options_.manifest_preallocation_size);
     std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-        std::move(file), manifest, file_options, env_, io_tracer_,
-        nullptr /* stats */, immutable_db_options_.listeners));
+        std::move(file), manifest, file_options, immutable_db_options_.clock,
+        io_tracer_, nullptr /* stats */, immutable_db_options_.listeners,
+        nullptr, tmp_set.Contains(FileType::kDescriptorFile)));
     log::Writer log(std::move(file_writer), 0, false);
     std::string record;
     new_db.EncodeTo(&record);
     s = log.AddRecord(record);
     if (s.ok()) {
-      s = SyncManifest(env_, &immutable_db_options_, log.file());
+      s = SyncManifest(&immutable_db_options_, log.file());
     }
   }
   if (s.ok()) {
@@ -310,7 +315,7 @@ Status DBImpl::NewDB(std::vector<std::string>* new_filenames) {
           manifest.substr(manifest.find_last_of("/\\") + 1));
     }
   } else {
-    fs_->DeleteFile(manifest, IOOptions(), nullptr);
+    fs_->DeleteFile(manifest, IOOptions(), nullptr).PermitUncheckedError();
   }
   return s;
 }
@@ -1130,11 +1135,29 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& wal_numbers,
        immutable_db_options_.wal_recovery_mode ==
            WALRecoveryMode::kTolerateCorruptedTailRecords)) {
     for (auto cfd : *versions_->GetColumnFamilySet()) {
-      if (cfd->GetLogNumber() > corrupted_wal_number) {
+      // One special case cause cfd->GetLogNumber() > corrupted_wal_number but
+      // the CF is still consistent: If a new column family is created during
+      // the flush and the WAL sync fails at the same time, the new CF points to
+      // the new WAL but the old WAL is curropted. Since the new CF is empty, it
+      // is still consistent. We add the check of CF sst file size to avoid the
+      // false positive alert.
+
+      // Note that, the check of (cfd->GetLiveSstFilesSize() > 0) may leads to
+      // the ignorance of a very rare inconsistency case caused in data
+      // canclation. One CF is empty due to KV deletion. But those operations
+      // are in the WAL. If the WAL is corrupted, the status of this CF might
+      // not be consistent with others. However, the consistency check will be
+      // bypassed due to empty CF.
+      // TODO: a better and complete implementation is needed to ensure strict
+      // consistency check in WAL recovery including hanlding the tailing
+      // issues.
+      if (cfd->GetLogNumber() > corrupted_wal_number &&
+          cfd->GetLiveSstFilesSize() > 0) {
         ROCKS_LOG_ERROR(immutable_db_options_.info_log,
                         "Column family inconsistency: SST file contains data"
                         " beyond the point of corruption.");
-        return Status::Corruption("SST file is ahead of WALs");
+        return Status::Corruption("SST file is ahead of WALs in CF " +
+                                  cfd->GetName());
       }
     }
   }
@@ -1295,7 +1318,7 @@ Status DBImpl::RestoreAliveLogFiles(const std::vector<uint64_t>& wal_numbers) {
 Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
                                            MemTable* mem, VersionEdit* edit) {
   mutex_.AssertHeld();
-  const uint64_t start_micros = env_->NowMicros();
+  const uint64_t start_micros = immutable_db_options_.clock->NowMicros();
 
   FileMetaData meta;
   std::vector<BlobFileAddition> blob_file_additions;
@@ -1323,7 +1346,7 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
         cfd->GetLatestMutableCFOptions()->paranoid_file_checks;
 
     int64_t _current_time = 0;
-    env_->GetCurrentTime(&_current_time)
+    immutable_db_options_.clock->GetCurrentTime(&_current_time)
         .PermitUncheckedError();  // ignore error
     const uint64_t current_time = static_cast<uint64_t>(_current_time);
     meta.oldest_ancester_time = current_time;
@@ -1362,7 +1385,8 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
           io_tracer_, &event_logger_, job_id, Env::IO_HIGH,
           nullptr /* table_properties */, -1 /* level */, current_time,
           0 /* oldest_key_time */, write_hint, 0 /* file_creation_time */,
-          db_id_, db_session_id_);
+          db_id_, db_session_id_, nullptr /*full_history_ts_low*/,
+          &blob_callback_);
       LogFlush(immutable_db_options_.info_log);
       ROCKS_LOG_DEBUG(immutable_db_options_.info_log,
                       "[%s] [WriteLevel0TableForRecovery]"
@@ -1391,11 +1415,13 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
                   meta.oldest_ancester_time, meta.file_creation_time,
                   meta.file_checksum, meta.file_checksum_func_name);
 
-    edit->SetBlobFileAdditions(std::move(blob_file_additions));
+    for (const auto& blob : blob_file_additions) {
+      edit->AddBlobFile(blob);
+    }
   }
 
   InternalStats::CompactionStats stats(CompactionReason::kFlush, 1);
-  stats.micros = env_->NowMicros() - start_micros;
+  stats.micros = immutable_db_options_.clock->NowMicros() - start_micros;
 
   if (has_output) {
     stats.bytes_written = meta.fd.GetFileSize();
@@ -1404,14 +1430,15 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
 
   const auto& blobs = edit->GetBlobFileAdditions();
   for (const auto& blob : blobs) {
-    stats.bytes_written += blob.GetTotalBlobBytes();
+    stats.bytes_written_blob += blob.GetTotalBlobBytes();
   }
 
-  stats.num_output_files += static_cast<int>(blobs.size());
+  stats.num_output_files_blob = static_cast<int>(blobs.size());
 
   cfd->internal_stats()->AddCompactionStats(level, Env::Priority::USER, stats);
-  cfd->internal_stats()->AddCFStats(InternalStats::BYTES_FLUSHED,
-                                    stats.bytes_written);
+  cfd->internal_stats()->AddCFStats(
+      InternalStats::BYTES_FLUSHED,
+      stats.bytes_written + stats.bytes_written_blob);
   RecordTick(stats_, COMPACT_WRITE_BYTES, meta.fd.GetFileSize());
   return s;
 }
@@ -1485,9 +1512,11 @@ IOStatus DBImpl::CreateWAL(uint64_t log_file_num, uint64_t recycle_log_number,
     lfile->SetPreallocationBlockSize(preallocate_block_size);
 
     const auto& listeners = immutable_db_options_.listeners;
+    FileTypeSet tmp_set = immutable_db_options_.checksum_handoff_file_types;
     std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-        std::move(lfile), log_fname, opt_file_options, env_, io_tracer_,
-        nullptr /* stats */, listeners));
+        std::move(lfile), log_fname, opt_file_options,
+        immutable_db_options_.clock, io_tracer_, nullptr /* stats */, listeners,
+        nullptr, tmp_set.Contains(FileType::kWalFile)));
     *new_log = new log::Writer(std::move(file_writer), log_file_num,
                                immutable_db_options_.recycle_log_file_num > 0,
                                immutable_db_options_.manual_wal_flush);
@@ -1716,6 +1745,8 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
 
     std::vector<LiveFileMetaData> metadata;
 
+    // TODO: Once GetLiveFilesMetaData supports blob files, update the logic
+    // below to get known_file_sizes for blob files.
     impl->mutex_.Lock();
     impl->versions_->GetLiveFilesMetaData(&metadata);
     impl->mutex_.Unlock();
@@ -1748,13 +1779,12 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
         FileType file_type;
         std::string file_path = path + "/" + file_name;
         if (ParseFileName(file_name, &file_number, &file_type) &&
-            file_type == kTableFile) {
+            (file_type == kTableFile || file_type == kBlobFile)) {
           // TODO: Check for errors from OnAddFile?
           if (known_file_sizes.count(file_name)) {
             // We're assuming that each sst file name exists in at most one of
             // the paths.
-            sfm->OnAddFile(file_path, known_file_sizes.at(file_name),
-                           /* compaction */ false)
+            sfm->OnAddFile(file_path, known_file_sizes.at(file_name))
                 .PermitUncheckedError();
           } else {
             sfm->OnAddFile(file_path).PermitUncheckedError();

db/db_impl/db_impl_readonly.cc

@@ -143,9 +143,6 @@ Status OpenForReadOnlyCheckExistence(const DBOptions& db_options,
     uint64_t manifest_file_number;
     s = VersionSet::GetCurrentManifestPath(dbname, fs.get(), &manifest_path,
                                            &manifest_file_number);
-    if (!s.ok()) {
-      return Status::NotFound(CurrentFileName(dbname), "does not exist");
-    }
   }
   return s;
 }

db/db_impl/db_impl_secondary.cc

@@ -327,8 +327,8 @@ Status DBImplSecondary::GetImpl(const ReadOptions& read_options,
                                 ColumnFamilyHandle* column_family,
                                 const Slice& key, PinnableSlice* pinnable_val) {
   assert(pinnable_val != nullptr);
-  PERF_CPU_TIMER_GUARD(get_cpu_nanos, env_);
-  StopWatch sw(env_, stats_, DB_GET);
+  PERF_CPU_TIMER_GUARD(get_cpu_nanos, immutable_db_options_.clock);
+  StopWatch sw(immutable_db_options_.clock, stats_, DB_GET);
   PERF_TIMER_GUARD(get_snapshot_time);
 
   auto cfh = static_cast<ColumnFamilyHandleImpl*>(column_family);
@@ -519,7 +519,8 @@ Status DBImplSecondary::TryCatchUpWithPrimary() {
   {
     InstrumentedMutexLock lock_guard(&mutex_);
     s = static_cast_with_check<ReactiveVersionSet>(versions_.get())
-            ->ReadAndApply(&mutex_, &manifest_reader_, &cfds_changed);
+            ->ReadAndApply(&mutex_, &manifest_reader_,
+                           manifest_reader_status_.get(), &cfds_changed);
 
     ROCKS_LOG_INFO(immutable_db_options_.info_log, "Last sequence is %" PRIu64,
                    static_cast<uint64_t>(versions_->LastSequence()));

db/db_impl/db_impl_write.cc

@@ -160,7 +160,8 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
     RecordTick(stats_, WRITE_WITH_WAL);
   }
 
-  StopWatch write_sw(env_, immutable_db_options_.statistics.get(), DB_WRITE);
+  StopWatch write_sw(immutable_db_options_.clock,
+                     immutable_db_options_.statistics.get(), DB_WRITE);
 
   write_thread_.JoinBatchGroup(&w);
   if (w.state == WriteThread::STATE_PARALLEL_MEMTABLE_WRITER) {
@@ -465,7 +466,8 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
                                   uint64_t* log_used, uint64_t log_ref,
                                   bool disable_memtable, uint64_t* seq_used) {
   PERF_TIMER_GUARD(write_pre_and_post_process_time);
-  StopWatch write_sw(env_, immutable_db_options_.statistics.get(), DB_WRITE);
+  StopWatch write_sw(immutable_db_options_.clock,
+                     immutable_db_options_.statistics.get(), DB_WRITE);
 
   WriteContext write_context;
 
@@ -621,7 +623,8 @@ Status DBImpl::UnorderedWriteMemtable(const WriteOptions& write_options,
                                       SequenceNumber seq,
                                       const size_t sub_batch_cnt) {
   PERF_TIMER_GUARD(write_pre_and_post_process_time);
-  StopWatch write_sw(env_, immutable_db_options_.statistics.get(), DB_WRITE);
+  StopWatch write_sw(immutable_db_options_.clock,
+                     immutable_db_options_.statistics.get(), DB_WRITE);
 
   WriteThread::Writer w(write_options, my_batch, callback, log_ref,
                         false /*disable_memtable*/);
@@ -676,7 +679,8 @@ Status DBImpl::WriteImplWALOnly(
   WriteThread::Writer w(write_options, my_batch, callback, log_ref,
                         disable_memtable, sub_batch_cnt, pre_release_callback);
   RecordTick(stats_, WRITE_WITH_WAL);
-  StopWatch write_sw(env_, immutable_db_options_.statistics.get(), DB_WRITE);
+  StopWatch write_sw(immutable_db_options_.clock,
+                     immutable_db_options_.statistics.get(), DB_WRITE);
 
   write_thread->JoinBatchGroup(&w);
   assert(w.state != WriteThread::STATE_PARALLEL_MEMTABLE_WRITER);
@@ -1093,7 +1097,7 @@ IOStatus DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
   }
 
   if (io_s.ok() && need_log_sync) {
-    StopWatch sw(env_, stats_, WAL_FILE_SYNC_MICROS);
+    StopWatch sw(immutable_db_options_.clock, 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,
     //    so other threads won't pop from logs_ while we're here,
@@ -1457,8 +1461,10 @@ Status DBImpl::DelayWrite(uint64_t num_bytes,
   uint64_t time_delayed = 0;
   bool delayed = false;
   {
-    StopWatch sw(env_, stats_, WRITE_STALL, &time_delayed);
-    uint64_t delay = write_controller_.GetDelay(env_, num_bytes);
+    StopWatch sw(immutable_db_options_.clock, stats_, WRITE_STALL,
+                 &time_delayed);
+    uint64_t delay =
+        write_controller_.GetDelay(immutable_db_options_.clock, num_bytes);
     if (delay > 0) {
       if (write_options.no_slowdown) {
         return Status::Incomplete("Write stall");
@@ -1470,19 +1476,21 @@ Status DBImpl::DelayWrite(uint64_t num_bytes,
       write_thread_.BeginWriteStall();
       TEST_SYNC_POINT("DBImpl::DelayWrite:BeginWriteStallDone");
       mutex_.Unlock();
-      // We will delay the write until we have slept for delay ms or
-      // we don't need a delay anymore
-      const uint64_t kDelayInterval = 1000;
+      // We will delay the write until we have slept for `delay` microseconds
+      // or we don't need a delay anymore. We check for cancellation every 1ms
+      // (slightly longer because WriteController minimum delay is 1ms, in
+      // case of sleep imprecision, rounding, etc.)
+      const uint64_t kDelayInterval = 1001;
       uint64_t stall_end = sw.start_time() + delay;
       while (write_controller_.NeedsDelay()) {
-        if (env_->NowMicros() >= stall_end) {
+        if (immutable_db_options_.clock->NowMicros() >= stall_end) {
           // We already delayed this write `delay` microseconds
           break;
         }
 
         delayed = true;
         // Sleep for 0.001 seconds
-        env_->SleepForMicroseconds(kDelayInterval);
+        immutable_db_options_.clock->SleepForMicroseconds(kDelayInterval);
       }
       mutex_.Lock();
       write_thread_.EndWriteStall();

db/db_info_dumper.cc

@@ -64,9 +64,10 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
         break;
       case kWalFile:
         if (env->GetFileSize(dbname + "/" + file, &file_size).ok()) {
-          char str[16];
-          snprintf(str, sizeof(str), "%" PRIu64, file_size);
-          wal_info.append(file).append(" size: ").append(str).append(" ; ");
+          wal_info.append(file)
+              .append(" size: ")
+              .append(std::to_string(file_size))
+              .append(" ; ");
         } else {
           Error(options.info_log, "Error when reading LOG file: %s/%s\n",
                 dbname.c_str(), file.c_str());
@@ -120,9 +121,10 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
       if (ParseFileName(file, &number, &type)) {
         if (type == kWalFile) {
           if (env->GetFileSize(options.wal_dir + "/" + file, &file_size).ok()) {
-            char str[16];
-            snprintf(str, sizeof(str), "%" PRIu64, file_size);
-            wal_info.append(file).append(" size: ").append(str).append(" ; ");
+            wal_info.append(file)
+                .append(" size: ")
+                .append(std::to_string(file_size))
+                .append(" ; ");
           } else {
             Error(options.info_log, "Error when reading LOG file %s/%s\n",
                   options.wal_dir.c_str(), file.c_str());

db/db_iter.cc

@@ -8,9 +8,10 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 
 #include "db/db_iter.h"
-#include <string>
+
 #include <iostream>
 #include <limits>
+#include <string>
 
 #include "db/dbformat.h"
 #include "db/merge_context.h"
@@ -24,6 +25,7 @@
 #include "rocksdb/iterator.h"
 #include "rocksdb/merge_operator.h"
 #include "rocksdb/options.h"
+#include "rocksdb/system_clock.h"
 #include "table/internal_iterator.h"
 #include "table/iterator_wrapper.h"
 #include "trace_replay/trace_replay.h"
@@ -43,6 +45,7 @@ DBIter::DBIter(Env* _env, const ReadOptions& read_options,
                ColumnFamilyData* cfd, bool expose_blob_index)
     : prefix_extractor_(mutable_cf_options.prefix_extractor.get()),
       env_(_env),
+      clock_(_env->GetSystemClock().get()),
       logger_(cf_options.info_log),
       user_comparator_(cmp),
       merge_operator_(cf_options.merge_operator),
@@ -111,8 +114,7 @@ Status DBIter::GetProperty(std::string prop_name, std::string* prop) {
 }
 
 bool DBIter::ParseKey(ParsedInternalKey* ikey) {
-  Status s =
-      ParseInternalKey(iter_.key(), ikey, false /* log_err_key */);  // TODO
+  Status s = ParseInternalKey(iter_.key(), ikey, false /* log_err_key */);
   if (!s.ok()) {
     status_ = Status::Corruption("In DBIter: ", s.getState());
     valid_ = false;
@@ -127,7 +129,7 @@ void DBIter::Next() {
   assert(valid_);
   assert(status_.ok());
 
-  PERF_CPU_TIMER_GUARD(iter_next_cpu_nanos, env_);
+  PERF_CPU_TIMER_GUARD(iter_next_cpu_nanos, clock_);
   // Release temporarily pinned blocks from last operation
   ReleaseTempPinnedData();
   local_stats_.skip_count_ += num_internal_keys_skipped_;
@@ -190,8 +192,10 @@ bool DBIter::SetBlobValueIfNeeded(const Slice& user_key,
   read_options.read_tier = read_tier_;
   read_options.verify_checksums = verify_checksums_;
 
-  const Status s =
-      version_->GetBlob(read_options, user_key, blob_index, &blob_value_);
+  constexpr uint64_t* bytes_read = nullptr;
+
+  const Status s = version_->GetBlob(read_options, user_key, blob_index,
+                                     &blob_value_, bytes_read);
 
   if (!s.ok()) {
     status_ = s;
@@ -484,7 +488,7 @@ bool DBIter::FindNextUserEntryInternal(bool skipping_saved_key,
               &last_key,
               ParsedInternalKey(saved_key_.GetUserKey(), 0, kTypeDeletion));
         } else {
-          const std::string kTsMin(timestamp_size_, static_cast<char>(0));
+          const std::string kTsMin(timestamp_size_, '\0');
           AppendInternalKeyWithDifferentTimestamp(
               &last_key,
               ParsedInternalKey(saved_key_.GetUserKey(), 0, kTypeDeletion),
@@ -574,7 +578,7 @@ bool DBIter::MergeValuesNewToOld() {
       const Slice val = iter_.value();
       Status s = MergeHelper::TimedFullMerge(
           merge_operator_, ikey.user_key, &val, merge_context_.GetOperands(),
-          &saved_value_, logger_, statistics_, env_, &pinned_value_, true);
+          &saved_value_, logger_, statistics_, clock_, &pinned_value_, true);
       if (!s.ok()) {
         valid_ = false;
         status_ = s;
@@ -617,7 +621,7 @@ bool DBIter::MergeValuesNewToOld() {
   // client can differentiate this scenario and do things accordingly.
   Status s = MergeHelper::TimedFullMerge(
       merge_operator_, saved_key_.GetUserKey(), nullptr,
-      merge_context_.GetOperands(), &saved_value_, logger_, statistics_, env_,
+      merge_context_.GetOperands(), &saved_value_, logger_, statistics_, clock_,
       &pinned_value_, true);
   if (!s.ok()) {
     valid_ = false;
@@ -630,17 +634,10 @@ 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());
 
-  PERF_CPU_TIMER_GUARD(iter_prev_cpu_nanos, env_);
+  PERF_CPU_TIMER_GUARD(iter_prev_cpu_nanos, clock_);
   ReleaseTempPinnedData();
   ResetInternalKeysSkippedCounter();
   bool ok = true;
@@ -675,9 +672,16 @@ bool DBIter::ReverseToForward() {
   // If that's the case, seek iter_ to current key.
   if (!expect_total_order_inner_iter() || !iter_.Valid()) {
     IterKey last_key;
-    last_key.SetInternalKey(ParsedInternalKey(
-        saved_key_.GetUserKey(), kMaxSequenceNumber, kValueTypeForSeek));
+    ParsedInternalKey pikey(saved_key_.GetUserKey(), kMaxSequenceNumber,
+                            kValueTypeForSeek);
+    if (timestamp_size_ > 0) {
+      // TODO: pre-create kTsMax.
+      const std::string kTsMax(timestamp_size_, '\xff');
+      pikey.SetTimestamp(kTsMax);
+    }
+    last_key.SetInternalKey(pikey);
     iter_.Seek(last_key.GetInternalKey());
+    RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
   }
 
   direction_ = kForward;
@@ -728,6 +732,7 @@ bool DBIter::ReverseToBackward() {
         iter_.SeekToLast();
       }
     }
+    RecordTick(statistics_, NUMBER_OF_RESEEKS_IN_ITERATION);
   }
 
   direction_ = kReverse;
@@ -753,11 +758,13 @@ void DBIter::PrevInternal(const Slice* prefix) {
     }
 
     assert(iterate_lower_bound_ == nullptr || iter_.MayBeOutOfLowerBound() ||
-           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);
     if (iterate_lower_bound_ != nullptr && iter_.MayBeOutOfLowerBound() &&
-        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) {
       // We've iterated earlier than the user-specified lower bound.
       valid_ = false;
       return;
@@ -802,8 +809,8 @@ bool DBIter::FindValueForCurrentKey() {
   assert(iter_.Valid());
   merge_context_.Clear();
   current_entry_is_merged_ = false;
-  // last entry before merge (could be kTypeDeletion, kTypeSingleDeletion or
-  // kTypeValue)
+  // last entry before merge (could be kTypeDeletion,
+  // kTypeDeletionWithTimestamp, kTypeSingleDeletion or kTypeValue)
   ValueType last_not_merge_type = kTypeDeletion;
   ValueType last_key_entry_type = kTypeDeletion;
 
@@ -824,9 +831,13 @@ bool DBIter::FindValueForCurrentKey() {
                  timestamp_size_);
     }
     if (!IsVisible(ikey.sequence, ts) ||
-        !user_comparator_.Equal(ikey.user_key, saved_key_.GetUserKey())) {
+        !user_comparator_.EqualWithoutTimestamp(ikey.user_key,
+                                                saved_key_.GetUserKey())) {
       break;
     }
+    if (!ts.empty()) {
+      saved_timestamp_.assign(ts.data(), ts.size());
+    }
     if (TooManyInternalKeysSkipped()) {
       return false;
     }
@@ -866,6 +877,7 @@ bool DBIter::FindValueForCurrentKey() {
         }
         break;
       case kTypeDeletion:
+      case kTypeDeletionWithTimestamp:
       case kTypeSingleDeletion:
         merge_context_.Clear();
         last_not_merge_type = last_key_entry_type;
@@ -909,6 +921,7 @@ bool DBIter::FindValueForCurrentKey() {
   is_blob_ = false;
   switch (last_key_entry_type) {
     case kTypeDeletion:
+    case kTypeDeletionWithTimestamp:
     case kTypeSingleDeletion:
     case kTypeRangeDeletion:
       valid_ = false;
@@ -921,7 +934,7 @@ bool DBIter::FindValueForCurrentKey() {
         s = MergeHelper::TimedFullMerge(
             merge_operator_, saved_key_.GetUserKey(), nullptr,
             merge_context_.GetOperands(), &saved_value_, logger_, statistics_,
-            env_, &pinned_value_, true);
+            clock_, &pinned_value_, true);
       } else if (last_not_merge_type == kTypeBlobIndex) {
         status_ =
             Status::NotSupported("BlobDB does not support merge operator.");
@@ -932,7 +945,7 @@ bool DBIter::FindValueForCurrentKey() {
         s = MergeHelper::TimedFullMerge(
             merge_operator_, saved_key_.GetUserKey(), &pinned_value_,
             merge_context_.GetOperands(), &saved_value_, logger_, statistics_,
-            env_, &pinned_value_, true);
+            clock_, &pinned_value_, true);
       }
       break;
     case kTypeValue:
@@ -969,8 +982,17 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
   // FindValueForCurrentKeyUsingSeek()
   assert(pinned_iters_mgr_.PinningEnabled());
   std::string last_key;
-  AppendInternalKey(&last_key, ParsedInternalKey(saved_key_.GetUserKey(),
-                                                 sequence_, kValueTypeForSeek));
+  if (0 == timestamp_size_) {
+    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);
 
@@ -994,7 +1016,8 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
                  timestamp_size_);
     }
 
-    if (!user_comparator_.Equal(ikey.user_key, saved_key_.GetUserKey())) {
+    if (!user_comparator_.EqualWithoutTimestamp(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
       // the entries were discarded in a compaction.
@@ -1011,7 +1034,8 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
 
   if (ikey.type == kTypeDeletion || ikey.type == kTypeSingleDeletion ||
       range_del_agg_.ShouldDelete(
-          ikey, RangeDelPositioningMode::kBackwardTraversal)) {
+          ikey, RangeDelPositioningMode::kBackwardTraversal) ||
+      kTypeDeletionWithTimestamp == ikey.type) {
     valid_ = false;
     return true;
   }
@@ -1019,6 +1043,10 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
     valid_ = false;
     return false;
   }
+  if (timestamp_size_ > 0) {
+    Slice ts = ExtractTimestampFromUserKey(ikey.user_key, timestamp_size_);
+    saved_timestamp_.assign(ts.data(), ts.size());
+  }
   if (ikey.type == kTypeValue || ikey.type == kTypeBlobIndex) {
     assert(iter_.iter()->IsValuePinned());
     pinned_value_ = iter_.value();
@@ -1070,7 +1098,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
       Status s = MergeHelper::TimedFullMerge(
           merge_operator_, saved_key_.GetUserKey(), &val,
           merge_context_.GetOperands(), &saved_value_, logger_, statistics_,
-          env_, &pinned_value_, true);
+          clock_, &pinned_value_, true);
       if (!s.ok()) {
         valid_ = false;
         status_ = s;
@@ -1097,7 +1125,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {
 
   Status s = MergeHelper::TimedFullMerge(
       merge_operator_, saved_key_.GetUserKey(), nullptr,
-      merge_context_.GetOperands(), &saved_value_, logger_, statistics_, env_,
+      merge_context_.GetOperands(), &saved_value_, logger_, statistics_, clock_,
       &pinned_value_, true);
   if (!s.ok()) {
     valid_ = false;
@@ -1135,7 +1163,8 @@ bool DBIter::FindUserKeyBeforeSavedKey() {
       return false;
     }
 
-    if (user_comparator_.Compare(ikey.user_key, saved_key_.GetUserKey()) < 0) {
+    if (user_comparator_.CompareWithoutTimestamp(ikey.user_key,
+                                                 saved_key_.GetUserKey()) < 0) {
       return true;
     }
 
@@ -1159,8 +1188,14 @@ bool DBIter::FindUserKeyBeforeSavedKey() {
     if (num_skipped >= max_skip_) {
       num_skipped = 0;
       IterKey last_key;
-      last_key.SetInternalKey(ParsedInternalKey(
-          saved_key_.GetUserKey(), kMaxSequenceNumber, kValueTypeForSeek));
+      ParsedInternalKey pikey(saved_key_.GetUserKey(), kMaxSequenceNumber,
+                              kValueTypeForSeek);
+      if (timestamp_size_ > 0) {
+        // TODO: pre-create kTsMax.
+        const std::string kTsMax(timestamp_size_, '\xff');
+        pikey.SetTimestamp(kTsMax);
+      }
+      last_key.SetInternalKey(pikey);
       // It would be more efficient to use SeekForPrev() here, but some
       // iterators may not support it.
       iter_.Seek(last_key.GetInternalKey());
@@ -1237,24 +1272,50 @@ void DBIter::SetSavedKeyToSeekForPrevTarget(const Slice& target) {
   saved_key_.Clear();
   // now saved_key is used to store internal key.
   saved_key_.SetInternalKey(target, 0 /* sequence_number */,
-                            kValueTypeForSeekForPrev);
+                            kValueTypeForSeekForPrev, timestamp_ub_);
+
+  if (timestamp_size_ > 0) {
+    const std::string kTsMin(timestamp_size_, '\0');
+    Slice ts = kTsMin;
+    saved_key_.UpdateInternalKey(/*seq=*/0, kValueTypeForSeekForPrev, &ts);
+  }
 
   if (iterate_upper_bound_ != nullptr &&
-      user_comparator_.Compare(saved_key_.GetUserKey(),
-                               *iterate_upper_bound_) >= 0) {
+      user_comparator_.CompareWithoutTimestamp(
+          saved_key_.GetUserKey(), /*a_has_ts=*/true, *iterate_upper_bound_,
+          /*b_has_ts=*/false) >= 0) {
     saved_key_.Clear();
-    saved_key_.SetInternalKey(*iterate_upper_bound_, kMaxSequenceNumber);
+    saved_key_.SetInternalKey(*iterate_upper_bound_, kMaxSequenceNumber,
+                              kValueTypeForSeekForPrev, timestamp_ub_);
+    if (timestamp_size_ > 0) {
+      const std::string kTsMax(timestamp_size_, '\xff');
+      Slice ts = kTsMax;
+      saved_key_.UpdateInternalKey(kMaxSequenceNumber, kValueTypeForSeekForPrev,
+                                   &ts);
+    }
   }
 }
 
 void DBIter::Seek(const Slice& target) {
-  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, env_);
-  StopWatch sw(env_, statistics_, DB_SEEK);
+  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, clock_);
+  StopWatch sw(clock_, statistics_, DB_SEEK);
 
 #ifndef ROCKSDB_LITE
   if (db_impl_ != nullptr && cfd_ != nullptr) {
     // TODO: What do we do if this returns an error?
-    db_impl_->TraceIteratorSeek(cfd_->GetID(), target).PermitUncheckedError();
+    Slice lower_bound, upper_bound;
+    if (iterate_lower_bound_ != nullptr) {
+      lower_bound = *iterate_lower_bound_;
+    } else {
+      lower_bound = Slice("");
+    }
+    if (iterate_upper_bound_ != nullptr) {
+      upper_bound = *iterate_upper_bound_;
+    } else {
+      upper_bound = Slice("");
+    }
+    db_impl_->TraceIteratorSeek(cfd_->GetID(), target, lower_bound, upper_bound)
+        .PermitUncheckedError();
   }
 #endif  // ROCKSDB_LITE
 
@@ -1310,24 +1371,30 @@ void DBIter::Seek(const Slice& target) {
 }
 
 void DBIter::SeekForPrev(const Slice& target) {
-  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, env_);
-  StopWatch sw(env_, statistics_, DB_SEEK);
+  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, clock_);
+  StopWatch sw(clock_, statistics_, DB_SEEK);
 
 #ifndef ROCKSDB_LITE
   if (db_impl_ != nullptr && cfd_ != nullptr) {
     // TODO: What do we do if this returns an error?
-    db_impl_->TraceIteratorSeekForPrev(cfd_->GetID(), target)
+    Slice lower_bound, upper_bound;
+    if (iterate_lower_bound_ != nullptr) {
+      lower_bound = *iterate_lower_bound_;
+    } else {
+      lower_bound = Slice("");
+    }
+    if (iterate_upper_bound_ != nullptr) {
+      upper_bound = *iterate_upper_bound_;
+    } else {
+      upper_bound = Slice("");
+    }
+    db_impl_
+        ->TraceIteratorSeekForPrev(cfd_->GetID(), target, lower_bound,
+                                   upper_bound)
         .PermitUncheckedError();
   }
 #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();
@@ -1378,7 +1445,7 @@ void DBIter::SeekToFirst() {
     Seek(*iterate_lower_bound_);
     return;
   }
-  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, env_);
+  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, clock_);
   // Don't use iter_::Seek() if we set a prefix extractor
   // because prefix seek will be used.
   if (!expect_total_order_inner_iter()) {
@@ -1422,24 +1489,19 @@ 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_);
-    if (Valid() && user_comparator_.Equal(*iterate_upper_bound_, key())) {
+    if (Valid() && 0 == user_comparator_.CompareWithoutTimestamp(
+                            *iterate_upper_bound_, /*a_has_ts=*/false, key(),
+                            /*b_has_ts=*/false)) {
       ReleaseTempPinnedData();
       PrevInternal(nullptr);
     }
     return;
   }
 
-  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, env_);
+  PERF_CPU_TIMER_GUARD(iter_seek_cpu_nanos, clock_);
   // Don't use iter_::Seek() if we set a prefix extractor
   // because prefix seek will be used.
   if (!expect_total_order_inner_iter()) {

db/db_iter.h

@@ -8,8 +8,9 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 
 #pragma once
-#include <stdint.h>
+#include <cstdint>
 #include <string>
+
 #include "db/db_impl/db_impl.h"
 #include "db/dbformat.h"
 #include "db/range_del_aggregator.h"
@@ -21,7 +22,6 @@
 #include "util/autovector.h"
 
 namespace ROCKSDB_NAMESPACE {
-
 class Version;
 
 // This file declares the factory functions of DBIter, in its original form
@@ -68,7 +68,7 @@ class DBIter final : public Iterator {
   //        this->key().
   // (2) When moving backwards, the internal iterator is positioned
   //     just before all entries whose user key == this->key().
-  enum Direction { kForward, kReverse };
+  enum Direction : uint8_t { kForward, kReverse };
 
   // LocalStatistics contain Statistics counters that will be aggregated per
   // each iterator instance and then will be sent to the global statistics when
@@ -185,6 +185,9 @@ class DBIter final : public Iterator {
   Slice timestamp() const override {
     assert(valid_);
     assert(timestamp_size_ > 0);
+    if (direction_ == kReverse) {
+      return saved_timestamp_;
+    }
     const Slice ukey_and_ts = saved_key_.GetUserKey();
     assert(timestamp_size_ < ukey_and_ts.size());
     return ExtractTimestampFromUserKey(ukey_and_ts, timestamp_size_);
@@ -298,6 +301,7 @@ class DBIter final : public Iterator {
 
   const SliceTransform* prefix_extractor_;
   Env* const env_;
+  SystemClock* clock_;
   Logger* logger_;
   UserComparatorWrapper user_comparator_;
   const MergeOperator* const merge_operator_;
@@ -372,6 +376,7 @@ class DBIter final : public Iterator {
   const Slice* const timestamp_ub_;
   const Slice* const timestamp_lb_;
   const size_t timestamp_size_;
+  std::string saved_timestamp_;
 };
 
 // Return a new iterator that converts internal keys (yielded by

db/db_iterator_test.cc

@@ -18,6 +18,7 @@
 #include "rocksdb/perf_context.h"
 #include "table/block_based/flush_block_policy.h"
 #include "util/random.h"
+#include "utilities/merge_operators/string_append/stringappend2.h"
 
 namespace ROCKSDB_NAMESPACE {
 
@@ -619,6 +620,40 @@ TEST_P(DBIteratorTest, IterReseek) {
   delete iter;
 }
 
+TEST_F(DBIteratorTest, ReseekUponDirectionChange) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.prefix_extractor.reset(NewFixedPrefixTransform(1));
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
+  options.merge_operator.reset(
+      new StringAppendTESTOperator(/*delim_char=*/' '));
+  DestroyAndReopen(options);
+  ASSERT_OK(Put("foo", "value"));
+  ASSERT_OK(Put("bar", "value"));
+  {
+    std::unique_ptr<Iterator> it(db_->NewIterator(ReadOptions()));
+    it->SeekToLast();
+    it->Prev();
+    it->Next();
+  }
+  ASSERT_EQ(1,
+            options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
+
+  const std::string merge_key("good");
+  ASSERT_OK(Put(merge_key, "orig"));
+  ASSERT_OK(Merge(merge_key, "suffix"));
+  {
+    std::unique_ptr<Iterator> it(db_->NewIterator(ReadOptions()));
+    it->Seek(merge_key);
+    ASSERT_TRUE(it->Valid());
+    const uint64_t prev_reseek_count =
+        options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION);
+    it->Prev();
+    ASSERT_EQ(prev_reseek_count + 1, options.statistics->getTickerCount(
+                                         NUMBER_OF_RESEEKS_IN_ITERATION));
+  }
+}
+
 TEST_P(DBIteratorTest, IterSmallAndLargeMix) {
   do {
     CreateAndReopenWithCF({"pikachu"}, CurrentOptions());
@@ -1518,6 +1553,10 @@ TEST_P(DBIteratorTestForPinnedData, PinnedDataIteratorRandomizedFlush) {
   PinnedDataIteratorRandomized(TestConfig::FLUSH_EVERY_1000);
 }
 
+INSTANTIATE_TEST_CASE_P(DBIteratorTestForPinnedDataInstance,
+                        DBIteratorTestForPinnedData,
+                        testing::Values(true, false));
+
 #ifndef ROCKSDB_LITE
 TEST_P(DBIteratorTest, PinnedDataIteratorMultipleFiles) {
   Options options = CurrentOptions();
@@ -2997,18 +3036,18 @@ TEST_P(DBIteratorTest, Blob) {
   iter->Prev();
   ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
   iter->Next();
-  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 6);
   ASSERT_EQ(IterStatus(iter), "b->vb3");
 
   // Switch from forward to reverse
   iter->SeekToFirst();
-  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
-  iter->Next();
-  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 5);
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 6);
   iter->Next();
   ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 6);
-  iter->Prev();
+  iter->Next();
   ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 7);
+  iter->Prev();
+  ASSERT_EQ(TestGetTickerCount(options, NUMBER_OF_RESEEKS_IN_ITERATION), 8);
   ASSERT_EQ(IterStatus(iter), "b->vb3");
 }
 

db/db_kv_checksum_test.cc

@@ -0,0 +1,199 @@
+//  Copyright (c) 2020-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/db_test_util.h"
+#include "rocksdb/rocksdb_namespace.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+enum class WriteBatchOpType {
+  kPut = 0,
+  kDelete,
+  kSingleDelete,
+  kDeleteRange,
+  kMerge,
+  kBlobIndex,
+  kNum,
+};
+
+// Integer addition is needed for `::testing::Range()` to take the enum type.
+WriteBatchOpType operator+(WriteBatchOpType lhs, const int rhs) {
+  using T = std::underlying_type<WriteBatchOpType>::type;
+  return static_cast<WriteBatchOpType>(static_cast<T>(lhs) + rhs);
+}
+
+class DbKvChecksumTest
+    : public DBTestBase,
+      public ::testing::WithParamInterface<std::tuple<WriteBatchOpType, char>> {
+ public:
+  DbKvChecksumTest()
+      : DBTestBase("/db_kv_checksum_test", /*env_do_fsync=*/false) {
+    op_type_ = std::get<0>(GetParam());
+    corrupt_byte_addend_ = std::get<1>(GetParam());
+  }
+
+  std::pair<WriteBatch, Status> GetWriteBatch(size_t ts_sz,
+                                              ColumnFamilyHandle* cf_handle) {
+    Status s;
+    WriteBatch wb(0 /* reserved_bytes */, 0 /* max_bytes */, ts_sz,
+                  8 /* protection_bytes_per_entry */);
+    switch (op_type_) {
+      case WriteBatchOpType::kPut:
+        s = wb.Put(cf_handle, "key", "val");
+        break;
+      case WriteBatchOpType::kDelete:
+        s = wb.Delete(cf_handle, "key");
+        break;
+      case WriteBatchOpType::kSingleDelete:
+        s = wb.SingleDelete(cf_handle, "key");
+        break;
+      case WriteBatchOpType::kDeleteRange:
+        s = wb.DeleteRange(cf_handle, "begin", "end");
+        break;
+      case WriteBatchOpType::kMerge:
+        s = wb.Merge(cf_handle, "key", "val");
+        break;
+      case WriteBatchOpType::kBlobIndex:
+        // TODO(ajkr): use public API once available.
+        uint32_t cf_id;
+        if (cf_handle == nullptr) {
+          cf_id = 0;
+        } else {
+          cf_id = cf_handle->GetID();
+        }
+        s = WriteBatchInternal::PutBlobIndex(&wb, cf_id, "key", "val");
+        break;
+      case WriteBatchOpType::kNum:
+        assert(false);
+    }
+    return {std::move(wb), std::move(s)};
+  }
+
+  void CorruptNextByteCallBack(void* arg) {
+    Slice encoded = *static_cast<Slice*>(arg);
+    if (entry_len_ == port::kMaxSizet) {
+      // We learn the entry size on the first attempt
+      entry_len_ = encoded.size();
+    }
+    // All entries should be the same size
+    assert(entry_len_ == encoded.size());
+    char* buf = const_cast<char*>(encoded.data());
+    buf[corrupt_byte_offset_] += corrupt_byte_addend_;
+    ++corrupt_byte_offset_;
+  }
+
+  bool MoreBytesToCorrupt() { return corrupt_byte_offset_ < entry_len_; }
+
+ protected:
+  WriteBatchOpType op_type_;
+  char corrupt_byte_addend_;
+  size_t corrupt_byte_offset_ = 0;
+  size_t entry_len_ = port::kMaxSizet;
+};
+
+std::string GetTestNameSuffix(
+    ::testing::TestParamInfo<std::tuple<WriteBatchOpType, char>> info) {
+  std::ostringstream oss;
+  switch (std::get<0>(info.param)) {
+    case WriteBatchOpType::kPut:
+      oss << "Put";
+      break;
+    case WriteBatchOpType::kDelete:
+      oss << "Delete";
+      break;
+    case WriteBatchOpType::kSingleDelete:
+      oss << "SingleDelete";
+      break;
+    case WriteBatchOpType::kDeleteRange:
+      oss << "DeleteRange";
+      break;
+    case WriteBatchOpType::kMerge:
+      oss << "Merge";
+      break;
+    case WriteBatchOpType::kBlobIndex:
+      oss << "BlobIndex";
+      break;
+    case WriteBatchOpType::kNum:
+      assert(false);
+  }
+  oss << "Add"
+      << static_cast<int>(static_cast<unsigned char>(std::get<1>(info.param)));
+  return oss.str();
+}
+
+INSTANTIATE_TEST_CASE_P(
+    DbKvChecksumTest, DbKvChecksumTest,
+    ::testing::Combine(::testing::Range(static_cast<WriteBatchOpType>(0),
+                                        WriteBatchOpType::kNum),
+                       ::testing::Values(2, 103, 251)),
+    GetTestNameSuffix);
+
+TEST_P(DbKvChecksumTest, MemTableAddCorrupted) {
+  // This test repeatedly attempts to write `WriteBatch`es containing a single
+  // entry of type `op_type_`. Each attempt has one byte corrupted in its
+  // memtable entry by adding `corrupt_byte_addend_` to its original value. The
+  // test repeats until an attempt has been made on each byte in the encoded
+  // memtable entry. All attempts are expected to fail with `Status::Corruption`
+  SyncPoint::GetInstance()->SetCallBack(
+      "MemTable::Add:Encoded",
+      std::bind(&DbKvChecksumTest::CorruptNextByteCallBack, this,
+                std::placeholders::_1));
+
+  while (MoreBytesToCorrupt()) {
+    // Failed memtable insert always leads to read-only mode, so we have to
+    // reopen for every attempt.
+    Options options = CurrentOptions();
+    if (op_type_ == WriteBatchOpType::kMerge) {
+      options.merge_operator = MergeOperators::CreateStringAppendOperator();
+    }
+    Reopen(options);
+
+    SyncPoint::GetInstance()->EnableProcessing();
+    auto batch_and_status =
+        GetWriteBatch(0 /* ts_sz */, nullptr /* cf_handle */);
+    ASSERT_OK(batch_and_status.second);
+    ASSERT_TRUE(
+        db_->Write(WriteOptions(), &batch_and_status.first).IsCorruption());
+    SyncPoint::GetInstance()->DisableProcessing();
+  }
+}
+
+TEST_P(DbKvChecksumTest, MemTableAddWithColumnFamilyCorrupted) {
+  // This test repeatedly attempts to write `WriteBatch`es containing a single
+  // entry of type `op_type_` to a non-default column family. Each attempt has
+  // one byte corrupted in its memtable entry by adding `corrupt_byte_addend_`
+  // to its original value. The test repeats until an attempt has been made on
+  // each byte in the encoded memtable entry. All attempts are expected to fail
+  // with `Status::Corruption`.
+  Options options = CurrentOptions();
+  if (op_type_ == WriteBatchOpType::kMerge) {
+    options.merge_operator = MergeOperators::CreateStringAppendOperator();
+  }
+  CreateAndReopenWithCF({"pikachu"}, options);
+  SyncPoint::GetInstance()->SetCallBack(
+      "MemTable::Add:Encoded",
+      std::bind(&DbKvChecksumTest::CorruptNextByteCallBack, this,
+                std::placeholders::_1));
+
+  while (MoreBytesToCorrupt()) {
+    // Failed memtable insert always leads to read-only mode, so we have to
+    // reopen for every attempt.
+    ReopenWithColumnFamilies({kDefaultColumnFamilyName, "pikachu"}, options);
+
+    SyncPoint::GetInstance()->EnableProcessing();
+    auto batch_and_status = GetWriteBatch(0 /* ts_sz */, handles_[1]);
+    ASSERT_OK(batch_and_status.second);
+    ASSERT_TRUE(
+        db_->Write(WriteOptions(), &batch_and_status.first).IsCorruption());
+    SyncPoint::GetInstance()->DisableProcessing();
+  }
+}
+
+}  // namespace ROCKSDB_NAMESPACE
+
+int main(int argc, char** argv) {
+  ::testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}

db/db_log_iter_test.cc

@@ -13,6 +13,7 @@
 #if !defined(ROCKSDB_LITE)
 
 #include "db/db_test_util.h"
+#include "env/mock_env.h"
 #include "port/stack_trace.h"
 
 namespace ROCKSDB_NAMESPACE {

db/db_memtable_test.cc

@@ -145,15 +145,25 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
                                kMaxSequenceNumber, 0 /* column_family_id */);
 
   // Write some keys and make sure it returns false on duplicates
-  ASSERT_OK(mem->Add(seq, kTypeValue, "key", "value2"));
-  ASSERT_TRUE(mem->Add(seq, kTypeValue, "key", "value2").IsTryAgain());
+  ASSERT_OK(
+      mem->Add(seq, kTypeValue, "key", "value2", nullptr /* kv_prot_info */));
+  ASSERT_TRUE(
+      mem->Add(seq, kTypeValue, "key", "value2", nullptr /* kv_prot_info */)
+          .IsTryAgain());
   // Changing the type should still cause the duplicatae key
-  ASSERT_TRUE(mem->Add(seq, kTypeMerge, "key", "value2").IsTryAgain());
+  ASSERT_TRUE(
+      mem->Add(seq, kTypeMerge, "key", "value2", nullptr /* kv_prot_info */)
+          .IsTryAgain());
   // Changing the seq number will make the key fresh
-  ASSERT_OK(mem->Add(seq + 1, kTypeMerge, "key", "value2"));
+  ASSERT_OK(mem->Add(seq + 1, kTypeMerge, "key", "value2",
+                     nullptr /* kv_prot_info */));
   // Test with different types for duplicate keys
-  ASSERT_TRUE(mem->Add(seq, kTypeDeletion, "key", "").IsTryAgain());
-  ASSERT_TRUE(mem->Add(seq, kTypeSingleDeletion, "key", "").IsTryAgain());
+  ASSERT_TRUE(
+      mem->Add(seq, kTypeDeletion, "key", "", nullptr /* kv_prot_info */)
+          .IsTryAgain());
+  ASSERT_TRUE(
+      mem->Add(seq, kTypeSingleDeletion, "key", "", nullptr /* kv_prot_info */)
+          .IsTryAgain());
 
   // Test the duplicate keys under stress
   for (int i = 0; i < 10000; i++) {
@@ -161,7 +171,8 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
     if (!insert_dup) {
       seq++;
     }
-    Status s = mem->Add(seq, kTypeValue, "foo", "value" + ToString(seq));
+    Status s = mem->Add(seq, kTypeValue, "foo", "value" + ToString(seq),
+                        nullptr /* kv_prot_info */);
     if (insert_dup) {
       ASSERT_TRUE(s.IsTryAgain());
     } else {
@@ -177,8 +188,11 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
   mem = new MemTable(cmp, ioptions, MutableCFOptions(options), &wb,
                      kMaxSequenceNumber, 0 /* column_family_id */);
   // Insert a duplicate key with _ in it
-  ASSERT_OK(mem->Add(seq, kTypeValue, "key_1", "value"));
-  ASSERT_TRUE(mem->Add(seq, kTypeValue, "key_1", "value").IsTryAgain());
+  ASSERT_OK(
+      mem->Add(seq, kTypeValue, "key_1", "value", nullptr /* kv_prot_info */));
+  ASSERT_TRUE(
+      mem->Add(seq, kTypeValue, "key_1", "value", nullptr /* kv_prot_info */)
+          .IsTryAgain());
   delete mem;
 
   // Test when InsertConcurrently will be invoked
@@ -187,11 +201,11 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
   mem = new MemTable(cmp, ioptions, MutableCFOptions(options), &wb,
                      kMaxSequenceNumber, 0 /* column_family_id */);
   MemTablePostProcessInfo post_process_info;
-  ASSERT_OK(
-      mem->Add(seq, kTypeValue, "key", "value", true, &post_process_info));
-  ASSERT_TRUE(
-      mem->Add(seq, kTypeValue, "key", "value", true, &post_process_info)
-          .IsTryAgain());
+  ASSERT_OK(mem->Add(seq, kTypeValue, "key", "value",
+                     nullptr /* kv_prot_info */, true, &post_process_info));
+  ASSERT_TRUE(mem->Add(seq, kTypeValue, "key", "value",
+                       nullptr /* kv_prot_info */, true, &post_process_info)
+                  .IsTryAgain());
   delete mem;
 }
 
@@ -217,7 +231,7 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
 
   // Put 0 as the base
   PutFixed64(&value, static_cast<uint64_t>(0));
-  ASSERT_OK(mem->Add(0, kTypeValue, "key", value));
+  ASSERT_OK(mem->Add(0, kTypeValue, "key", value, nullptr /* kv_prot_info */));
   value.clear();
 
   // Write Merge concurrently
@@ -226,8 +240,8 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
     std::string v1;
     for (int seq = 1; seq < num_ops / 2; seq++) {
       PutFixed64(&v1, seq);
-      ASSERT_OK(
-          mem->Add(seq, kTypeMerge, "key", v1, true, &post_process_info1));
+      ASSERT_OK(mem->Add(seq, kTypeMerge, "key", v1, nullptr /* kv_prot_info */,
+                         true, &post_process_info1));
       v1.clear();
     }
   });
@@ -236,8 +250,8 @@ TEST_F(DBMemTableTest, ConcurrentMergeWrite) {
     std::string v2;
     for (int seq = num_ops / 2; seq < num_ops; seq++) {
       PutFixed64(&v2, seq);
-      ASSERT_OK(
-          mem->Add(seq, kTypeMerge, "key", v2, true, &post_process_info2));
+      ASSERT_OK(mem->Add(seq, kTypeMerge, "key", v2, nullptr /* kv_prot_info */,
+                         true, &post_process_info2));
       v2.clear();
     }
   });

db/db_options_test.cc

@@ -129,6 +129,83 @@ TEST_F(DBOptionsTest, GetLatestCFOptions) {
             GetMutableCFOptionsMap(dbfull()->GetOptions(handles_[1])));
 }
 
+TEST_F(DBOptionsTest, SetMutableTableOptions) {
+  Options options;
+  options.create_if_missing = true;
+  options.env = env_;
+  options.blob_file_size = 16384;
+  BlockBasedTableOptions bbto;
+  bbto.no_block_cache = true;
+  bbto.block_size = 8192;
+  bbto.block_restart_interval = 7;
+
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  Reopen(options);
+
+  ColumnFamilyHandle* cfh = dbfull()->DefaultColumnFamily();
+  Options c_opts = dbfull()->GetOptions(cfh);
+  const auto* c_bbto =
+      c_opts.table_factory->GetOptions<BlockBasedTableOptions>();
+  ASSERT_NE(c_bbto, nullptr);
+  ASSERT_EQ(c_opts.blob_file_size, 16384);
+  ASSERT_EQ(c_bbto->no_block_cache, true);
+  ASSERT_EQ(c_bbto->block_size, 8192);
+  ASSERT_EQ(c_bbto->block_restart_interval, 7);
+  ASSERT_OK(dbfull()->SetOptions(
+      cfh, {{"table_factory.block_size", "16384"},
+            {"table_factory.block_restart_interval", "11"}}));
+  ASSERT_EQ(c_bbto->block_size, 16384);
+  ASSERT_EQ(c_bbto->block_restart_interval, 11);
+
+  // Now set an option that is not mutable - options should not change
+  ASSERT_NOK(
+      dbfull()->SetOptions(cfh, {{"table_factory.no_block_cache", "false"}}));
+  ASSERT_EQ(c_bbto->no_block_cache, true);
+  ASSERT_EQ(c_bbto->block_size, 16384);
+  ASSERT_EQ(c_bbto->block_restart_interval, 11);
+
+  // Set some that are mutable and some that are not - options should not change
+  ASSERT_NOK(dbfull()->SetOptions(
+      cfh, {{"table_factory.no_block_cache", "false"},
+            {"table_factory.block_size", "8192"},
+            {"table_factory.block_restart_interval", "7"}}));
+  ASSERT_EQ(c_bbto->no_block_cache, true);
+  ASSERT_EQ(c_bbto->block_size, 16384);
+  ASSERT_EQ(c_bbto->block_restart_interval, 11);
+
+  // Set some that are mutable and some that do not exist - options should not
+  // change
+  ASSERT_NOK(dbfull()->SetOptions(
+      cfh, {{"table_factory.block_size", "8192"},
+            {"table_factory.does_not_exist", "true"},
+            {"table_factory.block_restart_interval", "7"}}));
+  ASSERT_EQ(c_bbto->no_block_cache, true);
+  ASSERT_EQ(c_bbto->block_size, 16384);
+  ASSERT_EQ(c_bbto->block_restart_interval, 11);
+
+  // Trying to change the table factory fails
+  ASSERT_NOK(dbfull()->SetOptions(
+      cfh, {{"table_factory", TableFactory::kPlainTableName()}}));
+
+  // Set some on the table and some on the Column Family
+  ASSERT_OK(dbfull()->SetOptions(
+      cfh, {{"table_factory.block_size", "16384"},
+            {"blob_file_size", "32768"},
+            {"table_factory.block_restart_interval", "13"}}));
+  c_opts = dbfull()->GetOptions(cfh);
+  ASSERT_EQ(c_opts.blob_file_size, 32768);
+  ASSERT_EQ(c_bbto->block_size, 16384);
+  ASSERT_EQ(c_bbto->block_restart_interval, 13);
+  // Set some on the table and a bad one on the ColumnFamily - options should
+  // not change
+  ASSERT_NOK(dbfull()->SetOptions(
+      cfh, {{"table_factory.block_size", "1024"},
+            {"no_such_option", "32768"},
+            {"table_factory.block_restart_interval", "7"}}));
+  ASSERT_EQ(c_bbto->block_size, 16384);
+  ASSERT_EQ(c_bbto->block_restart_interval, 13);
+}
+
 TEST_F(DBOptionsTest, SetBytesPerSync) {
   const size_t kValueSize = 1024 * 1024;  // 1MB
   Options options;

db/db_range_del_test.cc

@@ -73,6 +73,15 @@ TEST_F(DBRangeDelTest, FlushOutputHasOnlyRangeTombstones) {
   } while (ChangeOptions(kRangeDelSkipConfigs));
 }
 
+TEST_F(DBRangeDelTest, DictionaryCompressionWithOnlyRangeTombstones) {
+  Options opts = CurrentOptions();
+  opts.compression_opts.max_dict_bytes = 16384;
+  Reopen(opts);
+  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr1",
+                             "dr2"));
+  ASSERT_OK(db_->Flush(FlushOptions()));
+}
+
 TEST_F(DBRangeDelTest, CompactionOutputHasOnlyRangeTombstone) {
   do {
     Options opts = CurrentOptions();

db/db_secondary_test.cc

@@ -459,20 +459,6 @@ TEST_F(DBSecondaryTest, MissingTableFileDuringOpen) {
 }
 
 TEST_F(DBSecondaryTest, MissingTableFile) {
-  int table_files_not_exist = 0;
-  SyncPoint::GetInstance()->DisableProcessing();
-  SyncPoint::GetInstance()->ClearAllCallBacks();
-  SyncPoint::GetInstance()->SetCallBack(
-      "ReactiveVersionSet::ApplyOneVersionEditToBuilder:AfterLoadTableHandlers",
-      [&](void* arg) {
-        Status s = *reinterpret_cast<Status*>(arg);
-        if (s.IsPathNotFound()) {
-          ++table_files_not_exist;
-        } else if (!s.ok()) {
-          assert(false);  // Should not reach here
-        }
-      });
-  SyncPoint::GetInstance()->EnableProcessing();
   Options options;
   options.env = env_;
   options.level0_file_num_compaction_trigger = 4;
@@ -499,7 +485,6 @@ TEST_F(DBSecondaryTest, MissingTableFile) {
   ASSERT_NOK(db_secondary_->Get(ropts, "bar", &value));
 
   ASSERT_OK(db_secondary_->TryCatchUpWithPrimary());
-  ASSERT_EQ(options.level0_file_num_compaction_trigger, table_files_not_exist);
   ASSERT_OK(db_secondary_->Get(ropts, "foo", &value));
   ASSERT_EQ("foo_value" +
                 std::to_string(options.level0_file_num_compaction_trigger - 1),
@@ -615,10 +600,7 @@ TEST_F(DBSecondaryTest, SwitchManifest) {
   range_scan_db();
 }
 
-// Here, "Snapshot" refers to the version edits written by
-// VersionSet::WriteSnapshot() at the beginning of the new MANIFEST after
-// switching from the old one.
-TEST_F(DBSecondaryTest, SkipSnapshotAfterManifestSwitch) {
+TEST_F(DBSecondaryTest, SwitchManifestTwice) {
   Options options;
   options.env = env_;
   options.disable_auto_compactions = true;
@@ -640,10 +622,15 @@ TEST_F(DBSecondaryTest, SkipSnapshotAfterManifestSwitch) {
 
   Reopen(options);
   ASSERT_OK(dbfull()->SetOptions({{"disable_auto_compactions", "false"}}));
+  Reopen(options);
+  ASSERT_OK(Put("0", "value1"));
   ASSERT_OK(db_secondary_->TryCatchUpWithPrimary());
+
+  ASSERT_OK(db_secondary_->Get(ropts, "0", &value));
+  ASSERT_EQ("value1", value);
 }
 
-TEST_F(DBSecondaryTest, SwitchWAL) {
+TEST_F(DBSecondaryTest, DISABLED_SwitchWAL) {
   const int kNumKeysPerMemtable = 1;
   Options options;
   options.env = env_;
@@ -692,7 +679,7 @@ TEST_F(DBSecondaryTest, SwitchWAL) {
   }
 }
 
-TEST_F(DBSecondaryTest, SwitchWALMultiColumnFamilies) {
+TEST_F(DBSecondaryTest, DISABLED_SwitchWALMultiColumnFamilies) {
   const int kNumKeysPerMemtable = 1;
   SyncPoint::GetInstance()->DisableProcessing();
   SyncPoint::GetInstance()->LoadDependency(

db/db_sst_test.cc

@@ -306,13 +306,13 @@ TEST_F(DBSSTTest, DBWithSstFileManager) {
     ASSERT_OK(dbfull()->TEST_WaitForCompact());
     // Verify that we are tracking all sst files in dbname_
     std::unordered_map<std::string, uint64_t> files_in_db;
-    ASSERT_OK(GetAllSSTFiles(&files_in_db));
+    ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db));
     ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
   }
   ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
 
   std::unordered_map<std::string, uint64_t> files_in_db;
-  ASSERT_OK(GetAllSSTFiles(&files_in_db));
+  ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db));
   // Verify that we are tracking all sst files in dbname_
   ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
   // Verify the total files size
@@ -346,6 +346,260 @@ TEST_F(DBSSTTest, DBWithSstFileManager) {
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
 }
 
+TEST_F(DBSSTTest, DBWithSstFileManagerForBlobFiles) {
+  std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
+  auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
+
+  int files_added = 0;
+  int files_deleted = 0;
+  int files_moved = 0;
+  int files_scheduled_to_delete = 0;
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::OnAddFile", [&](void* arg) {
+        const std::string* const file_path =
+            static_cast<const std::string*>(arg);
+        if (file_path->find(".blob") != std::string::npos) {
+          files_added++;
+        }
+      });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::OnDeleteFile", [&](void* arg) {
+        const std::string* const file_path =
+            static_cast<const std::string*>(arg);
+        if (file_path->find(".blob") != std::string::npos) {
+          files_deleted++;
+        }
+      });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::ScheduleFileDeletion", [&](void* arg) {
+        assert(arg);
+        const std::string* const file_path =
+            static_cast<const std::string*>(arg);
+        if (file_path->find(".blob") != std::string::npos) {
+          ++files_scheduled_to_delete;
+        }
+      });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::OnMoveFile", [&](void* /*arg*/) { files_moved++; });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  Options options = CurrentOptions();
+  options.sst_file_manager = sst_file_manager;
+  options.enable_blob_files = true;
+  options.blob_file_size = 32;  // create one blob per file
+  DestroyAndReopen(options);
+  Random rnd(301);
+
+  for (int i = 0; i < 10; i++) {
+    ASSERT_OK(Put("Key_" + std::to_string(i), "Value_" + std::to_string(i)));
+    ASSERT_OK(Flush());
+    ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
+    ASSERT_OK(dbfull()->TEST_WaitForCompact());
+
+    // Verify that we are tracking all sst and blob files in dbname_
+    std::unordered_map<std::string, uint64_t> files_in_db;
+    ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db));
+    ASSERT_OK(GetAllDataFiles(kBlobFile, &files_in_db));
+    ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
+  }
+
+  std::vector<uint64_t> blob_files = GetBlobFileNumbers();
+  ASSERT_EQ(files_added, blob_files.size());
+  // No blob file is obsoleted.
+  ASSERT_EQ(files_deleted, 0);
+  ASSERT_EQ(files_scheduled_to_delete, 0);
+  // No files were moved.
+  ASSERT_EQ(files_moved, 0);
+
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
+
+  std::unordered_map<std::string, uint64_t> files_in_db;
+  ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db));
+  ASSERT_OK(GetAllDataFiles(kBlobFile, &files_in_db));
+
+  // Verify that we are tracking all sst and blob files in dbname_
+  ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
+  // Verify the total files size
+  uint64_t total_files_size = 0;
+  for (auto& file_to_size : files_in_db) {
+    total_files_size += file_to_size.second;
+  }
+  ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
+  Close();
+
+  Reopen(options);
+  ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
+  ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
+
+  // Verify that we track all the files again after the DB is closed and opened.
+  Close();
+
+  sst_file_manager.reset(NewSstFileManager(env_));
+  options.sst_file_manager = sst_file_manager;
+  sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
+
+  Reopen(options);
+
+  ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
+  ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
+
+  // Destroy DB and it will remove all the blob files from sst file manager and
+  // blob files deletion will go through ScheduleFileDeletion.
+  ASSERT_EQ(files_deleted, 0);
+  ASSERT_EQ(files_scheduled_to_delete, 0);
+  Close();
+  ASSERT_OK(DestroyDB(dbname_, options));
+  ASSERT_EQ(files_deleted, blob_files.size());
+  ASSERT_EQ(files_scheduled_to_delete, blob_files.size());
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
+TEST_F(DBSSTTest, DBWithSstFileManagerForBlobFilesWithGC) {
+  std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
+  auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
+  Options options = CurrentOptions();
+  options.sst_file_manager = sst_file_manager;
+  options.enable_blob_files = true;
+  options.blob_file_size = 32;  // create one blob per file
+  options.disable_auto_compactions = true;
+  options.enable_blob_garbage_collection = true;
+  options.blob_garbage_collection_age_cutoff = 0.5;
+
+  int files_added = 0;
+  int files_deleted = 0;
+  int files_moved = 0;
+  int files_scheduled_to_delete = 0;
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::OnAddFile", [&](void* arg) {
+        const std::string* const file_path =
+            static_cast<const std::string*>(arg);
+        if (file_path->find(".blob") != std::string::npos) {
+          files_added++;
+        }
+      });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::OnDeleteFile", [&](void* arg) {
+        const std::string* const file_path =
+            static_cast<const std::string*>(arg);
+        if (file_path->find(".blob") != std::string::npos) {
+          files_deleted++;
+        }
+      });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::ScheduleFileDeletion", [&](void* arg) {
+        assert(arg);
+        const std::string* const file_path =
+            static_cast<const std::string*>(arg);
+        if (file_path->find(".blob") != std::string::npos) {
+          ++files_scheduled_to_delete;
+        }
+      });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "SstFileManagerImpl::OnMoveFile", [&](void* /*arg*/) { files_moved++; });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  DestroyAndReopen(options);
+  Random rnd(301);
+
+  constexpr char first_key[] = "first_key";
+  constexpr char first_value[] = "first_value";
+  constexpr char second_key[] = "second_key";
+  constexpr char second_value[] = "second_value";
+
+  ASSERT_OK(Put(first_key, first_value));
+  ASSERT_OK(Put(second_key, second_value));
+  ASSERT_OK(Flush());
+
+  constexpr char third_key[] = "third_key";
+  constexpr char third_value[] = "third_value";
+  constexpr char fourth_key[] = "fourth_key";
+  constexpr char fourth_value[] = "fourth_value";
+  constexpr char fifth_key[] = "fifth_key";
+  constexpr char fifth_value[] = "fifth_value";
+
+  ASSERT_OK(Put(third_key, third_value));
+  ASSERT_OK(Put(fourth_key, fourth_value));
+  ASSERT_OK(Put(fifth_key, fifth_value));
+  ASSERT_OK(Flush());
+
+  const std::vector<uint64_t> original_blob_files = GetBlobFileNumbers();
+
+  ASSERT_EQ(original_blob_files.size(), 5);
+  ASSERT_EQ(files_added, 5);
+  ASSERT_EQ(files_deleted, 0);
+  ASSERT_EQ(files_scheduled_to_delete, 0);
+  ASSERT_EQ(files_moved, 0);
+  {
+    // Verify that we are tracking all sst and blob files in dbname_
+    std::unordered_map<std::string, uint64_t> files_in_db;
+    ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db));
+    ASSERT_OK(GetAllDataFiles(kBlobFile, &files_in_db));
+    ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
+  }
+
+  const size_t cutoff_index = static_cast<size_t>(
+      options.blob_garbage_collection_age_cutoff * original_blob_files.size());
+
+  size_t expected_number_of_files = original_blob_files.size();
+  // Note: turning off enable_blob_files before the compaction results in
+  // garbage collected values getting inlined.
+  ASSERT_OK(db_->SetOptions({{"enable_blob_files", "false"}}));
+  expected_number_of_files -= cutoff_index;
+  files_added = 0;
+
+  constexpr Slice* begin = nullptr;
+  constexpr Slice* end = nullptr;
+
+  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), begin, end));
+
+  ASSERT_EQ(Get(first_key), first_value);
+  ASSERT_EQ(Get(second_key), second_value);
+  ASSERT_EQ(Get(third_key), third_value);
+  ASSERT_EQ(Get(fourth_key), fourth_value);
+  ASSERT_EQ(Get(fifth_key), fifth_value);
+
+  const std::vector<uint64_t> new_blob_files = GetBlobFileNumbers();
+
+  ASSERT_EQ(new_blob_files.size(), expected_number_of_files);
+  // No new file is added.
+  ASSERT_EQ(files_added, 0);
+  ASSERT_EQ(files_deleted, cutoff_index);
+  ASSERT_EQ(files_scheduled_to_delete, cutoff_index);
+  ASSERT_EQ(files_moved, 0);
+
+  // Original blob files below the cutoff should be gone, original blob files at
+  // or above the cutoff should be still there
+  for (size_t i = cutoff_index; i < original_blob_files.size(); ++i) {
+    ASSERT_EQ(new_blob_files[i - cutoff_index], original_blob_files[i]);
+  }
+
+  {
+    // Verify that we are tracking all sst and blob files in dbname_
+    std::unordered_map<std::string, uint64_t> files_in_db;
+    ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db));
+    ASSERT_OK(GetAllDataFiles(kBlobFile, &files_in_db));
+    ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
+  }
+
+  Close();
+  ASSERT_OK(DestroyDB(dbname_, options));
+  ASSERT_EQ(files_deleted, 5);
+  ASSERT_EQ(files_scheduled_to_delete, 5);
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
+}
+
 TEST_F(DBSSTTest, RateLimitedDelete) {
   Destroy(last_options_);
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency({
@@ -780,7 +1034,7 @@ TEST_F(DBSSTTest, DBWithMaxSpaceAllowed) {
 
   uint64_t first_file_size = 0;
   std::unordered_map<std::string, uint64_t> files_in_db;
-  ASSERT_OK(GetAllSSTFiles(&files_in_db, &first_file_size));
+  ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db, &first_file_size));
   ASSERT_EQ(sfm->GetTotalSize(), first_file_size);
 
   // Set the maximum allowed space usage to the current total size
@@ -791,6 +1045,60 @@ TEST_F(DBSSTTest, DBWithMaxSpaceAllowed) {
   ASSERT_NOK(Flush());
 }
 
+TEST_F(DBSSTTest, DBWithMaxSpaceAllowedWithBlobFiles) {
+  std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
+  auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
+
+  Options options = CurrentOptions();
+  options.sst_file_manager = sst_file_manager;
+  options.disable_auto_compactions = true;
+  options.enable_blob_files = true;
+  DestroyAndReopen(options);
+
+  Random rnd(301);
+
+  // Generate a file containing keys.
+  for (int i = 0; i < 10; i++) {
+    ASSERT_OK(Put(Key(i), rnd.RandomString(50)));
+  }
+  ASSERT_OK(Flush());
+
+  uint64_t files_size = 0;
+  uint64_t total_files_size = 0;
+  std::unordered_map<std::string, uint64_t> files_in_db;
+
+  ASSERT_OK(GetAllDataFiles(kBlobFile, &files_in_db, &files_size));
+  // Make sure blob files are considered by SSTFileManage in size limits.
+  ASSERT_GT(files_size, 0);
+  total_files_size = files_size;
+  ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db, &files_size));
+  total_files_size += files_size;
+  ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
+
+  // Set the maximum allowed space usage to the current total size.
+  sfm->SetMaxAllowedSpaceUsage(files_size + 1);
+
+  bool max_allowed_space_reached = false;
+  bool delete_blob_file = false;
+  // Sync point called after blob file is closed and max allowed space is
+  // checked.
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BlobFileCompletionCallback::CallBack::MaxAllowedSpaceReached",
+      [&](void* /*arg*/) { max_allowed_space_reached = true; });
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->SetCallBack(
+      "BuildTable::AfterDeleteFile",
+      [&](void* /*arg*/) { delete_blob_file = true; });
+
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  ASSERT_OK(Put("key1", "val1"));
+  // This flush will fail
+  ASSERT_NOK(Flush());
+  ASSERT_TRUE(max_allowed_space_reached);
+  ASSERT_TRUE(delete_blob_file);
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+}
+
 TEST_F(DBSSTTest, CancellingCompactionsWorks) {
   std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
   auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
@@ -821,7 +1129,7 @@ TEST_F(DBSSTTest, CancellingCompactionsWorks) {
   ASSERT_OK(Flush());
   uint64_t total_file_size = 0;
   std::unordered_map<std::string, uint64_t> files_in_db;
-  ASSERT_OK(GetAllSSTFiles(&files_in_db, &total_file_size));
+  ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db, &total_file_size));
   // Set the maximum allowed space usage to the current total size
   sfm->SetMaxAllowedSpaceUsage(2 * total_file_size + 1);
 
@@ -869,7 +1177,7 @@ TEST_F(DBSSTTest, CancellingManualCompactionsWorks) {
   ASSERT_OK(Flush());
   uint64_t total_file_size = 0;
   std::unordered_map<std::string, uint64_t> files_in_db;
-  ASSERT_OK(GetAllSSTFiles(&files_in_db, &total_file_size));
+  ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db, &total_file_size));
   // Set the maximum allowed space usage to the current total size
   sfm->SetMaxAllowedSpaceUsage(2 * total_file_size + 1);
 
@@ -986,7 +1294,7 @@ TEST_F(DBSSTTest, DBWithMaxSpaceAllowedRandomized) {
     ASSERT_TRUE(bg_error_set);
     uint64_t total_sst_files_size = 0;
     std::unordered_map<std::string, uint64_t> files_in_db;
-    ASSERT_OK(GetAllSSTFiles(&files_in_db, &total_sst_files_size));
+    ASSERT_OK(GetAllDataFiles(kTableFile, &files_in_db, &total_sst_files_size));
     ASSERT_GE(total_sst_files_size, limit_mb * 1024 * 1024);
     ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
   }

db/db_test.cc

@@ -247,17 +247,21 @@ TEST_F(DBTest, SkipDelay) {
       wo.sync = sync;
       wo.disableWAL = disableWAL;
       wo.no_slowdown = true;
-      dbfull()->Put(wo, "foo", "bar");
+      // Large enough to exceed allowance for one time interval
+      std::string large_value(1024, 'x');
+      // Perhaps ideally this first write would fail because of delay, but
+      // the current implementation does not guarantee that.
+      dbfull()->Put(wo, "foo", large_value).PermitUncheckedError();
       // We need the 2nd write to trigger delay. This is because delay is
       // estimated based on the last write size which is 0 for the first write.
-      ASSERT_NOK(dbfull()->Put(wo, "foo2", "bar2"));
+      ASSERT_NOK(dbfull()->Put(wo, "foo2", large_value));
       ASSERT_GE(sleep_count.load(), 0);
       ASSERT_GE(wait_count.load(), 0);
       token.reset();
 
-      token = dbfull()->TEST_write_controler().GetDelayToken(1000000000);
+      token = dbfull()->TEST_write_controler().GetDelayToken(1000000);
       wo.no_slowdown = false;
-      ASSERT_OK(dbfull()->Put(wo, "foo3", "bar3"));
+      ASSERT_OK(dbfull()->Put(wo, "foo3", large_value));
       ASSERT_GE(sleep_count.load(), 1);
       token.reset();
     }
@@ -2332,6 +2336,13 @@ TEST_F(DBTest, ReadonlyDBGetLiveManifestSize) {
 }
 
 TEST_F(DBTest, GetLiveBlobFiles) {
+  // Note: the following prevents an otherwise harmless data race between the
+  // test setup code (AddBlobFile) below and the periodic stat dumping thread.
+  Options options = CurrentOptions();
+  options.stats_dump_period_sec = 0;
+
+  Reopen(options);
+
   VersionSet* const versions = dbfull()->TEST_GetVersionSet();
   assert(versions);
   assert(versions->GetColumnFamilySet());

db/db_test2.cc

@@ -154,8 +154,14 @@ class PartitionedIndexTestListener : public EventListener {
 };
 
 TEST_F(DBTest2, PartitionedIndexUserToInternalKey) {
+  const int kValueSize = 10500;
+  const int kNumEntriesPerFile = 1000;
+  const int kNumFiles = 3;
+  const int kNumDistinctKeys = 30;
+
   BlockBasedTableOptions table_options;
   Options options = CurrentOptions();
+  options.disable_auto_compactions = true;
   table_options.index_type = BlockBasedTableOptions::kTwoLevelIndexSearch;
   PartitionedIndexTestListener* listener = new PartitionedIndexTestListener();
   options.table_factory.reset(NewBlockBasedTableFactory(table_options));
@@ -164,14 +170,15 @@ TEST_F(DBTest2, PartitionedIndexUserToInternalKey) {
   Reopen(options);
   Random rnd(301);
 
-  for (int i = 0; i < 3000; i++) {
-    int j = i % 30;
-    std::string value = rnd.RandomString(10500);
-    ASSERT_OK(Put("keykey_" + std::to_string(j), value));
-    snapshots.push_back(db_->GetSnapshot());
+  for (int i = 0; i < kNumFiles; i++) {
+    for (int j = 0; j < kNumEntriesPerFile; j++) {
+      int key_id = (i * kNumEntriesPerFile + j) % kNumDistinctKeys;
+      std::string value = rnd.RandomString(kValueSize);
+      ASSERT_OK(Put("keykey_" + std::to_string(key_id), value));
+      snapshots.push_back(db_->GetSnapshot());
+    }
+    ASSERT_OK(Flush());
   }
-  ASSERT_OK(Flush());
-  dbfull()->TEST_WaitForFlushMemTable();
 
   for (auto s : snapshots) {
     db_->ReleaseSnapshot(s);
@@ -1410,67 +1417,94 @@ INSTANTIATE_TEST_CASE_P(
 
 TEST_P(PresetCompressionDictTest, Flush) {
   // Verifies that dictionary is generated and written during flush only when
-  // `ColumnFamilyOptions::compression` enables dictionary.
+  // `ColumnFamilyOptions::compression` enables dictionary. Also verifies the
+  // size of the dictionary is within expectations according to the limit on
+  // buffering set by `CompressionOptions::max_dict_buffer_bytes`.
   const size_t kValueLen = 256;
   const size_t kKeysPerFile = 1 << 10;
-  const size_t kDictLen = 4 << 10;
+  const size_t kDictLen = 16 << 10;
+  const size_t kBlockLen = 4 << 10;
 
   Options options = CurrentOptions();
   if (bottommost_) {
     options.bottommost_compression = compression_type_;
     options.bottommost_compression_opts.enabled = true;
     options.bottommost_compression_opts.max_dict_bytes = kDictLen;
+    options.bottommost_compression_opts.max_dict_buffer_bytes = kBlockLen;
   } else {
     options.compression = compression_type_;
     options.compression_opts.max_dict_bytes = kDictLen;
+    options.compression_opts.max_dict_buffer_bytes = kBlockLen;
   }
   options.memtable_factory.reset(new SpecialSkipListFactory(kKeysPerFile));
   options.statistics = CreateDBStatistics();
   BlockBasedTableOptions bbto;
+  bbto.block_size = kBlockLen;
   bbto.cache_index_and_filter_blocks = true;
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   Reopen(options);
 
-  uint64_t prev_compression_dict_misses =
-      TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_MISS);
   Random rnd(301);
   for (size_t i = 0; i <= kKeysPerFile; ++i) {
     ASSERT_OK(Put(Key(static_cast<int>(i)), rnd.RandomString(kValueLen)));
   }
   ASSERT_OK(dbfull()->TEST_WaitForFlushMemTable());
 
-  // If there's a compression dictionary, it should have been loaded when the
-  // flush finished, incurring a cache miss.
-  uint64_t expected_compression_dict_misses;
+  // We can use `BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT` to detect whether a
+  // compression dictionary exists since dictionaries would be preloaded when
+  // the flush finishes.
   if (bottommost_) {
-    expected_compression_dict_misses = prev_compression_dict_misses;
+    // Flush is never considered bottommost. This should change in the future
+    // since flushed files may have nothing underneath them, like the one in
+    // this test case.
+    ASSERT_EQ(
+        TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+        0);
   } else {
-    expected_compression_dict_misses = prev_compression_dict_misses + 1;
+    ASSERT_GT(
+        TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+        0);
+    // TODO(ajkr): fix the below assertion to work with ZSTD. The expectation on
+    // number of bytes needs to be adjusted in case the cached block is in
+    // ZSTD's digested dictionary format.
+    if (compression_type_ != kZSTD &&
+        compression_type_ != kZSTDNotFinalCompression) {
+      // Although we limited buffering to `kBlockLen`, there may be up to two
+      // blocks of data included in the dictionary since we only check limit
+      // after each block is built.
+      ASSERT_LE(TestGetTickerCount(options,
+                                   BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+                2 * kBlockLen);
+    }
   }
-  ASSERT_EQ(expected_compression_dict_misses,
-            TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_MISS));
 }
 
 TEST_P(PresetCompressionDictTest, CompactNonBottommost) {
   // Verifies that dictionary is generated and written during compaction to
   // non-bottommost level only when `ColumnFamilyOptions::compression` enables
-  // dictionary.
+  // dictionary. Also verifies the size of the dictionary is within expectations
+  // according to the limit on buffering set by
+  // `CompressionOptions::max_dict_buffer_bytes`.
   const size_t kValueLen = 256;
   const size_t kKeysPerFile = 1 << 10;
-  const size_t kDictLen = 4 << 10;
+  const size_t kDictLen = 16 << 10;
+  const size_t kBlockLen = 4 << 10;
 
   Options options = CurrentOptions();
   if (bottommost_) {
     options.bottommost_compression = compression_type_;
     options.bottommost_compression_opts.enabled = true;
     options.bottommost_compression_opts.max_dict_bytes = kDictLen;
+    options.bottommost_compression_opts.max_dict_buffer_bytes = kBlockLen;
   } else {
     options.compression = compression_type_;
     options.compression_opts.max_dict_bytes = kDictLen;
+    options.compression_opts.max_dict_buffer_bytes = kBlockLen;
   }
   options.disable_auto_compactions = true;
   options.statistics = CreateDBStatistics();
   BlockBasedTableOptions bbto;
+  bbto.block_size = kBlockLen;
   bbto.cache_index_and_filter_blocks = true;
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   Reopen(options);
@@ -1492,8 +1526,8 @@ TEST_P(PresetCompressionDictTest, CompactNonBottommost) {
   ASSERT_EQ("2,0,1", FilesPerLevel(0));
 #endif  // ROCKSDB_LITE
 
-  uint64_t prev_compression_dict_misses =
-      TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_MISS);
+  uint64_t prev_compression_dict_bytes_inserted =
+      TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT);
   // This L0->L1 compaction merges the two L0 files into L1. The produced L1
   // file is not bottommost due to the existing L2 file covering the same key-
   // range.
@@ -1501,38 +1535,58 @@ TEST_P(PresetCompressionDictTest, CompactNonBottommost) {
 #ifndef ROCKSDB_LITE
   ASSERT_EQ("0,1,1", FilesPerLevel(0));
 #endif  // ROCKSDB_LITE
-  // If there's a compression dictionary, it should have been loaded when the
-  // compaction finished, incurring a cache miss.
-  uint64_t expected_compression_dict_misses;
+  // We can use `BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT` to detect whether a
+  // compression dictionary exists since dictionaries would be preloaded when
+  // the compaction finishes.
   if (bottommost_) {
-    expected_compression_dict_misses = prev_compression_dict_misses;
+    ASSERT_EQ(
+        TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+        prev_compression_dict_bytes_inserted);
   } else {
-    expected_compression_dict_misses = prev_compression_dict_misses + 1;
+    ASSERT_GT(
+        TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+        prev_compression_dict_bytes_inserted);
+    // TODO(ajkr): fix the below assertion to work with ZSTD. The expectation on
+    // number of bytes needs to be adjusted in case the cached block is in
+    // ZSTD's digested dictionary format.
+    if (compression_type_ != kZSTD &&
+        compression_type_ != kZSTDNotFinalCompression) {
+      // Although we limited buffering to `kBlockLen`, there may be up to two
+      // blocks of data included in the dictionary since we only check limit
+      // after each block is built.
+      ASSERT_LE(TestGetTickerCount(options,
+                                   BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+                prev_compression_dict_bytes_inserted + 2 * kBlockLen);
+    }
   }
-  ASSERT_EQ(expected_compression_dict_misses,
-            TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_MISS));
 }
 
 TEST_P(PresetCompressionDictTest, CompactBottommost) {
   // Verifies that dictionary is generated and written during compaction to
   // non-bottommost level only when either `ColumnFamilyOptions::compression` or
-  // `ColumnFamilyOptions::bottommost_compression` enables dictionary.
+  // `ColumnFamilyOptions::bottommost_compression` enables dictionary. Also
+  // verifies the size of the dictionary is within expectations according to the
+  // limit on buffering set by `CompressionOptions::max_dict_buffer_bytes`.
   const size_t kValueLen = 256;
   const size_t kKeysPerFile = 1 << 10;
-  const size_t kDictLen = 4 << 10;
+  const size_t kDictLen = 16 << 10;
+  const size_t kBlockLen = 4 << 10;
 
   Options options = CurrentOptions();
   if (bottommost_) {
     options.bottommost_compression = compression_type_;
     options.bottommost_compression_opts.enabled = true;
     options.bottommost_compression_opts.max_dict_bytes = kDictLen;
+    options.bottommost_compression_opts.max_dict_buffer_bytes = kBlockLen;
   } else {
     options.compression = compression_type_;
     options.compression_opts.max_dict_bytes = kDictLen;
+    options.compression_opts.max_dict_buffer_bytes = kBlockLen;
   }
   options.disable_auto_compactions = true;
   options.statistics = CreateDBStatistics();
   BlockBasedTableOptions bbto;
+  bbto.block_size = kBlockLen;
   bbto.cache_index_and_filter_blocks = true;
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   Reopen(options);
@@ -1548,17 +1602,28 @@ TEST_P(PresetCompressionDictTest, CompactBottommost) {
   ASSERT_EQ("2", FilesPerLevel(0));
 #endif  // ROCKSDB_LITE
 
-  uint64_t prev_compression_dict_misses =
-      TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_MISS);
+  uint64_t prev_compression_dict_bytes_inserted =
+      TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT);
   CompactRangeOptions cro;
   ASSERT_OK(db_->CompactRange(cro, nullptr, nullptr));
 #ifndef ROCKSDB_LITE
   ASSERT_EQ("0,1", FilesPerLevel(0));
 #endif  // ROCKSDB_LITE
-  // If there's a compression dictionary, it should have been loaded when the
-  // compaction finished, incurring a cache miss.
-  ASSERT_EQ(prev_compression_dict_misses + 1,
-            TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_MISS));
+  ASSERT_GT(
+      TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+      prev_compression_dict_bytes_inserted);
+  // TODO(ajkr): fix the below assertion to work with ZSTD. The expectation on
+  // number of bytes needs to be adjusted in case the cached block is in ZSTD's
+  // digested dictionary format.
+  if (compression_type_ != kZSTD &&
+      compression_type_ != kZSTDNotFinalCompression) {
+    // Although we limited buffering to `kBlockLen`, there may be up to two
+    // blocks of data included in the dictionary since we only check limit after
+    // each block is built.
+    ASSERT_LE(
+        TestGetTickerCount(options, BLOCK_CACHE_COMPRESSION_DICT_BYTES_INSERT),
+        prev_compression_dict_bytes_inserted + 2 * kBlockLen);
+  }
 }
 
 class CompactionCompressionListener : public EventListener {
@@ -5363,6 +5428,98 @@ TEST_F(DBTest2, AutoPrefixMode1) {
     ASSERT_EQ("a1", iterator->key().ToString());
   }
 }
+
+class RenameCurrentTest : public DBTestBase,
+                          public testing::WithParamInterface<std::string> {
+ public:
+  RenameCurrentTest()
+      : DBTestBase("rename_current_test", /*env_do_fsync=*/true),
+        sync_point_(GetParam()) {}
+
+  ~RenameCurrentTest() override {}
+
+  void SetUp() override {
+    env_->no_file_overwrite_.store(true, std::memory_order_release);
+  }
+
+  void TearDown() override {
+    env_->no_file_overwrite_.store(false, std::memory_order_release);
+  }
+
+  void SetupSyncPoints() {
+    SyncPoint::GetInstance()->DisableProcessing();
+    SyncPoint::GetInstance()->SetCallBack(sync_point_, [&](void* arg) {
+      Status* s = reinterpret_cast<Status*>(arg);
+      assert(s);
+      *s = Status::IOError("Injected IO error.");
+    });
+  }
+
+  const std::string sync_point_;
+};
+
+INSTANTIATE_TEST_CASE_P(DistributedFS, RenameCurrentTest,
+                        ::testing::Values("SetCurrentFile:BeforeRename",
+                                          "SetCurrentFile:AfterRename"));
+
+TEST_P(RenameCurrentTest, Open) {
+  Destroy(last_options_);
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  SetupSyncPoints();
+  SyncPoint::GetInstance()->EnableProcessing();
+  Status s = TryReopen(options);
+  ASSERT_NOK(s);
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  Reopen(options);
+}
+
+TEST_P(RenameCurrentTest, Flush) {
+  Destroy(last_options_);
+  Options options = GetDefaultOptions();
+  options.max_manifest_file_size = 1;
+  options.create_if_missing = true;
+  Reopen(options);
+  ASSERT_OK(Put("key", "value"));
+  SetupSyncPoints();
+  SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_NOK(Flush());
+
+  ASSERT_NOK(Put("foo", "value"));
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  Reopen(options);
+  ASSERT_EQ("value", Get("key"));
+  ASSERT_EQ("NOT_FOUND", Get("foo"));
+}
+
+TEST_P(RenameCurrentTest, Compaction) {
+  Destroy(last_options_);
+  Options options = GetDefaultOptions();
+  options.max_manifest_file_size = 1;
+  options.create_if_missing = true;
+  Reopen(options);
+  ASSERT_OK(Put("a", "a_value"));
+  ASSERT_OK(Put("c", "c_value"));
+  ASSERT_OK(Flush());
+
+  ASSERT_OK(Put("b", "b_value"));
+  ASSERT_OK(Put("d", "d_value"));
+  ASSERT_OK(Flush());
+
+  SetupSyncPoints();
+  SyncPoint::GetInstance()->EnableProcessing();
+  ASSERT_NOK(db_->CompactRange(CompactRangeOptions(), /*begin=*/nullptr,
+                               /*end=*/nullptr));
+
+  ASSERT_NOK(Put("foo", "value"));
+
+  SyncPoint::GetInstance()->DisableProcessing();
+  Reopen(options);
+  ASSERT_EQ("NOT_FOUND", Get("foo"));
+  ASSERT_EQ("d_value", Get("d"));
+}
 #endif  // ROCKSDB_LITE
 
 // WAL recovery mode is WALRecoveryMode::kPointInTimeRecovery.
@@ -5390,6 +5547,35 @@ TEST_F(DBTest2, PointInTimeRecoveryWithIOErrorWhileReadingWal) {
   Status s = TryReopen(options);
   ASSERT_TRUE(s.IsIOError());
 }
+
+TEST_F(DBTest2, PointInTimeRecoveryWithSyncFailureInCFCreation) {
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
+      {{"DBImpl::BackgroundCallFlush:Start:1",
+        "PointInTimeRecoveryWithSyncFailureInCFCreation:1"},
+       {"PointInTimeRecoveryWithSyncFailureInCFCreation:2",
+        "DBImpl::BackgroundCallFlush:Start:2"}});
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->EnableProcessing();
+
+  CreateColumnFamilies({"test1"}, Options());
+  ASSERT_OK(Put("foo", "bar"));
+
+  // Creating a CF when a flush is going on, log is synced but the
+  // closed log file is not synced and corrupted.
+  port::Thread flush_thread([&]() { ASSERT_NOK(Flush()); });
+  TEST_SYNC_POINT("PointInTimeRecoveryWithSyncFailureInCFCreation:1");
+  CreateColumnFamilies({"test2"}, Options());
+  env_->corrupt_in_sync_ = true;
+  TEST_SYNC_POINT("PointInTimeRecoveryWithSyncFailureInCFCreation:2");
+  flush_thread.join();
+  env_->corrupt_in_sync_ = false;
+  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->DisableProcessing();
+
+  // Reopening the DB should not corrupt anything
+  Options options = CurrentOptions();
+  options.wal_recovery_mode = WALRecoveryMode::kPointInTimeRecovery;
+  ReopenWithColumnFamilies({"default", "test1", "test2"}, options);
+}
+
 }  // namespace ROCKSDB_NAMESPACE
 
 #ifdef ROCKSDB_UNITTESTS_WITH_CUSTOM_OBJECTS_FROM_STATIC_LIBS

db/db_test_util.cc

@@ -10,6 +10,7 @@
 #include "db/db_test_util.h"
 
 #include "db/forward_iterator.h"
+#include "env/mock_env.h"
 #include "rocksdb/convenience.h"
 #include "rocksdb/env_encryption.h"
 #include "rocksdb/utilities/object_registry.h"
@@ -43,6 +44,7 @@ SpecialEnv::SpecialEnv(Env* base, bool time_elapse_only_sleep)
   manifest_sync_error_.store(false, std::memory_order_release);
   manifest_write_error_.store(false, std::memory_order_release);
   log_write_error_.store(false, std::memory_order_release);
+  no_file_overwrite_.store(false, std::memory_order_release);
   random_file_open_counter_.store(0, std::memory_order_relaxed);
   delete_count_.store(0, std::memory_order_relaxed);
   num_open_wal_file_.store(0);
@@ -499,6 +501,7 @@ Options DBTestBase::GetOptions(
       break;
     }
     case kBlockBasedTableWithPartitionedIndex: {
+      table_options.format_version = 3;
       table_options.index_type = BlockBasedTableOptions::kTwoLevelIndexSearch;
       options.prefix_extractor.reset(NewNoopTransform());
       break;
@@ -1125,8 +1128,34 @@ std::string DBTestBase::FilesPerLevel(int cf) {
   result.resize(last_non_zero_offset);
   return result;
 }
+
 #endif  // !ROCKSDB_LITE
 
+std::vector<uint64_t> DBTestBase::GetBlobFileNumbers() {
+  VersionSet* const versions = dbfull()->TEST_GetVersionSet();
+  assert(versions);
+
+  ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
+  assert(cfd);
+
+  Version* const current = cfd->current();
+  assert(current);
+
+  const VersionStorageInfo* const storage_info = current->storage_info();
+  assert(storage_info);
+
+  const auto& blob_files = storage_info->GetBlobFiles();
+
+  std::vector<uint64_t> result;
+  result.reserve(blob_files.size());
+
+  for (const auto& blob_file : blob_files) {
+    result.emplace_back(blob_file.first);
+  }
+
+  return result;
+}
+
 size_t DBTestBase::CountFiles() {
   size_t count = 0;
   std::vector<std::string> files;
@@ -1435,26 +1464,26 @@ void DBTestBase::CopyFile(const std::string& source,
   ASSERT_OK(destfile->Close());
 }
 
-Status DBTestBase::GetAllSSTFiles(
-    std::unordered_map<std::string, uint64_t>* sst_files,
+Status DBTestBase::GetAllDataFiles(
+    const FileType file_type, std::unordered_map<std::string, uint64_t>* files,
     uint64_t* total_size /* = nullptr */) {
   if (total_size) {
     *total_size = 0;
   }
-  std::vector<std::string> files;
-  Status s = env_->GetChildren(dbname_, &files);
+  std::vector<std::string> children;
+  Status s = env_->GetChildren(dbname_, &children);
   if (s.ok()) {
-    for (auto& file_name : files) {
+    for (auto& file_name : children) {
       uint64_t number;
       FileType type;
-      if (ParseFileName(file_name, &number, &type) && type == kTableFile) {
+      if (ParseFileName(file_name, &number, &type) && type == file_type) {
         std::string file_path = dbname_ + "/" + file_name;
         uint64_t file_size = 0;
         s = env_->GetFileSize(file_path, &file_size);
         if (!s.ok()) {
           break;
         }
-        (*sst_files)[file_path] = file_size;
+        (*files)[file_path] = file_size;
         if (total_size) {
           *total_size += file_size;
         }

db/db_test_util.h

@@ -23,7 +23,6 @@
 
 #include "db/db_impl/db_impl.h"
 #include "db/dbformat.h"
-#include "env/mock_env.h"
 #include "file/filename.h"
 #include "memtable/hash_linklist_rep.h"
 #include "rocksdb/cache.h"
@@ -40,7 +39,6 @@
 #include "rocksdb/utilities/checkpoint.h"
 #include "table/mock_table.h"
 #include "table/scoped_arena_iterator.h"
-#include "test_util/mock_time_env.h"
 #include "test_util/sync_point.h"
 #include "test_util/testharness.h"
 #include "util/cast_util.h"
@@ -50,6 +48,7 @@
 #include "utilities/merge_operators.h"
 
 namespace ROCKSDB_NAMESPACE {
+class MockEnv;
 
 namespace anon {
 class AtomicCounter {
@@ -230,6 +229,11 @@ class SpecialEnv : public EnvWrapper {
           return base_->Append(data);
         }
       }
+      Status Append(
+          const Slice& data,
+          const DataVerificationInfo& /* verification_info */) override {
+        return Append(data);
+      }
       Status PositionedAppend(const Slice& data, uint64_t offset) override {
         if (env_->table_write_callback_) {
           (*env_->table_write_callback_)();
@@ -244,6 +248,11 @@ class SpecialEnv : public EnvWrapper {
           return base_->PositionedAppend(data, offset);
         }
       }
+      Status PositionedAppend(
+          const Slice& data, uint64_t offset,
+          const DataVerificationInfo& /* verification_info */) override {
+        return PositionedAppend(data, offset);
+      }
       Status Truncate(uint64_t size) override { return base_->Truncate(size); }
       Status RangeSync(uint64_t offset, uint64_t nbytes) override {
         Status s = base_->RangeSync(offset, nbytes);
@@ -306,6 +315,12 @@ class SpecialEnv : public EnvWrapper {
           return base_->Append(data);
         }
       }
+      Status Append(
+          const Slice& data,
+          const DataVerificationInfo& /*verification_info*/) override {
+        return Append(data);
+      }
+
       Status Truncate(uint64_t size) override { return base_->Truncate(size); }
       Status Close() override { return base_->Close(); }
       Status Flush() override { return base_->Flush(); }
@@ -357,6 +372,11 @@ class SpecialEnv : public EnvWrapper {
 #endif
         return s;
       }
+      Status Append(
+          const Slice& data,
+          const DataVerificationInfo& /* verification_info */) override {
+        return Append(data);
+      }
       Status Truncate(uint64_t size) override { return base_->Truncate(size); }
       Status Close() override {
 // SyncPoint is not supported in Released Windows Mode.
@@ -373,6 +393,10 @@ class SpecialEnv : public EnvWrapper {
       Status Flush() override { return base_->Flush(); }
       Status Sync() override {
         ++env_->sync_counter_;
+        if (env_->corrupt_in_sync_) {
+          Append(std::string(33000, ' '));
+          return Status::IOError("Ingested Sync Failure");
+        }
         if (env_->skip_fsync_) {
           return Status::OK();
         } else {
@@ -395,6 +419,11 @@ class SpecialEnv : public EnvWrapper {
       OtherFile(SpecialEnv* env, std::unique_ptr<WritableFile>&& b)
           : env_(env), base_(std::move(b)) {}
       Status Append(const Slice& data) override { return base_->Append(data); }
+      Status Append(
+          const Slice& data,
+          const DataVerificationInfo& /*verification_info*/) override {
+        return Append(data);
+      }
       Status Truncate(uint64_t size) override { return base_->Truncate(size); }
       Status Close() override { return base_->Close(); }
       Status Flush() override { return base_->Flush(); }
@@ -415,6 +444,11 @@ class SpecialEnv : public EnvWrapper {
       std::unique_ptr<WritableFile> base_;
     };
 
+    if (no_file_overwrite_.load(std::memory_order_acquire) &&
+        target()->FileExists(f).ok()) {
+      return Status::NotSupported("SpecialEnv::no_file_overwrite_ is true.");
+    }
+
     if (non_writeable_rate_.load(std::memory_order_acquire) > 0) {
       uint32_t random_number;
       {
@@ -662,6 +696,9 @@ class SpecialEnv : public EnvWrapper {
   // Slow down every log write, in micro-seconds.
   std::atomic<int> log_write_slowdown_;
 
+  // If true, returns Status::NotSupported for file overwrite.
+  std::atomic<bool> no_file_overwrite_;
+
   // Number of WAL files that are still open for write.
   std::atomic<int> num_open_wal_file_;
 
@@ -684,6 +721,9 @@ class SpecialEnv : public EnvWrapper {
   // If true, all fsync to files and directories are skipped.
   bool skip_fsync_ = false;
 
+  // If true, ingest the corruption to file during sync.
+  bool corrupt_in_sync_ = false;
+
   std::atomic<uint32_t> non_writeable_rate_;
 
   std::atomic<uint32_t> new_writable_count_;
@@ -1065,6 +1105,8 @@ class DBTestBase : public testing::Test {
   int TotalTableFiles(int cf = 0, int levels = -1);
 #endif  // ROCKSDB_LITE
 
+  std::vector<uint64_t> GetBlobFileNumbers();
+
   // Return spread of files per level
   std::string FilesPerLevel(int cf = 0);
 
@@ -1154,8 +1196,9 @@ class DBTestBase : public testing::Test {
   void CopyFile(const std::string& source, const std::string& destination,
                 uint64_t size = 0);
 
-  Status GetAllSSTFiles(std::unordered_map<std::string, uint64_t>* sst_files,
-                        uint64_t* total_size = nullptr);
+  Status GetAllDataFiles(const FileType file_type,
+                         std::unordered_map<std::string, uint64_t>* sst_files,
+                         uint64_t* total_size = nullptr);
 
   std::vector<std::uint64_t> ListTableFiles(Env* env, const std::string& path);
 

db/db_wal_test.cc

@@ -8,12 +8,13 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 
 #include "db/db_test_util.h"
-#include "env/composite_env_wrapper.h"
 #include "options/options_helper.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
+#include "rocksdb/file_system.h"
 #include "test_util/sync_point.h"
 #include "utilities/fault_injection_env.h"
+#include "utilities/fault_injection_fs.h"
 
 namespace ROCKSDB_NAMESPACE {
 class DBWALTestBase : public DBTestBase {
@@ -424,28 +425,174 @@ TEST_F(DBWALTest, RecoverWithBlob) {
   const InternalStats* const internal_stats = cfd->internal_stats();
   ASSERT_NE(internal_stats, nullptr);
 
-  const uint64_t expected_bytes =
-      table_file->fd.GetFileSize() + blob_file->GetTotalBlobBytes();
-
   const auto& compaction_stats = internal_stats->TEST_GetCompactionStats();
   ASSERT_FALSE(compaction_stats.empty());
-  ASSERT_EQ(compaction_stats[0].bytes_written, expected_bytes);
-  ASSERT_EQ(compaction_stats[0].num_output_files, 2);
+  ASSERT_EQ(compaction_stats[0].bytes_written, table_file->fd.GetFileSize());
+  ASSERT_EQ(compaction_stats[0].bytes_written_blob,
+            blob_file->GetTotalBlobBytes());
+  ASSERT_EQ(compaction_stats[0].num_output_files, 1);
+  ASSERT_EQ(compaction_stats[0].num_output_files_blob, 1);
 
   const uint64_t* const cf_stats_value = internal_stats->TEST_GetCFStatsValue();
-  ASSERT_EQ(cf_stats_value[InternalStats::BYTES_FLUSHED], expected_bytes);
+  ASSERT_EQ(cf_stats_value[InternalStats::BYTES_FLUSHED],
+            compaction_stats[0].bytes_written +
+                compaction_stats[0].bytes_written_blob);
 #endif  // ROCKSDB_LITE
 }
 
+TEST_F(DBWALTest, RecoverWithBlobMultiSST) {
+  // Write several large (4 KB) values without flushing. Note that blob files
+  // are not actually enabled at this point.
+  std::string large_value(1 << 12, 'a');
+
+  constexpr int num_keys = 64;
+
+  for (int i = 0; i < num_keys; ++i) {
+    ASSERT_OK(Put(Key(i), large_value));
+  }
+
+  // There should be no files just yet since we haven't flushed.
+  {
+    VersionSet* const versions = dbfull()->TEST_GetVersionSet();
+    ASSERT_NE(versions, nullptr);
+
+    ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
+    ASSERT_NE(cfd, nullptr);
+
+    Version* const current = cfd->current();
+    ASSERT_NE(current, nullptr);
+
+    const VersionStorageInfo* const storage_info = current->storage_info();
+    ASSERT_NE(storage_info, nullptr);
+
+    ASSERT_EQ(storage_info->num_non_empty_levels(), 0);
+    ASSERT_TRUE(storage_info->GetBlobFiles().empty());
+  }
+
+  // Reopen the database with blob files enabled and write buffer size set to a
+  // smaller value. Multiple table files+blob files should be written and added
+  // to the Version during recovery.
+  Options options;
+  options.write_buffer_size = 1 << 16;  // 64 KB
+  options.enable_blob_files = true;
+  options.avoid_flush_during_recovery = false;
+  options.disable_auto_compactions = true;
+  options.env = env_;
+
+  Reopen(options);
+
+  for (int i = 0; i < num_keys; ++i) {
+    ASSERT_EQ(Get(Key(i)), large_value);
+  }
+
+  VersionSet* const versions = dbfull()->TEST_GetVersionSet();
+  ASSERT_NE(versions, nullptr);
+
+  ColumnFamilyData* const cfd = versions->GetColumnFamilySet()->GetDefault();
+  ASSERT_NE(cfd, nullptr);
+
+  Version* const current = cfd->current();
+  ASSERT_NE(current, nullptr);
+
+  const VersionStorageInfo* const storage_info = current->storage_info();
+  ASSERT_NE(storage_info, nullptr);
+
+  const auto& l0_files = storage_info->LevelFiles(0);
+  ASSERT_GT(l0_files.size(), 1);
+
+  const auto& blob_files = storage_info->GetBlobFiles();
+  ASSERT_GT(blob_files.size(), 1);
+
+  ASSERT_EQ(l0_files.size(), blob_files.size());
+}
+
+TEST_F(DBWALTest, WALWithChecksumHandoff) {
+#ifndef ROCKSDB_ASSERT_STATUS_CHECKED
+  if (mem_env_ || encrypted_env_) {
+    ROCKSDB_GTEST_SKIP("Test requires non-mem or non-encrypted environment");
+    return;
+  }
+  std::shared_ptr<FaultInjectionTestFS> fault_fs(
+      new FaultInjectionTestFS(FileSystem::Default()));
+  std::unique_ptr<Env> fault_fs_env(NewCompositeEnv(fault_fs));
+  do {
+    Options options = CurrentOptions();
+
+    options.checksum_handoff_file_types.Add(FileType::kWalFile);
+    options.env = fault_fs_env.get();
+    fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+
+    CreateAndReopenWithCF({"pikachu"}, options);
+    WriteOptions writeOpt = WriteOptions();
+    writeOpt.disableWAL = true;
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "foo", "v1"));
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "bar", "v1"));
+
+    ReopenWithColumnFamilies({"default", "pikachu"}, options);
+    ASSERT_EQ("v1", Get(1, "foo"));
+    ASSERT_EQ("v1", Get(1, "bar"));
+
+    writeOpt.disableWAL = false;
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "bar", "v2"));
+    writeOpt.disableWAL = true;
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "foo", "v2"));
+
+    ReopenWithColumnFamilies({"default", "pikachu"}, options);
+    // Both value's should be present.
+    ASSERT_EQ("v2", Get(1, "bar"));
+    ASSERT_EQ("v2", Get(1, "foo"));
+
+    writeOpt.disableWAL = true;
+    // This put, data is persisted by Flush
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "bar", "v3"));
+    ReopenWithColumnFamilies({"default", "pikachu"}, options);
+    writeOpt.disableWAL = false;
+    // Data is persisted in the WAL
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "zoo", "v3"));
+    // The hash does not match, write fails
+    fault_fs->SetChecksumHandoffFuncType(ChecksumType::kxxHash);
+    writeOpt.disableWAL = false;
+    ASSERT_NOK(dbfull()->Put(writeOpt, handles_[1], "foo", "v3"));
+
+    ReopenWithColumnFamilies({"default", "pikachu"}, options);
+    // Due to the write failure, Get should not find
+    ASSERT_NE("v3", Get(1, "foo"));
+    ASSERT_EQ("v3", Get(1, "zoo"));
+    ASSERT_EQ("v3", Get(1, "bar"));
+
+    fault_fs->SetChecksumHandoffFuncType(ChecksumType::kCRC32c);
+    // Each write will be similated as corrupted.
+    fault_fs->IngestDataCorruptionBeforeWrite();
+    writeOpt.disableWAL = true;
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "bar", "v4"));
+    writeOpt.disableWAL = false;
+    ASSERT_NOK(dbfull()->Put(writeOpt, handles_[1], "foo", "v4"));
+    ReopenWithColumnFamilies({"default", "pikachu"}, options);
+    ASSERT_NE("v4", Get(1, "foo"));
+    ASSERT_NE("v4", Get(1, "bar"));
+    fault_fs->NoDataCorruptionBeforeWrite();
+
+    fault_fs->SetChecksumHandoffFuncType(ChecksumType::kNoChecksum);
+    // The file system does not provide checksum method and verification.
+    writeOpt.disableWAL = true;
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "bar", "v5"));
+    writeOpt.disableWAL = false;
+    ASSERT_OK(dbfull()->Put(writeOpt, handles_[1], "foo", "v5"));
+    ReopenWithColumnFamilies({"default", "pikachu"}, options);
+    ASSERT_EQ("v5", Get(1, "foo"));
+    ASSERT_EQ("v5", Get(1, "bar"));
+
+    Destroy(options);
+  } while (ChangeWalOptions());
+#endif  // ROCKSDB_ASSERT_STATUS_CHECKED
+}
+
 class DBRecoveryTestBlobError
     : public DBWALTest,
       public testing::WithParamInterface<std::string> {
  public:
-  DBRecoveryTestBlobError()
-      : fault_injection_env_(env_), sync_point_(GetParam()) {}
-  ~DBRecoveryTestBlobError() { Close(); }
+  DBRecoveryTestBlobError() : sync_point_(GetParam()) {}
 
-  FaultInjectionTestEnv fault_injection_env_;
   std::string sync_point_;
 };
 
@@ -460,21 +607,19 @@ TEST_P(DBRecoveryTestBlobError, RecoverWithBlobError) {
 
   // Reopen with blob files enabled but make blob file writing fail during
   // recovery.
-  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* /* arg */) {
-    fault_injection_env_.SetFilesystemActive(false,
-                                             Status::IOError(sync_point_));
+  SyncPoint::GetInstance()->SetCallBack(sync_point_, [this](void* arg) {
+    Status* const s = static_cast<Status*>(arg);
+    assert(s);
+
+    (*s) = Status::IOError(sync_point_);
   });
-  SyncPoint::GetInstance()->SetCallBack(
-      "BuildTable:BeforeDeleteFile", [this](void* /* arg */) {
-        fault_injection_env_.SetFilesystemActive(true);
-      });
   SyncPoint::GetInstance()->EnableProcessing();
 
   Options options;
   options.enable_blob_files = true;
   options.avoid_flush_during_recovery = false;
   options.disable_auto_compactions = true;
-  options.env = &fault_injection_env_;
+  options.env = env_;
 
   ASSERT_NOK(TryReopen(options));
 
@@ -1086,7 +1231,7 @@ class RecoveryTestHelper {
     *count = 0;
 
     std::shared_ptr<Cache> table_cache = NewLRUCache(50, 0);
-    EnvOptions env_options;
+    FileOptions file_options;
     WriteBufferManager write_buffer_manager(db_options.db_write_buffer_size);
 
     std::unique_ptr<VersionSet> versions;
@@ -1094,22 +1239,22 @@ class RecoveryTestHelper {
     WriteController write_controller;
 
     versions.reset(new VersionSet(
-        test->dbname_, &db_options, env_options, table_cache.get(),
+        test->dbname_, &db_options, file_options, table_cache.get(),
         &write_buffer_manager, &write_controller,
         /*block_cache_tracer=*/nullptr, /*io_tracer=*/nullptr));
 
     wal_manager.reset(
-        new WalManager(db_options, env_options, /*io_tracer=*/nullptr));
+        new WalManager(db_options, file_options, /*io_tracer=*/nullptr));
 
     std::unique_ptr<log::Writer> current_log_writer;
 
     for (size_t j = kWALFileOffset; j < wal_count + kWALFileOffset; j++) {
       uint64_t current_log_number = j;
       std::string fname = LogFileName(test->dbname_, current_log_number);
-      std::unique_ptr<WritableFile> file;
-      ASSERT_OK(db_options.env->NewWritableFile(fname, &file, env_options));
-      std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-          NewLegacyWritableFileWrapper(std::move(file)), fname, env_options));
+      std::unique_ptr<WritableFileWriter> file_writer;
+      ASSERT_OK(WritableFileWriter::Create(db_options.env->GetFileSystem(),
+                                           fname, file_options, &file_writer,
+                                           nullptr));
       current_log_writer.reset(
           new log::Writer(std::move(file_writer), current_log_number,
                           db_options.recycle_log_file_num > 0));

db/db_with_timestamp_basic_test.cc

@@ -223,6 +223,111 @@ TEST_F(DBBasicTestWithTimestamp, CompactRangeWithSpecifiedRange) {
   Close();
 }
 
+TEST_F(DBBasicTestWithTimestamp, UpdateFullHistoryTsLow) {
+  Options options = CurrentOptions();
+  options.env = env_;
+  options.create_if_missing = true;
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+
+  const std::string kKey = "test kKey";
+
+  // Test set ts_low first and flush()
+  int current_ts_low = 5;
+  std::string ts_low_str = Timestamp(current_ts_low, 0);
+  Slice ts_low = ts_low_str;
+  CompactRangeOptions comp_opts;
+  comp_opts.full_history_ts_low = &ts_low;
+  comp_opts.bottommost_level_compaction = BottommostLevelCompaction::kForce;
+
+  ASSERT_OK(db_->CompactRange(comp_opts, nullptr, nullptr));
+
+  auto* cfd =
+      static_cast_with_check<ColumnFamilyHandleImpl>(db_->DefaultColumnFamily())
+          ->cfd();
+  auto result_ts_low = cfd->GetFullHistoryTsLow();
+
+  ASSERT_TRUE(test_cmp.CompareTimestamp(ts_low, result_ts_low) == 0);
+
+  for (int i = 0; i < 10; i++) {
+    WriteOptions write_opts;
+    std::string ts_str = Timestamp(i, 0);
+    Slice ts = ts_str;
+    write_opts.timestamp = &ts;
+    ASSERT_OK(db_->Put(write_opts, kKey, Key(i)));
+  }
+  ASSERT_OK(Flush());
+
+  for (int i = 0; i < 10; i++) {
+    ReadOptions read_opts;
+    std::string ts_str = Timestamp(i, 0);
+    Slice ts = ts_str;
+    read_opts.timestamp = &ts;
+    std::string value;
+    Status status = db_->Get(read_opts, kKey, &value);
+    if (i < current_ts_low) {
+      ASSERT_TRUE(status.IsNotFound());
+    } else {
+      ASSERT_OK(status);
+      ASSERT_TRUE(value.compare(Key(i)) == 0);
+    }
+  }
+
+  // Test set ts_low and then trigger compaction
+  for (int i = 10; i < 20; i++) {
+    WriteOptions write_opts;
+    std::string ts_str = Timestamp(i, 0);
+    Slice ts = ts_str;
+    write_opts.timestamp = &ts;
+    ASSERT_OK(db_->Put(write_opts, kKey, Key(i)));
+  }
+
+  ASSERT_OK(Flush());
+
+  current_ts_low = 15;
+  ts_low_str = Timestamp(current_ts_low, 0);
+  ts_low = ts_low_str;
+  comp_opts.full_history_ts_low = &ts_low;
+  ASSERT_OK(db_->CompactRange(comp_opts, nullptr, nullptr));
+  result_ts_low = cfd->GetFullHistoryTsLow();
+  ASSERT_TRUE(test_cmp.CompareTimestamp(ts_low, result_ts_low) == 0);
+
+  for (int i = 0; i < 20; i++) {
+    ReadOptions read_opts;
+    std::string ts_str = Timestamp(i, 0);
+    Slice ts = ts_str;
+    read_opts.timestamp = &ts;
+    std::string value;
+    Status status = db_->Get(read_opts, kKey, &value);
+    if (i < current_ts_low) {
+      ASSERT_TRUE(status.IsNotFound());
+    } else {
+      ASSERT_OK(status);
+      ASSERT_TRUE(value.compare(Key(i)) == 0);
+    }
+  }
+
+  // Test invalid compaction with range
+  Slice start(kKey), end(kKey);
+  Status s = db_->CompactRange(comp_opts, &start, &end);
+  ASSERT_TRUE(s.IsInvalidArgument());
+  s = db_->CompactRange(comp_opts, &start, nullptr);
+  ASSERT_TRUE(s.IsInvalidArgument());
+  s = db_->CompactRange(comp_opts, nullptr, &end);
+  ASSERT_TRUE(s.IsInvalidArgument());
+
+  // Test invalid compaction with the decreasing ts_low
+  ts_low_str = Timestamp(current_ts_low - 1, 0);
+  ts_low = ts_low_str;
+  comp_opts.full_history_ts_low = &ts_low;
+  s = db_->CompactRange(comp_opts, nullptr, nullptr);
+  ASSERT_TRUE(s.IsInvalidArgument());
+
+  Close();
+}
+
 TEST_F(DBBasicTestWithTimestamp, GetApproximateSizes) {
   Options options = CurrentOptions();
   options.write_buffer_size = 100000000;  // Large write buffer
@@ -297,12 +402,11 @@ TEST_F(DBBasicTestWithTimestamp, GetApproximateSizes) {
   ASSERT_OK(
       db_->GetApproximateSizes(size_approx_options, default_cf, &r, 1, &size));
   ASSERT_EQ(size, 0);
-  std::cout << size << std::endl;
 
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterate) {
+TEST_F(DBBasicTestWithTimestamp, SimpleIterate) {
   const int kNumKeysPerFile = 128;
   const uint64_t kMaxKey = 1024;
   Options options = CurrentOptions();
@@ -334,6 +438,7 @@ TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterate) {
     std::unique_ptr<Iterator> it(db_->NewIterator(read_opts));
     int count = 0;
     uint64_t key = 0;
+    // Forward iterate.
     for (it->Seek(Key1(0)), key = start_keys[i]; it->Valid();
          it->Next(), ++count, ++key) {
       CheckIterUserEntry(it.get(), Key1(key), kTypeValue,
@@ -342,7 +447,16 @@ TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterate) {
     size_t expected_count = kMaxKey - start_keys[i] + 1;
     ASSERT_EQ(expected_count, count);
 
-    // SeekToFirst() with lower bound.
+    // Backward iterate.
+    count = 0;
+    for (it->SeekForPrev(Key1(kMaxKey)), key = kMaxKey; it->Valid();
+         it->Prev(), ++count, --key) {
+      CheckIterUserEntry(it.get(), Key1(key), kTypeValue,
+                         "value" + std::to_string(i), write_timestamps[i]);
+    }
+    ASSERT_EQ(static_cast<size_t>(kMaxKey) - start_keys[i] + 1, count);
+
+    // SeekToFirst()/SeekToLast() with lower/upper bounds.
     // Then iter with lower and upper bounds.
     uint64_t l = 0;
     uint64_t r = kMaxKey + 1;
@@ -360,6 +474,12 @@ TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterate) {
                            "value" + std::to_string(i), write_timestamps[i]);
       }
       ASSERT_EQ(r - std::max(l, start_keys[i]), count);
+
+      for (it->SeekToLast(), key = std::min(r, kMaxKey + 1), count = 0;
+           it->Valid(); it->Prev(), --key, ++count) {
+        CheckIterUserEntry(it.get(), Key1(key - 1), kTypeValue,
+                           "value" + std::to_string(i), write_timestamps[i]);
+      }
       l += (kMaxKey / 100);
       r -= (kMaxKey / 100);
     }
@@ -367,7 +487,93 @@ TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterate) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLessThanKey) {
+class DBBasicTestWithTimestampTableOptions
+    : public DBBasicTestWithTimestampBase,
+      public testing::WithParamInterface<BlockBasedTableOptions::IndexType> {
+ public:
+  explicit DBBasicTestWithTimestampTableOptions()
+      : DBBasicTestWithTimestampBase(
+            "db_basic_test_with_timestamp_table_options") {}
+};
+
+INSTANTIATE_TEST_CASE_P(
+    Timestamp, DBBasicTestWithTimestampTableOptions,
+    testing::Values(
+        BlockBasedTableOptions::IndexType::kBinarySearch,
+        BlockBasedTableOptions::IndexType::kHashSearch,
+        BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch,
+        BlockBasedTableOptions::IndexType::kBinarySearchWithFirstKey));
+
+TEST_P(DBBasicTestWithTimestampTableOptions, GetAndMultiGet) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.prefix_extractor.reset(NewFixedPrefixTransform(3));
+  options.compression = kNoCompression;
+  BlockBasedTableOptions bbto;
+  bbto.index_type = GetParam();
+  bbto.block_size = 100;
+  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator cmp(kTimestampSize);
+  options.comparator = &cmp;
+  DestroyAndReopen(options);
+  constexpr uint64_t kNumKeys = 1024;
+  for (uint64_t k = 0; k < kNumKeys; ++k) {
+    WriteOptions write_opts;
+    std::string ts_str = Timestamp(1, 0);
+    Slice ts = ts_str;
+    write_opts.timestamp = &ts;
+    ASSERT_OK(db_->Put(write_opts, Key1(k), "value" + std::to_string(k)));
+  }
+  ASSERT_OK(Flush());
+  {
+    ReadOptions read_opts;
+    read_opts.total_order_seek = true;
+    std::string ts_str = Timestamp(2, 0);
+    Slice ts = ts_str;
+    read_opts.timestamp = &ts;
+    std::unique_ptr<Iterator> it(db_->NewIterator(read_opts));
+    // verify Get()
+    for (it->SeekToFirst(); it->Valid(); it->Next()) {
+      std::string value_from_get;
+      std::string key_str(it->key().data(), it->key().size());
+      std::string timestamp;
+      ASSERT_OK(db_->Get(read_opts, key_str, &value_from_get, &timestamp));
+      ASSERT_EQ(it->value(), value_from_get);
+      ASSERT_EQ(Timestamp(1, 0), timestamp);
+    }
+
+    // verify MultiGet()
+    constexpr uint64_t step = 2;
+    static_assert(0 == (kNumKeys % step),
+                  "kNumKeys must be a multiple of step");
+    for (uint64_t k = 0; k < kNumKeys; k += 2) {
+      std::vector<std::string> key_strs;
+      std::vector<Slice> keys;
+      for (size_t i = 0; i < step; ++i) {
+        key_strs.push_back(Key1(k + i));
+      }
+      for (size_t i = 0; i < step; ++i) {
+        keys.emplace_back(key_strs[i]);
+      }
+      std::vector<std::string> values;
+      std::vector<std::string> timestamps;
+      std::vector<Status> statuses =
+          db_->MultiGet(read_opts, keys, &values, &timestamps);
+      ASSERT_EQ(step, statuses.size());
+      ASSERT_EQ(step, values.size());
+      ASSERT_EQ(step, timestamps.size());
+      for (uint64_t i = 0; i < step; ++i) {
+        ASSERT_OK(statuses[i]);
+        ASSERT_EQ("value" + std::to_string(k + i), values[i]);
+        ASSERT_EQ(Timestamp(1, 0), timestamps[i]);
+      }
+    }
+  }
+  Close();
+}
+
+TEST_P(DBBasicTestWithTimestampTableOptions, SeekWithPrefixLessThanKey) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -378,6 +584,7 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLessThanKey) {
   bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
   bbto.cache_index_and_filter_blocks = true;
   bbto.whole_key_filtering = true;
+  bbto.index_type = GetParam();
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   const size_t kTimestampSize = Timestamp(0, 0).size();
   TestComparator test_cmp(kTimestampSize);
@@ -422,7 +629,7 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLessThanKey) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLargerThanKey) {
+TEST_P(DBBasicTestWithTimestampTableOptions, SeekWithPrefixLongerThanKey) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -433,6 +640,7 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLargerThanKey) {
   bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
   bbto.cache_index_and_filter_blocks = true;
   bbto.whole_key_filtering = true;
+  bbto.index_type = GetParam();
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   const size_t kTimestampSize = Timestamp(0, 0).size();
   TestComparator test_cmp(kTimestampSize);
@@ -475,7 +683,7 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithPrefixLargerThanKey) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, SeekWithBound) {
+TEST_P(DBBasicTestWithTimestampTableOptions, SeekWithBound) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -484,6 +692,7 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithBound) {
   bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
   bbto.cache_index_and_filter_blocks = true;
   bbto.whole_key_filtering = true;
+  bbto.index_type = GetParam();
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   const size_t kTimestampSize = Timestamp(0, 0).size();
   TestComparator test_cmp(kTimestampSize);
@@ -495,38 +704,135 @@ TEST_F(DBBasicTestWithTimestamp, SeekWithBound) {
   Slice ts = ts_str;
   write_opts.timestamp = &ts;
 
-  ASSERT_OK(db_->Put(write_opts, "foo1", "bar"));
+  ASSERT_OK(db_->Put(write_opts, "foo1", "bar1"));
   ASSERT_OK(Flush());
 
-  ASSERT_OK(db_->Put(write_opts, "foo2", "bar"));
+  ASSERT_OK(db_->Put(write_opts, "foo2", "bar2"));
   ASSERT_OK(Flush());
 
   // Move sst file to next level
   ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
 
-  ASSERT_OK(db_->Put(write_opts, "foo3", "bar"));
+  for (int i = 3; i < 9; ++i) {
+    ASSERT_OK(db_->Put(write_opts, "foo" + std::to_string(i),
+                       "bar" + std::to_string(i)));
+  }
   ASSERT_OK(Flush());
 
   ReadOptions read_opts;
   std::string read_ts = Timestamp(2, 0);
   ts = read_ts;
   read_opts.timestamp = &ts;
-  std::string up_bound = "foo5";
+  std::string up_bound = "foo5";  // exclusive
   Slice up_bound_slice = up_bound;
+  std::string lo_bound = "foo2";  // inclusive
+  Slice lo_bound_slice = lo_bound;
   read_opts.iterate_upper_bound = &up_bound_slice;
+  read_opts.iterate_lower_bound = &lo_bound_slice;
   read_opts.auto_prefix_mode = true;
   {
     std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts));
     // Make sure the prefix extractor doesn't include timestamp, otherwise it
     // may return invalid result.
     iter->Seek("foo");
-    ASSERT_TRUE(iter->Valid());
-    ASSERT_OK(iter->status());
+    CheckIterUserEntry(iter.get(), lo_bound, kTypeValue, "bar2",
+                       Timestamp(1, 0));
+    iter->SeekToFirst();
+    CheckIterUserEntry(iter.get(), lo_bound, kTypeValue, "bar2",
+                       Timestamp(1, 0));
+    iter->SeekForPrev("g");
+    CheckIterUserEntry(iter.get(), "foo4", kTypeValue, "bar4", Timestamp(1, 0));
+    iter->SeekToLast();
+    CheckIterUserEntry(iter.get(), "foo4", kTypeValue, "bar4", Timestamp(1, 0));
   }
 
   Close();
 }
 
+TEST_F(DBBasicTestWithTimestamp, ChangeIterationDirection) {
+  Options options = GetDefaultOptions();
+  options.create_if_missing = true;
+  options.env = env_;
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  options.prefix_extractor.reset(NewFixedPrefixTransform(1));
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
+  DestroyAndReopen(options);
+  const std::vector<std::string> timestamps = {Timestamp(1, 1), Timestamp(0, 2),
+                                               Timestamp(4, 3)};
+  const std::vector<std::tuple<std::string, std::string>> kvs = {
+      std::make_tuple("aa", "value1"), std::make_tuple("ab", "value2")};
+  for (const auto& ts : timestamps) {
+    WriteBatch wb(0, 0, kTimestampSize);
+    for (const auto& kv : kvs) {
+      const std::string& key = std::get<0>(kv);
+      const std::string& value = std::get<1>(kv);
+      ASSERT_OK(wb.Put(key, value));
+    }
+
+    ASSERT_OK(wb.AssignTimestamp(ts));
+    ASSERT_OK(db_->Write(WriteOptions(), &wb));
+  }
+  std::string read_ts_str = Timestamp(5, 3);
+  Slice read_ts = read_ts_str;
+  ReadOptions read_opts;
+  read_opts.timestamp = &read_ts;
+  std::unique_ptr<Iterator> it(db_->NewIterator(read_opts));
+
+  it->SeekToFirst();
+  ASSERT_TRUE(it->Valid());
+  it->Prev();
+  ASSERT_FALSE(it->Valid());
+
+  it->SeekToLast();
+  ASSERT_TRUE(it->Valid());
+  uint64_t prev_reseek_count =
+      options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION);
+  ASSERT_EQ(0, prev_reseek_count);
+  it->Next();
+  ASSERT_FALSE(it->Valid());
+  ASSERT_EQ(1 + prev_reseek_count,
+            options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
+
+  it->Seek(std::get<0>(kvs[0]));
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[0]), kTypeValue,
+                     std::get<1>(kvs[0]), Timestamp(4, 3));
+  it->Next();
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[1]), kTypeValue,
+                     std::get<1>(kvs[1]), Timestamp(4, 3));
+  it->Prev();
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[0]), kTypeValue,
+                     std::get<1>(kvs[0]), Timestamp(4, 3));
+
+  prev_reseek_count =
+      options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION);
+  ASSERT_EQ(1, prev_reseek_count);
+  it->Next();
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[1]), kTypeValue,
+                     std::get<1>(kvs[1]), Timestamp(4, 3));
+  ASSERT_EQ(1 + prev_reseek_count,
+            options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
+
+  it->SeekForPrev(std::get<0>(kvs[1]));
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[1]), kTypeValue,
+                     std::get<1>(kvs[1]), Timestamp(4, 3));
+  it->Prev();
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[0]), kTypeValue,
+                     std::get<1>(kvs[0]), Timestamp(4, 3));
+
+  prev_reseek_count =
+      options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION);
+  it->Next();
+  CheckIterUserEntry(it.get(), std::get<0>(kvs[1]), kTypeValue,
+                     std::get<1>(kvs[1]), Timestamp(4, 3));
+  ASSERT_EQ(1 + prev_reseek_count,
+            options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
+
+  it.reset();
+  Close();
+}
+
 TEST_F(DBBasicTestWithTimestamp, SimpleForwardIterateLowerTsBound) {
   constexpr int kNumKeysPerFile = 128;
   constexpr uint64_t kMaxKey = 1024;
@@ -725,6 +1031,16 @@ TEST_F(DBBasicTestWithTimestamp, ReseekToTargetTimestamp) {
     CheckIterUserEntry(iter.get(), "foo", kTypeValue, "value0", ts_str);
     ASSERT_EQ(
         1, options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
+
+    ts_str = Timestamp(kNumKeys, 0);
+    ts = ts_str;
+    read_opts.timestamp = &ts;
+    iter.reset(db_->NewIterator(read_opts));
+    iter->SeekToLast();
+    CheckIterUserEntry(iter.get(), "foo", kTypeValue,
+                       "value" + std::to_string(kNumKeys - 1), ts_str);
+    ASSERT_EQ(
+        2, options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
   }
   Close();
 }
@@ -775,6 +1091,47 @@ TEST_F(DBBasicTestWithTimestamp, ReseekToNextUserKey) {
   Close();
 }
 
+TEST_F(DBBasicTestWithTimestamp, ReseekToUserKeyBeforeSavedKey) {
+  Options options = GetDefaultOptions();
+  options.env = env_;
+  options.create_if_missing = true;
+  constexpr size_t kNumKeys = 16;
+  options.max_sequential_skip_in_iterations = kNumKeys / 2;
+  options.statistics = ROCKSDB_NAMESPACE::CreateDBStatistics();
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+  for (size_t i = 0; i < kNumKeys; ++i) {
+    std::string ts_str = Timestamp(static_cast<uint64_t>(i + 1), 0);
+    Slice ts = ts_str;
+    WriteOptions write_opts;
+    write_opts.timestamp = &ts;
+    Status s = db_->Put(write_opts, "b", "value" + std::to_string(i));
+    ASSERT_OK(s);
+  }
+  {
+    std::string ts_str = Timestamp(1, 0);
+    Slice ts = ts_str;
+    WriteOptions write_opts;
+    write_opts.timestamp = &ts;
+    ASSERT_OK(db_->Put(write_opts, "a", "value"));
+  }
+  {
+    ReadOptions read_opts;
+    std::string ts_str = Timestamp(1, 0);
+    Slice ts = ts_str;
+    read_opts.timestamp = &ts;
+    std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts));
+    iter->SeekToLast();
+    iter->Prev();
+    CheckIterUserEntry(iter.get(), "a", kTypeValue, "value", ts_str);
+    ASSERT_EQ(
+        1, options.statistics->getTickerCount(NUMBER_OF_RESEEKS_IN_ITERATION));
+  }
+  Close();
+}
+
 TEST_F(DBBasicTestWithTimestamp, MultiGetWithFastLocalBloom) {
   Options options = CurrentOptions();
   options.env = env_;
@@ -815,7 +1172,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetWithFastLocalBloom) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, MultiGetWithPrefix) {
+TEST_P(DBBasicTestWithTimestampTableOptions, MultiGetWithPrefix) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -824,6 +1181,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetWithPrefix) {
   bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
   bbto.cache_index_and_filter_blocks = true;
   bbto.whole_key_filtering = false;
+  bbto.index_type = GetParam();
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   const size_t kTimestampSize = Timestamp(0, 0).size();
   TestComparator test_cmp(kTimestampSize);
@@ -856,7 +1214,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetWithPrefix) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, MultiGetWithMemBloomFilter) {
+TEST_P(DBBasicTestWithTimestampTableOptions, MultiGetWithMemBloomFilter) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -865,6 +1223,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetWithMemBloomFilter) {
   bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
   bbto.cache_index_and_filter_blocks = true;
   bbto.whole_key_filtering = false;
+  bbto.index_type = GetParam();
   options.memtable_prefix_bloom_size_ratio = 0.1;
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   const size_t kTimestampSize = Timestamp(0, 0).size();
@@ -954,7 +1313,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetRangeFiltering) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, MultiGetPrefixFilter) {
+TEST_P(DBBasicTestWithTimestampTableOptions, MultiGetPrefixFilter) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -963,6 +1322,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetPrefixFilter) {
   bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
   bbto.cache_index_and_filter_blocks = true;
   bbto.whole_key_filtering = false;
+  bbto.index_type = GetParam();
   options.memtable_prefix_bloom_size_ratio = 0.1;
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   const size_t kTimestampSize = Timestamp(0, 0).size();
@@ -997,7 +1357,7 @@ TEST_F(DBBasicTestWithTimestamp, MultiGetPrefixFilter) {
   Close();
 }
 
-TEST_F(DBBasicTestWithTimestamp, MaxKeysSkipped) {
+TEST_F(DBBasicTestWithTimestamp, MaxKeysSkippedDuringNext) {
   Options options = CurrentOptions();
   options.env = env_;
   options.create_if_missing = true;
@@ -1036,6 +1396,45 @@ TEST_F(DBBasicTestWithTimestamp, MaxKeysSkipped) {
   Close();
 }
 
+TEST_F(DBBasicTestWithTimestamp, MaxKeysSkippedDuringPrev) {
+  Options options = GetDefaultOptions();
+  options.env = env_;
+  options.create_if_missing = true;
+  const size_t kTimestampSize = Timestamp(0, 0).size();
+  TestComparator test_cmp(kTimestampSize);
+  options.comparator = &test_cmp;
+  DestroyAndReopen(options);
+  constexpr size_t max_skippable_internal_keys = 2;
+  const size_t kNumKeys = max_skippable_internal_keys + 2;
+  WriteOptions write_opts;
+  Status s;
+  {
+    std::string ts_str = Timestamp(1, 0);
+    Slice ts = ts_str;
+    write_opts.timestamp = &ts;
+    ASSERT_OK(db_->Put(write_opts, "b", "value"));
+  }
+  for (size_t i = 0; i < kNumKeys; ++i) {
+    std::string ts_str = Timestamp(static_cast<uint64_t>(i + 1), 0);
+    Slice ts = ts_str;
+    write_opts.timestamp = &ts;
+    s = db_->Put(write_opts, "a", "value" + std::to_string(i));
+    ASSERT_OK(s);
+  }
+  {
+    ReadOptions read_opts;
+    read_opts.max_skippable_internal_keys = max_skippable_internal_keys;
+    std::string ts_str = Timestamp(1, 0);
+    Slice ts = ts_str;
+    read_opts.timestamp = &ts;
+    std::unique_ptr<Iterator> iter(db_->NewIterator(read_opts));
+    iter->SeekToLast();
+    iter->Prev();
+    ASSERT_TRUE(iter->status().IsIncomplete());
+  }
+  Close();
+}
+
 // Create two L0, and compact them to a new L1. In this test, L1 is L_bottom.
 // Two L0s:
 //       f1                                  f2
@@ -1099,7 +1498,8 @@ class DBBasicTestWithTimestampFilterPrefixSettings
     : public DBBasicTestWithTimestampBase,
       public testing::WithParamInterface<
           std::tuple<std::shared_ptr<const FilterPolicy>, bool, bool,
-                     std::shared_ptr<const SliceTransform>, bool, double>> {
+                     std::shared_ptr<const SliceTransform>, bool, double,
+                     BlockBasedTableOptions::IndexType>> {
  public:
   DBBasicTestWithTimestampFilterPrefixSettings()
       : DBBasicTestWithTimestampBase(
@@ -1114,6 +1514,7 @@ TEST_P(DBBasicTestWithTimestampFilterPrefixSettings, GetAndMultiGet) {
   bbto.filter_policy = std::get<0>(GetParam());
   bbto.whole_key_filtering = std::get<1>(GetParam());
   bbto.cache_index_and_filter_blocks = std::get<2>(GetParam());
+  bbto.index_type = std::get<6>(GetParam());
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   options.prefix_extractor = std::get<3>(GetParam());
   options.memtable_whole_key_filtering = std::get<4>(GetParam());
@@ -1218,7 +1619,12 @@ INSTANTIATE_TEST_CASE_P(
             std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(4)),
             std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(7)),
             std::shared_ptr<const SliceTransform>(NewFixedPrefixTransform(8))),
-        ::testing::Bool(), ::testing::Values(0, 0.1)));
+        ::testing::Bool(), ::testing::Values(0, 0.1),
+        ::testing::Values(
+            BlockBasedTableOptions::IndexType::kBinarySearch,
+            BlockBasedTableOptions::IndexType::kHashSearch,
+            BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch,
+            BlockBasedTableOptions::IndexType::kBinarySearchWithFirstKey)));
 
 class DataVisibilityTest : public DBBasicTestWithTimestampBase {
  public:
@@ -2299,14 +2705,15 @@ class DBBasicTestWithTimestampPrefixSeek
     : public DBBasicTestWithTimestampBase,
       public testing::WithParamInterface<
           std::tuple<std::shared_ptr<const SliceTransform>,
-                     std::shared_ptr<const FilterPolicy>, bool>> {
+                     std::shared_ptr<const FilterPolicy>, bool,
+                     BlockBasedTableOptions::IndexType>> {
  public:
   DBBasicTestWithTimestampPrefixSeek()
       : DBBasicTestWithTimestampBase(
             "/db_basic_test_with_timestamp_prefix_seek") {}
 };
 
-TEST_P(DBBasicTestWithTimestampPrefixSeek, ForwardIterateWithPrefix) {
+TEST_P(DBBasicTestWithTimestampPrefixSeek, IterateWithPrefix) {
   const size_t kNumKeysPerFile = 128;
   Options options = CurrentOptions();
   options.env = env_;
@@ -2318,6 +2725,7 @@ TEST_P(DBBasicTestWithTimestampPrefixSeek, ForwardIterateWithPrefix) {
   options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeysPerFile));
   BlockBasedTableOptions bbto;
   bbto.filter_policy = std::get<1>(GetParam());
+  bbto.index_type = std::get<3>(GetParam());
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   DestroyAndReopen(options);
 
@@ -2356,6 +2764,13 @@ TEST_P(DBBasicTestWithTimestampPrefixSeek, ForwardIterateWithPrefix) {
                          "value" + std::to_string(i), write_ts_list[i]);
       iter->Next();
       ASSERT_FALSE(iter->Valid());
+
+      // Seek to kMinKey
+      iter->Seek(Key1(kMinKey));
+      CheckIterUserEntry(iter.get(), Key1(kMinKey), kTypeValue,
+                         "value" + std::to_string(i), write_ts_list[i]);
+      iter->Prev();
+      ASSERT_FALSE(iter->Valid());
     }
     const std::vector<uint64_t> targets = {kMinKey, kMinKey + 0x10,
                                            kMinKey + 0x100, kMaxKey};
@@ -2371,6 +2786,7 @@ TEST_P(DBBasicTestWithTimestampPrefixSeek, ForwardIterateWithPrefix) {
       Slice read_ts = read_ts_list[i];
       read_opts.timestamp = &read_ts;
       std::unique_ptr<Iterator> it(db_->NewIterator(read_opts));
+      // Forward and backward iterate.
       for (size_t j = 0; j != targets.size(); ++j) {
         std::string start_key = Key1(targets[j]);
         uint64_t expected_ub =
@@ -2394,6 +2810,24 @@ TEST_P(DBBasicTestWithTimestampPrefixSeek, ForwardIterateWithPrefix) {
           it->Next();
         }
         ASSERT_EQ(expected_ub - targets[j] + 1, count);
+
+        count = 0;
+        expected_key = targets[j];
+        it->SeekForPrev(start_key);
+        uint64_t expected_lb = (targets[j] & kPrefixMask);
+        while (it->Valid()) {
+          // Out of prefix
+          if (!read_opts.prefix_same_as_start &&
+              pe->Transform(it->key()) != pe->Transform(start_key)) {
+            break;
+          }
+          CheckIterUserEntry(it.get(), Key1(expected_key), kTypeValue,
+                             "value" + std::to_string(i), write_ts_list[i]);
+          ++count;
+          --expected_key;
+          it->Prev();
+        }
+        ASSERT_EQ(targets[j] - std::max(expected_lb, kMinKey) + 1, count);
       }
     }
   }
@@ -2416,19 +2850,25 @@ INSTANTIATE_TEST_CASE_P(
                           std::shared_ptr<const FilterPolicy>(
                               NewBloomFilterPolicy(20 /*bits_per_key*/,
                                                    false))),
-        ::testing::Bool()));
+        ::testing::Bool(),
+        ::testing::Values(
+            BlockBasedTableOptions::IndexType::kBinarySearch,
+            BlockBasedTableOptions::IndexType::kHashSearch,
+            BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch,
+            BlockBasedTableOptions::IndexType::kBinarySearchWithFirstKey)));
 
 class DBBasicTestWithTsIterTombstones
     : public DBBasicTestWithTimestampBase,
       public testing::WithParamInterface<
           std::tuple<std::shared_ptr<const SliceTransform>,
-                     std::shared_ptr<const FilterPolicy>, int>> {
+                     std::shared_ptr<const FilterPolicy>, int,
+                     BlockBasedTableOptions::IndexType>> {
  public:
   DBBasicTestWithTsIterTombstones()
       : DBBasicTestWithTimestampBase("/db_basic_ts_iter_tombstones") {}
 };
 
-TEST_P(DBBasicTestWithTsIterTombstones, ForwardIterDelete) {
+TEST_P(DBBasicTestWithTsIterTombstones, IterWithDelete) {
   constexpr size_t kNumKeysPerFile = 128;
   Options options = CurrentOptions();
   options.env = env_;
@@ -2439,6 +2879,7 @@ TEST_P(DBBasicTestWithTsIterTombstones, ForwardIterDelete) {
   options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeysPerFile));
   BlockBasedTableOptions bbto;
   bbto.filter_policy = std::get<1>(GetParam());
+  bbto.index_type = std::get<3>(GetParam());
   options.table_factory.reset(NewBlockBasedTableFactory(bbto));
   options.num_levels = std::get<2>(GetParam());
   DestroyAndReopen(options);
@@ -2458,7 +2899,7 @@ TEST_P(DBBasicTestWithTsIterTombstones, ForwardIterDelete) {
     }
     ++key;
   } while (true);
-  // Delete them all
+
   ts = write_ts_strs[1];
   write_opts.timestamp = &ts;
   for (key = kMaxKey; key >= kMinKey; --key) {
@@ -2485,6 +2926,13 @@ TEST_P(DBBasicTestWithTsIterTombstones, ForwardIterDelete) {
       ASSERT_EQ("value1" + std::to_string(key), iter->value());
     }
     ASSERT_EQ((kMaxKey - kMinKey + 1) / 2, count);
+
+    for (iter->SeekToLast(), count = 0, key = kMaxKey; iter->Valid();
+         key -= 2, ++count, iter->Prev()) {
+      ASSERT_EQ(Key1(key), iter->key());
+      ASSERT_EQ("value1" + std::to_string(key), iter->value());
+    }
+    ASSERT_EQ((kMaxKey - kMinKey + 1) / 2, count);
   }
   Close();
 }
@@ -2500,7 +2948,12 @@ INSTANTIATE_TEST_CASE_P(
                               NewBloomFilterPolicy(10, false)),
                           std::shared_ptr<const FilterPolicy>(
                               NewBloomFilterPolicy(20, false))),
-        ::testing::Values(2, 6)));
+        ::testing::Values(2, 6),
+        ::testing::Values(
+            BlockBasedTableOptions::IndexType::kBinarySearch,
+            BlockBasedTableOptions::IndexType::kHashSearch,
+            BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch,
+            BlockBasedTableOptions::IndexType::kBinarySearchWithFirstKey)));
 
 }  // namespace ROCKSDB_NAMESPACE
 

db/dbformat.h

@@ -122,7 +122,7 @@ struct ParsedInternalKey {
 
   void SetTimestamp(const Slice& ts) {
     assert(ts.size() <= user_key.size());
-    const char* addr = user_key.data() - ts.size();
+    const char* addr = user_key.data() + user_key.size() - ts.size();
     memcpy(const_cast<char*>(addr), ts.data(), ts.size());
   }
 };
@@ -146,8 +146,10 @@ inline void UnPackSequenceAndType(uint64_t packed, uint64_t* seq,
   *seq = packed >> 8;
   *t = static_cast<ValueType>(packed & 0xff);
 
-  assert(*seq <= kMaxSequenceNumber);
-  assert(IsExtendedValueType(*t));
+  // Commented the following two assertions in order to test key-value checksum
+  // on corrupted keys without crashing ("DbKvChecksumTest").
+  // assert(*seq <= kMaxSequenceNumber);
+  // assert(IsExtendedValueType(*t));
 }
 
 EntryType GetEntryType(ValueType value_type);
@@ -510,6 +512,7 @@ class IterKey {
 
   bool IsKeyPinned() const { return (key_ != buf_); }
 
+  // user_key does not have timestamp.
   void SetInternalKey(const Slice& key_prefix, const Slice& user_key,
                       SequenceNumber s,
                       ValueType value_type = kValueTypeForSeek,

db/error_handler.cc

@@ -4,9 +4,11 @@
 //  (found in the LICENSE.Apache file in the root directory).
 //
 #include "db/error_handler.h"
+
 #include "db/db_impl/db_impl.h"
 #include "db/event_helpers.h"
 #include "file/sst_file_manager_impl.h"
+#include "logging/logging.h"
 
 namespace ROCKSDB_NAMESPACE {
 
@@ -274,6 +276,12 @@ const Status& ErrorHandler::SetBGError(const Status& bg_err,
     return bg_err;
   }
 
+  if (bg_error_stats_ != nullptr) {
+    RecordTick(bg_error_stats_.get(), ERROR_HANDLER_BG_ERROR_COUNT);
+  }
+  ROCKS_LOG_INFO(db_options_.info_log,
+                 "ErrorHandler: Set regular background error\n");
+
   bool paranoid = db_options_.paranoid_checks;
   Status::Severity sev = Status::Severity::kFatalError;
   Status new_bg_err;
@@ -399,6 +407,13 @@ const Status& ErrorHandler::SetBGError(const IOStatus& bg_io_err,
     if (recovery_in_prog_ && recovery_error_.ok()) {
       recovery_error_ = bg_err;
     }
+    if (bg_error_stats_ != nullptr) {
+      RecordTick(bg_error_stats_.get(), ERROR_HANDLER_BG_ERROR_COUNT);
+      RecordTick(bg_error_stats_.get(), ERROR_HANDLER_BG_IO_ERROR_COUNT);
+    }
+    ROCKS_LOG_INFO(
+        db_options_.info_log,
+        "ErrorHandler: Set background IO error as unrecoverable error\n");
     EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason,
                                           &bg_err, db_mutex_, &auto_recovery);
     recover_context_ = context;
@@ -416,12 +431,26 @@ const Status& ErrorHandler::SetBGError(const IOStatus& bg_io_err,
     EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason,
                                           &new_bg_io_err, db_mutex_,
                                           &auto_recovery);
+    if (bg_error_stats_ != nullptr) {
+      RecordTick(bg_error_stats_.get(), ERROR_HANDLER_BG_ERROR_COUNT);
+      RecordTick(bg_error_stats_.get(), ERROR_HANDLER_BG_IO_ERROR_COUNT);
+      RecordTick(bg_error_stats_.get(),
+                 ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT);
+    }
+    ROCKS_LOG_INFO(db_options_.info_log,
+                   "ErrorHandler: Set background retryable IO error\n");
     if (BackgroundErrorReason::kCompaction == reason) {
       // We map the retryable IO error during compaction to soft error. Since
       // compaction can reschedule by itself. We will not set the BG error in
       // this case
       // TODO:  a better way to set or clean the retryable IO error which
       // happens during compaction SST file write.
+      if (bg_error_stats_ != nullptr) {
+        RecordTick(bg_error_stats_.get(), ERROR_HANDLER_AUTORESUME_COUNT);
+      }
+      ROCKS_LOG_INFO(
+          db_options_.info_log,
+          "ErrorHandler: Compaction will schedule by itself to resume\n");
       return bg_error_;
     } else if (BackgroundErrorReason::kFlushNoWAL == reason ||
                BackgroundErrorReason::kManifestWriteNoWAL == reason) {
@@ -455,6 +484,9 @@ const Status& ErrorHandler::SetBGError(const IOStatus& bg_io_err,
       return StartRecoverFromRetryableBGIOError(bg_io_err);
     }
   } else {
+    if (bg_error_stats_ != nullptr) {
+      RecordTick(bg_error_stats_.get(), ERROR_HANDLER_BG_IO_ERROR_COUNT);
+    }
     return SetBGError(new_bg_io_err, reason);
   }
 }
@@ -603,7 +635,12 @@ const Status& ErrorHandler::StartRecoverFromRetryableBGIOError(
     // Auto resume BG error is not enabled, directly return bg_error_.
     return bg_error_;
   }
-
+  if (bg_error_stats_ != nullptr) {
+    RecordTick(bg_error_stats_.get(), ERROR_HANDLER_AUTORESUME_COUNT);
+  }
+  ROCKS_LOG_INFO(
+      db_options_.info_log,
+      "ErrorHandler: Call StartRecoverFromRetryableBGIOError to resume\n");
   if (recovery_thread_) {
     // In this case, if recovery_in_prog_ is false, current thread should
     // wait the previous recover thread to finish and create a new thread
@@ -642,6 +679,7 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
   DBRecoverContext context = recover_context_;
   int resume_count = db_options_.max_bgerror_resume_count;
   uint64_t wait_interval = db_options_.bgerror_resume_retry_interval;
+  uint64_t retry_count = 0;
   // Recover from the retryable error. Create a separate thread to do it.
   while (resume_count > 0) {
     if (end_recovery_) {
@@ -651,15 +689,24 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
     TEST_SYNC_POINT("RecoverFromRetryableBGIOError:BeforeResume1");
     recovery_io_error_ = IOStatus::OK();
     recovery_error_ = Status::OK();
+    retry_count++;
     Status s = db_->ResumeImpl(context);
     TEST_SYNC_POINT("RecoverFromRetryableBGIOError:AfterResume0");
     TEST_SYNC_POINT("RecoverFromRetryableBGIOError:AfterResume1");
+    if (bg_error_stats_ != nullptr) {
+      RecordTick(bg_error_stats_.get(),
+                 ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT);
+    }
     if (s.IsShutdownInProgress() ||
         bg_error_.severity() >= Status::Severity::kFatalError) {
       // If DB shutdown in progress or the error severity is higher than
       // Hard Error, stop auto resume and returns.
       TEST_SYNC_POINT("RecoverFromRetryableBGIOError:RecoverFail0");
       recovery_in_prog_ = false;
+      if (bg_error_stats_ != nullptr) {
+        RecordInHistogram(bg_error_stats_.get(),
+                          ERROR_HANDLER_AUTORESUME_RETRY_COUNT, retry_count);
+      }
       return;
     }
     if (!recovery_io_error_.ok() &&
@@ -670,7 +717,7 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
       // a period of time and redo auto resume if it is allowed.
       TEST_SYNC_POINT("RecoverFromRetryableBGIOError:BeforeWait0");
       TEST_SYNC_POINT("RecoverFromRetryableBGIOError:BeforeWait1");
-      int64_t wait_until = db_->env_->NowMicros() + wait_interval;
+      int64_t wait_until = db_options_.clock->NowMicros() + wait_interval;
       cv_.TimedWait(wait_until);
       TEST_SYNC_POINT("RecoverFromRetryableBGIOError:AfterWait0");
     } else {
@@ -686,6 +733,12 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
         bg_error_.PermitUncheckedError();
         EventHelpers::NotifyOnErrorRecoveryCompleted(db_options_.listeners,
                                                      old_bg_error, db_mutex_);
+        if (bg_error_stats_ != nullptr) {
+          RecordTick(bg_error_stats_.get(),
+                     ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT);
+          RecordInHistogram(bg_error_stats_.get(),
+                            ERROR_HANDLER_AUTORESUME_RETRY_COUNT, retry_count);
+        }
         recovery_in_prog_ = false;
         if (soft_error_no_bg_work_) {
           soft_error_no_bg_work_ = false;
@@ -696,6 +749,10 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
         // In this case: 1) recovery_io_error is more serious or not retryable
         // 2) other Non IO recovery_error happens. The auto recovery stops.
         recovery_in_prog_ = false;
+        if (bg_error_stats_ != nullptr) {
+          RecordInHistogram(bg_error_stats_.get(),
+                            ERROR_HANDLER_AUTORESUME_RETRY_COUNT, retry_count);
+        }
         return;
       }
     }
@@ -703,6 +760,10 @@ void ErrorHandler::RecoverFromRetryableBGIOError() {
   }
   recovery_in_prog_ = false;
   TEST_SYNC_POINT("RecoverFromRetryableBGIOError:LoopOut");
+  if (bg_error_stats_ != nullptr) {
+    RecordInHistogram(bg_error_stats_.get(),
+                      ERROR_HANDLER_AUTORESUME_RETRY_COUNT, retry_count);
+  }
   return;
 #else
   return;

db/error_handler.h

@@ -37,7 +37,8 @@ class ErrorHandler {
          db_mutex_(db_mutex),
          auto_recovery_(false),
          recovery_in_prog_(false),
-         soft_error_no_bg_work_(false) {
+         soft_error_no_bg_work_(false),
+         bg_error_stats_(db_options.statistics) {
      // Clear the checked flag for uninitialized errors
      bg_error_.PermitUncheckedError();
      recovery_error_.PermitUncheckedError();
@@ -108,6 +109,9 @@ class ErrorHandler {
     // Used to store the context for recover, such as flush reason.
     DBRecoverContext recover_context_;
 
+    // The pointer of DB statistics.
+    std::shared_ptr<Statistics> bg_error_stats_;
+
     Status OverrideNoSpaceError(const Status& bg_error, bool* auto_recovery);
     void RecoverFromNoSpace();
     const Status& StartRecoverFromRetryableBGIOError(const IOStatus& io_error);

db/error_handler_fs_test.cc

@@ -9,9 +9,9 @@
 #ifndef ROCKSDB_LITE
 
 #include "db/db_test_util.h"
+#include "file/sst_file_manager_impl.h"
 #include "port/stack_trace.h"
 #include "rocksdb/io_status.h"
-#include "rocksdb/perf_context.h"
 #include "rocksdb/sst_file_manager.h"
 #if !defined(ROCKSDB_LITE)
 #include "test_util/sync_point.h"
@@ -158,6 +158,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWriteError) {
   options.env = fault_env_.get();
   options.create_if_missing = true;
   options.listeners.emplace_back(listener);
+  options.statistics = CreateDBStatistics();
   Status s;
 
   listener->EnableAutoRecovery(false);
@@ -174,13 +175,25 @@ TEST_F(DBErrorHandlingFSTest, FLushWriteError) {
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
   ASSERT_OK(s);
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_IO_ERROR_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT));
 
   Reopen(options);
   ASSERT_EQ("val", Get(Key(0)));
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, FLushWritRetryableError) {
+TEST_F(DBErrorHandlingFSTest, FLushWriteRetryableError) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
   Options options = GetDefaultOptions();
@@ -188,6 +201,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableError) {
   options.create_if_missing = true;
   options.listeners.emplace_back(listener);
   options.max_bgerror_resume_count = 0;
+  options.statistics = CreateDBStatistics();
   Status s;
 
   listener->EnableAutoRecovery(false);
@@ -202,11 +216,23 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
   ASSERT_OK(s);
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_IO_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT));
   Reopen(options);
   ASSERT_EQ("val1", Get(Key(1)));
 
@@ -216,7 +242,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
@@ -230,7 +256,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
@@ -241,7 +267,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableError) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, FLushWritFileScopeError) {
+TEST_F(DBErrorHandlingFSTest, FLushWriteFileScopeError) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
   Options options = GetDefaultOptions();
@@ -266,7 +292,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritFileScopeError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
@@ -280,7 +306,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritFileScopeError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
@@ -294,7 +320,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritFileScopeError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
@@ -314,7 +340,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritFileScopeError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   s = dbfull()->Resume();
@@ -325,6 +351,95 @@ TEST_F(DBErrorHandlingFSTest, FLushWritFileScopeError) {
   Destroy(options);
 }
 
+TEST_F(DBErrorHandlingFSTest, FLushWALWriteRetryableError) {
+  std::shared_ptr<ErrorHandlerFSListener> listener(
+      new ErrorHandlerFSListener());
+  Options options = GetDefaultOptions();
+  options.env = fault_env_.get();
+  options.create_if_missing = true;
+  options.listeners.emplace_back(listener);
+  options.max_bgerror_resume_count = 0;
+  Status s;
+
+  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
+  error_msg.SetRetryable(true);
+
+  listener->EnableAutoRecovery(false);
+  SyncPoint::GetInstance()->SetCallBack(
+      "DBImpl::SyncClosedLogs:Start",
+      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  CreateAndReopenWithCF({"pikachu, sdfsdfsdf"}, options);
+
+  WriteOptions wo = WriteOptions();
+  wo.disableWAL = false;
+  ASSERT_OK(Put(Key(1), "val1", wo));
+
+  s = Flush();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  SyncPoint::GetInstance()->DisableProcessing();
+  fault_fs_->SetFilesystemActive(true);
+  auto cfh = dbfull()->GetColumnFamilyHandle(1);
+  s = dbfull()->DropColumnFamily(cfh);
+
+  s = dbfull()->Resume();
+  ASSERT_OK(s);
+  ASSERT_EQ("val1", Get(Key(1)));
+  ASSERT_OK(Put(Key(3), "val3", wo));
+  ASSERT_EQ("val3", Get(Key(3)));
+  s = Flush();
+  ASSERT_OK(s);
+  ASSERT_EQ("val3", Get(Key(3)));
+
+  Destroy(options);
+}
+
+TEST_F(DBErrorHandlingFSTest, FLushWALAtomicWriteRetryableError) {
+  std::shared_ptr<ErrorHandlerFSListener> listener(
+      new ErrorHandlerFSListener());
+  Options options = GetDefaultOptions();
+  options.env = fault_env_.get();
+  options.create_if_missing = true;
+  options.listeners.emplace_back(listener);
+  options.max_bgerror_resume_count = 0;
+  options.atomic_flush = true;
+  Status s;
+
+  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
+  error_msg.SetRetryable(true);
+
+  listener->EnableAutoRecovery(false);
+  SyncPoint::GetInstance()->SetCallBack(
+      "DBImpl::SyncClosedLogs:Start",
+      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
+  SyncPoint::GetInstance()->EnableProcessing();
+
+  CreateAndReopenWithCF({"pikachu, sdfsdfsdf"}, options);
+
+  WriteOptions wo = WriteOptions();
+  wo.disableWAL = false;
+  ASSERT_OK(Put(Key(1), "val1", wo));
+
+  s = Flush();
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  SyncPoint::GetInstance()->DisableProcessing();
+  fault_fs_->SetFilesystemActive(true);
+  auto cfh = dbfull()->GetColumnFamilyHandle(1);
+  s = dbfull()->DropColumnFamily(cfh);
+
+  s = dbfull()->Resume();
+  ASSERT_OK(s);
+  ASSERT_EQ("val1", Get(Key(1)));
+  ASSERT_OK(Put(Key(3), "val3", wo));
+  ASSERT_EQ("val3", Get(Key(3)));
+  s = Flush();
+  ASSERT_OK(s);
+  ASSERT_EQ("val3", Get(Key(3)));
+
+  Destroy(options);
+}
+
 TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError1) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
@@ -333,6 +448,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError1) {
   options.create_if_missing = true;
   options.listeners.emplace_back(listener);
   options.max_bgerror_resume_count = 0;
+  options.statistics = CreateDBStatistics();
   Status s;
 
   listener->EnableAutoRecovery(false);
@@ -363,11 +479,23 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError1) {
   s = Flush();
   ASSERT_OK(s);
   ASSERT_EQ("val3", Get(Key(3)));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_IO_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT));
 
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError2) {
+TEST_F(DBErrorHandlingFSTest, FLushWriteNoWALRetryableError2) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
   Options options = GetDefaultOptions();
@@ -410,7 +538,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError2) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableError3) {
+TEST_F(DBErrorHandlingFSTest, FLushWriteNoWALRetryableError3) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
   Options options = GetDefaultOptions();
@@ -521,7 +649,7 @@ TEST_F(DBErrorHandlingFSTest, ManifestWriteRetryableError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->ClearAllCallBacks();
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
@@ -567,7 +695,7 @@ TEST_F(DBErrorHandlingFSTest, ManifestWriteFileScopeError) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->ClearAllCallBacks();
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
@@ -864,7 +992,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteError) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableError) {
+TEST_F(DBErrorHandlingFSTest, DISABLED_CompactionWriteRetryableError) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
   Options options = GetDefaultOptions();
@@ -911,7 +1039,7 @@ TEST_F(DBErrorHandlingFSTest, CompactionWriteRetryableError) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, CompactionWriteFileScopeError) {
+TEST_F(DBErrorHandlingFSTest, DISABLED_CompactionWriteFileScopeError) {
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
   Options options = GetDefaultOptions();
@@ -1010,6 +1138,7 @@ TEST_F(DBErrorHandlingFSTest, AutoRecoverFlushError) {
   options.env = fault_env_.get();
   options.create_if_missing = true;
   options.listeners.emplace_back(listener);
+  options.statistics = CreateDBStatistics();
   Status s;
 
   listener->EnableAutoRecovery();
@@ -1028,6 +1157,18 @@ TEST_F(DBErrorHandlingFSTest, AutoRecoverFlushError) {
 
   s = Put(Key(1), "val");
   ASSERT_OK(s);
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_IO_ERROR_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT));
 
   Reopen(options);
   ASSERT_EQ("val", Get(Key(0)));
@@ -1396,6 +1537,10 @@ TEST_F(DBErrorHandlingFSTest, MultiDBCompactionError) {
     EXPECT_EQ(atoi(prop.c_str()), 1);
   }
 
+  SstFileManagerImpl* sfmImpl =
+      static_cast_with_check<SstFileManagerImpl>(sfm.get());
+  sfmImpl->Close();
+
   for (auto i = 0; i < kNumDbInstances; ++i) {
     char buf[16];
     snprintf(buf, sizeof(buf), "_%d", i);
@@ -1529,6 +1674,10 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {
     EXPECT_EQ(atoi(prop.c_str()), 1);
   }
 
+  SstFileManagerImpl* sfmImpl =
+      static_cast_with_check<SstFileManagerImpl>(sfm.get());
+  sfmImpl->Close();
+
   for (auto i = 0; i < kNumDbInstances; ++i) {
     char buf[16];
     snprintf(buf, sizeof(buf), "_%d", i);
@@ -1549,7 +1698,7 @@ TEST_F(DBErrorHandlingFSTest, MultiDBVariousErrors) {
 // to soft error and trigger auto resume. During auto resume, SwitchMemtable
 // is disabled to avoid small SST tables. Write can still be applied before
 // the bg error is cleaned unless the memtable is full.
-TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover1) {
+TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableErrorAutoRecover1) {
   // Activate the FS before the first resume
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
@@ -1559,6 +1708,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover1) {
   options.listeners.emplace_back(listener);
   options.max_bgerror_resume_count = 2;
   options.bgerror_resume_retry_interval = 100000;  // 0.1 second
+  options.statistics = CreateDBStatistics();
   Status s;
 
   listener->EnableAutoRecovery(false);
@@ -1586,6 +1736,22 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover1) {
   ASSERT_EQ("val1", Get(Key(1)));
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
+  ASSERT_EQ(3, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_ERROR_COUNT));
+  ASSERT_EQ(3, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_IO_ERROR_COUNT));
+  ASSERT_EQ(3, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_COUNT));
+  ASSERT_EQ(2, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT));
+  ASSERT_EQ(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT));
+  HistogramData autoresume_retry;
+  options.statistics->histogramData(ERROR_HANDLER_AUTORESUME_RETRY_COUNT,
+                                    &autoresume_retry);
+  ASSERT_EQ(autoresume_retry.max, 2);
   ASSERT_OK(Put(Key(2), "val2", wo));
   s = Flush();
   // Since auto resume fails, the bg error is not cleand, flush will
@@ -1602,7 +1768,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover1) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover2) {
+TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableErrorAutoRecover2) {
   // Activate the FS before the first resume
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
@@ -1612,6 +1778,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover2) {
   options.listeners.emplace_back(listener);
   options.max_bgerror_resume_count = 2;
   options.bgerror_resume_retry_interval = 100000;  // 0.1 second
+  options.statistics = CreateDBStatistics();
   Status s;
 
   listener->EnableAutoRecovery(false);
@@ -1635,6 +1802,22 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover2) {
   fault_fs_->SetFilesystemActive(true);
   ASSERT_EQ(listener->WaitForRecovery(5000000), true);
   ASSERT_EQ("val1", Get(Key(1)));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_IO_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_BG_RETRYABLE_IO_ERROR_COUNT));
+  ASSERT_EQ(1, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_COUNT));
+  ASSERT_LE(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_RETRY_TOTAL_COUNT));
+  ASSERT_LE(0, options.statistics->getAndResetTickerCount(
+                   ERROR_HANDLER_AUTORESUME_SUCCESS_COUNT));
+  HistogramData autoresume_retry;
+  options.statistics->histogramData(ERROR_HANDLER_AUTORESUME_RETRY_COUNT,
+                                    &autoresume_retry);
+  ASSERT_GE(autoresume_retry.max, 0);
   ASSERT_OK(Put(Key(2), "val2", wo));
   s = Flush();
   // Since auto resume is successful, the bg error is cleaned, flush will
@@ -1644,56 +1827,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritNoWALRetryableeErrorAutoRecover2) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover1) {
-  // Fail the first resume and make the second resume successful
-  std::shared_ptr<ErrorHandlerFSListener> listener(
-      new ErrorHandlerFSListener());
-  Options options = GetDefaultOptions();
-  options.env = fault_env_.get();
-  options.create_if_missing = true;
-  options.listeners.emplace_back(listener);
-  options.max_bgerror_resume_count = 2;
-  options.bgerror_resume_retry_interval = 100000;  // 0.1 second
-  Status s;
-
-  listener->EnableAutoRecovery(false);
-  DestroyAndReopen(options);
-
-  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
-  error_msg.SetRetryable(true);
-
-  ASSERT_OK(Put(Key(1), "val1"));
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
-      {{"RecoverFromRetryableBGIOError:BeforeWait0",
-        "FLushWritRetryableeErrorAutoRecover1:0"},
-       {"FLushWritRetryableeErrorAutoRecover1:1",
-        "RecoverFromRetryableBGIOError:BeforeWait1"},
-       {"RecoverFromRetryableBGIOError:RecoverSuccess",
-        "FLushWritRetryableeErrorAutoRecover1:2"}});
-  SyncPoint::GetInstance()->SetCallBack(
-      "BuildTable:BeforeFinishBuildTable",
-      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
-  SyncPoint::GetInstance()->EnableProcessing();
-  s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover1:0");
-  fault_fs_->SetFilesystemActive(true);
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover1:1");
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover1:2");
-  SyncPoint::GetInstance()->DisableProcessing();
-
-  ASSERT_EQ("val1", Get(Key(1)));
-  Reopen(options);
-  ASSERT_EQ("val1", Get(Key(1)));
-  ASSERT_OK(Put(Key(2), "val2"));
-  ASSERT_OK(Flush());
-  ASSERT_EQ("val2", Get(Key(2)));
-
-  Destroy(options);
-}
-
-TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover2) {
+TEST_F(DBErrorHandlingFSTest, FLushWritRetryableErrorAutoRecover1) {
   // Activate the FS before the first resume
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
@@ -1718,7 +1852,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover2) {
 
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   SyncPoint::GetInstance()->DisableProcessing();
   fault_fs_->SetFilesystemActive(true);
   ASSERT_EQ(listener->WaitForRecovery(5000000), true);
@@ -1733,7 +1867,7 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover2) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover3) {
+TEST_F(DBErrorHandlingFSTest, FLushWritRetryableErrorAutoRecover2) {
   // Fail all the resume and let user to resume
   std::shared_ptr<ErrorHandlerFSListener> listener(
       new ErrorHandlerFSListener());
@@ -1753,18 +1887,18 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover3) {
 
   ASSERT_OK(Put(Key(1), "val1"));
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
-      {{"FLushWritRetryableeErrorAutoRecover3:0",
+      {{"FLushWritRetryableeErrorAutoRecover2:0",
         "RecoverFromRetryableBGIOError:BeforeStart"},
        {"RecoverFromRetryableBGIOError:LoopOut",
-        "FLushWritRetryableeErrorAutoRecover3:1"}});
+        "FLushWritRetryableeErrorAutoRecover2:1"}});
   SyncPoint::GetInstance()->SetCallBack(
       "BuildTable:BeforeFinishBuildTable",
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover3:0");
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover3:1");
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
+  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover2:0");
+  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover2:1");
   fault_fs_->SetFilesystemActive(true);
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
   SyncPoint::GetInstance()->DisableProcessing();
@@ -1782,173 +1916,6 @@ TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover3) {
   Destroy(options);
 }
 
-TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover4) {
-  // Fail the first resume and does not do resume second time because
-  // the IO error severity is Fatal Error and not Retryable.
-  std::shared_ptr<ErrorHandlerFSListener> listener(
-      new ErrorHandlerFSListener());
-  Options options = GetDefaultOptions();
-  options.env = fault_env_.get();
-  options.create_if_missing = true;
-  options.listeners.emplace_back(listener);
-  options.max_bgerror_resume_count = 2;
-  options.bgerror_resume_retry_interval = 10;  // 0.1 second
-  Status s;
-
-  listener->EnableAutoRecovery(false);
-  DestroyAndReopen(options);
-
-  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
-  error_msg.SetRetryable(true);
-  IOStatus nr_msg = IOStatus::IOError("No Retryable Fatal IO Error");
-  nr_msg.SetRetryable(false);
-
-  ASSERT_OK(Put(Key(1), "val1"));
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
-      {{"RecoverFromRetryableBGIOError:BeforeStart",
-        "FLushWritRetryableeErrorAutoRecover4:0"},
-       {"FLushWritRetryableeErrorAutoRecover4:2",
-        "RecoverFromRetryableBGIOError:RecoverFail0"}});
-  SyncPoint::GetInstance()->SetCallBack(
-      "BuildTable:BeforeFinishBuildTable",
-      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
-  SyncPoint::GetInstance()->SetCallBack(
-      "RecoverFromRetryableBGIOError:BeforeResume1",
-      [&](void*) { fault_fs_->SetFilesystemActive(false, nr_msg); });
-
-  SyncPoint::GetInstance()->EnableProcessing();
-  s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover4:0");
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover4:2");
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
-  SyncPoint::GetInstance()->DisableProcessing();
-  fault_fs_->SetFilesystemActive(true);
-  // Even the FS is recoverd, due to the Fatal Error in bg_error_ the resume
-  // and flush will all fail.
-  ASSERT_EQ("val1", Get(Key(1)));
-  ASSERT_NOK(dbfull()->Resume());
-  ASSERT_EQ("val1", Get(Key(1)));
-  ASSERT_OK(Put(Key(2), "val2"));
-  ASSERT_NOK(Flush());
-  ASSERT_EQ("NOT_FOUND", Get(Key(2)));
-
-  Reopen(options);
-  ASSERT_EQ("val1", Get(Key(1)));
-  ASSERT_OK(Put(Key(2), "val2"));
-  ASSERT_OK(Flush());
-  ASSERT_EQ("val2", Get(Key(2)));
-
-  Destroy(options);
-}
-
-TEST_F(DBErrorHandlingFSTest, DISABLED_FLushWritRetryableeErrorAutoRecover5) {
-  // During the resume, call DB->CLose, make sure the resume thread exist
-  // before close continues. Due to the shutdown, the resume is not successful
-  // and the FS does not become active, so close status is still IO error
-  std::shared_ptr<ErrorHandlerFSListener> listener(
-      new ErrorHandlerFSListener());
-  Options options = GetDefaultOptions();
-  options.env = fault_env_.get();
-  options.create_if_missing = true;
-  options.listeners.emplace_back(listener);
-  options.max_bgerror_resume_count = 2;
-  options.bgerror_resume_retry_interval = 10;  // 0.1 second
-  Status s;
-
-  listener->EnableAutoRecovery(false);
-  DestroyAndReopen(options);
-
-  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
-  error_msg.SetRetryable(true);
-
-  ASSERT_OK(Put(Key(1), "val1"));
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
-      {{"RecoverFromRetryableBGIOError:BeforeStart",
-        "FLushWritRetryableeErrorAutoRecover5:0"}});
-  SyncPoint::GetInstance()->SetCallBack(
-      "BuildTable:BeforeFinishBuildTable",
-      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
-  SyncPoint::GetInstance()->EnableProcessing();
-  s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover5:0");
-  // The first resume will cause recovery_error and its severity is the
-  // Fatal error
-  s = dbfull()->Close();
-  ASSERT_NOK(s);
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
-  SyncPoint::GetInstance()->DisableProcessing();
-  fault_fs_->SetFilesystemActive(true);
-
-  Reopen(options);
-  ASSERT_NE("val1", Get(Key(1)));
-  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_OK(s);
-  ASSERT_EQ("val2", Get(Key(2)));
-
-  Destroy(options);
-}
-
-TEST_F(DBErrorHandlingFSTest, FLushWritRetryableeErrorAutoRecover6) {
-  // During the resume, call DB->CLose, make sure the resume thread exist
-  // before close continues. Due to the shutdown, the resume is not successful
-  // and the FS does not become active, so close status is still IO error
-  std::shared_ptr<ErrorHandlerFSListener> listener(
-      new ErrorHandlerFSListener());
-  Options options = GetDefaultOptions();
-  options.env = fault_env_.get();
-  options.create_if_missing = true;
-  options.listeners.emplace_back(listener);
-  options.max_bgerror_resume_count = 2;
-  options.bgerror_resume_retry_interval = 10;  // 0.1 second
-  Status s;
-
-  listener->EnableAutoRecovery(false);
-  DestroyAndReopen(options);
-
-  IOStatus error_msg = IOStatus::IOError("Retryable IO Error");
-  error_msg.SetRetryable(true);
-
-  ASSERT_OK(Put(Key(1), "val1"));
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
-      {{"FLushWritRetryableeErrorAutoRecover6:0",
-        "RecoverFromRetryableBGIOError:BeforeStart"},
-       {"RecoverFromRetryableBGIOError:BeforeWait0",
-        "FLushWritRetryableeErrorAutoRecover6:1"},
-       {"FLushWritRetryableeErrorAutoRecover6:2",
-        "RecoverFromRetryableBGIOError:BeforeWait1"},
-       {"RecoverFromRetryableBGIOError:AfterWait0",
-        "FLushWritRetryableeErrorAutoRecover6:3"}});
-  SyncPoint::GetInstance()->SetCallBack(
-      "BuildTable:BeforeFinishBuildTable",
-      [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
-  SyncPoint::GetInstance()->EnableProcessing();
-  s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover6:0");
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover6:1");
-  fault_fs_->SetFilesystemActive(true);
-  ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover6:2");
-  TEST_SYNC_POINT("FLushWritRetryableeErrorAutoRecover6:3");
-  // The first resume will cause recovery_error and its severity is the
-  // Fatal error
-  s = dbfull()->Close();
-  ASSERT_OK(s);
-  SyncPoint::GetInstance()->DisableProcessing();
-
-  Reopen(options);
-  ASSERT_EQ("val1", Get(Key(1)));
-  ASSERT_OK(Put(Key(2), "val2"));
-  s = Flush();
-  ASSERT_OK(s);
-  ASSERT_EQ("val2", Get(Key(2)));
-
-  Destroy(options);
-}
-
 TEST_F(DBErrorHandlingFSTest, ManifestWriteRetryableErrorAutoRecover) {
   // Fail the first resume and let the second resume be successful
   std::shared_ptr<ErrorHandlerFSListener> listener(
@@ -1985,7 +1952,7 @@ TEST_F(DBErrorHandlingFSTest, ManifestWriteRetryableErrorAutoRecover) {
       [&](void*) { fault_fs_->SetFilesystemActive(false, error_msg); });
   SyncPoint::GetInstance()->EnableProcessing();
   s = Flush();
-  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), ROCKSDB_NAMESPACE::Status::Severity::kSoftError);
   TEST_SYNC_POINT("ManifestWriteRetryableErrorAutoRecover:0");
   fault_fs_->SetFilesystemActive(true);
   ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->ClearAllCallBacks();
@@ -2244,7 +2211,8 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableErrorAutoRecover1) {
       ASSERT_OK(batch.Put(Key(i), rnd.RandomString(1024)));
     }
     ROCKSDB_NAMESPACE::SyncPoint::GetInstance()->LoadDependency(
-        {{"RecoverFromRetryableBGIOError:BeforeResume0", "WALWriteError1:0"},
+        {{"WALWriteErrorDone", "RecoverFromRetryableBGIOError:BeforeStart"},
+         {"RecoverFromRetryableBGIOError:BeforeResume0", "WALWriteError1:0"},
          {"WALWriteError1:1", "RecoverFromRetryableBGIOError:BeforeResume1"},
          {"RecoverFromRetryableBGIOError:RecoverSuccess", "WALWriteError1:2"}});
 
@@ -2260,6 +2228,7 @@ TEST_F(DBErrorHandlingFSTest, WALWriteRetryableErrorAutoRecover1) {
     wopts.sync = true;
     s = dbfull()->Write(wopts, &batch);
     ASSERT_EQ(true, s.IsIOError());
+    TEST_SYNC_POINT("WALWriteErrorDone");
 
     TEST_SYNC_POINT("WALWriteError1:0");
     fault_fs_->SetFilesystemActive(true);

db/external_sst_file_basic_test.cc

@@ -1542,6 +1542,44 @@ TEST_F(ExternalSSTFileBasicTest, OverlappingFiles) {
   ASSERT_EQ(2, NumTableFilesAtLevel(0));
 }
 
+TEST_F(ExternalSSTFileBasicTest, IngestFileAfterDBPut) {
+  // Repro https://github.com/facebook/rocksdb/issues/6245.
+  // Flush three files to L0. Ingest one more file to trigger L0->L1 compaction
+  // via trivial move. The bug happened when L1 files were incorrectly sorted
+  // resulting in an old value for "k" returned by `Get()`.
+  Options options = CurrentOptions();
+
+  ASSERT_OK(Put("k", "a"));
+  Flush();
+  ASSERT_OK(Put("k", "a"));
+  Flush();
+  ASSERT_OK(Put("k", "a"));
+  Flush();
+  SstFileWriter sst_file_writer(EnvOptions(), options);
+
+  // Current file size should be 0 after sst_file_writer init and before open a
+  // file.
+  ASSERT_EQ(sst_file_writer.FileSize(), 0);
+
+  std::string file1 = sst_files_dir_ + "file1.sst";
+  ASSERT_OK(sst_file_writer.Open(file1));
+  ASSERT_OK(sst_file_writer.Put("k", "b"));
+
+  ExternalSstFileInfo file1_info;
+  Status s = sst_file_writer.Finish(&file1_info);
+  ASSERT_OK(s) << s.ToString();
+
+  // Current file size should be non-zero after success write.
+  ASSERT_GT(sst_file_writer.FileSize(), 0);
+
+  IngestExternalFileOptions ifo;
+  s = db_->IngestExternalFile({file1}, ifo);
+  ASSERT_OK(s);
+  ASSERT_OK(dbfull()->TEST_WaitForCompact());
+
+  ASSERT_EQ(Get("k"), "b");
+}
+
 INSTANTIATE_TEST_CASE_P(ExternalSSTFileBasicTest, ExternalSSTFileBasicTest,
                         testing::Values(std::make_tuple(true, true),
                                         std::make_tuple(true, false),

db/external_sst_file_ingestion_job.cc

@@ -293,12 +293,13 @@ Status ExternalSstFileIngestionJob::Prepare(
 
   // TODO: The following is duplicated with Cleanup().
   if (!status.ok()) {
+    IOOptions io_opts;
     // We failed, remove all files that we copied into the db
     for (IngestedFileInfo& f : files_to_ingest_) {
       if (f.internal_file_path.empty()) {
         continue;
       }
-      Status s = env_->DeleteFile(f.internal_file_path);
+      Status s = fs_->DeleteFile(f.internal_file_path, io_opts, nullptr);
       if (!s.ok()) {
         ROCKS_LOG_WARN(db_options_.info_log,
                        "AddFile() clean up for file %s failed : %s",
@@ -367,9 +368,32 @@ Status ExternalSstFileIngestionJob::Run() {
           super_version, force_global_seqno, cfd_->ioptions()->compaction_style,
           last_seqno, &f, &assigned_seqno);
     }
+
+    // Modify the smallest/largest internal key to include the sequence number
+    // that we just learned. Only overwrite sequence number zero. There could
+    // be a nonzero sequence number already to indicate a range tombstone's
+    // exclusive endpoint.
+    ParsedInternalKey smallest_parsed, largest_parsed;
+    if (status.ok()) {
+      status = ParseInternalKey(*f.smallest_internal_key.rep(),
+                                &smallest_parsed, false /* log_err_key */);
+    }
+    if (status.ok()) {
+      status = ParseInternalKey(*f.largest_internal_key.rep(), &largest_parsed,
+                                false /* log_err_key */);
+    }
     if (!status.ok()) {
       return status;
     }
+    if (smallest_parsed.sequence == 0) {
+      UpdateInternalKey(f.smallest_internal_key.rep(), assigned_seqno,
+                        smallest_parsed.type);
+    }
+    if (largest_parsed.sequence == 0) {
+      UpdateInternalKey(f.largest_internal_key.rep(), assigned_seqno,
+                        largest_parsed.type);
+    }
+
     status = AssignGlobalSeqnoForIngestedFile(&f, assigned_seqno);
     TEST_SYNC_POINT_CALLBACK("ExternalSstFileIngestionJob::Run",
                              &assigned_seqno);
@@ -392,7 +416,7 @@ Status ExternalSstFileIngestionJob::Run() {
     int64_t temp_current_time = 0;
     uint64_t current_time = kUnknownFileCreationTime;
     uint64_t oldest_ancester_time = kUnknownOldestAncesterTime;
-    if (env_->GetCurrentTime(&temp_current_time).ok()) {
+    if (clock_->GetCurrentTime(&temp_current_time).ok()) {
       current_time = oldest_ancester_time =
           static_cast<uint64_t>(temp_current_time);
     }
@@ -410,7 +434,7 @@ void ExternalSstFileIngestionJob::UpdateStats() {
   // Update internal stats for new ingested files
   uint64_t total_keys = 0;
   uint64_t total_l0_files = 0;
-  uint64_t total_time = env_->NowMicros() - job_start_time_;
+  uint64_t total_time = clock_->NowMicros() - job_start_time_;
 
   EventLoggerStream stream = event_logger_->Log();
   stream << "event"
@@ -466,6 +490,7 @@ void ExternalSstFileIngestionJob::UpdateStats() {
 }
 
 void ExternalSstFileIngestionJob::Cleanup(const Status& status) {
+  IOOptions io_opts;
   if (!status.ok()) {
     // We failed to add the files to the database
     // remove all the files we copied
@@ -473,7 +498,7 @@ void ExternalSstFileIngestionJob::Cleanup(const Status& status) {
       if (f.internal_file_path.empty()) {
         continue;
       }
-      Status s = env_->DeleteFile(f.internal_file_path);
+      Status s = fs_->DeleteFile(f.internal_file_path, io_opts, nullptr);
       if (!s.ok()) {
         ROCKS_LOG_WARN(db_options_.info_log,
                        "AddFile() clean up for file %s failed : %s",
@@ -485,7 +510,7 @@ void ExternalSstFileIngestionJob::Cleanup(const Status& status) {
   } else if (status.ok() && ingestion_options_.move_files) {
     // The files were moved and added successfully, remove original file links
     for (IngestedFileInfo& f : files_to_ingest_) {
-      Status s = env_->DeleteFile(f.external_file_path);
+      Status s = fs_->DeleteFile(f.external_file_path, io_opts, nullptr);
       if (!s.ok()) {
         ROCKS_LOG_WARN(
             db_options_.info_log,
@@ -810,7 +835,8 @@ Status ExternalSstFileIngestionJob::AssignGlobalSeqnoForIngestedFile(
         fs_->NewRandomRWFile(file_to_ingest->internal_file_path, env_options_,
                              &rwfile, nullptr);
     if (status.ok()) {
-      FSRandomRWFilePtr fsptr(std::move(rwfile), io_tracer_);
+      FSRandomRWFilePtr fsptr(std::move(rwfile), io_tracer_,
+                              file_to_ingest->internal_file_path);
       std::string seqno_val;
       PutFixed64(&seqno_val, seqno);
       status = fsptr->Write(file_to_ingest->global_seqno_offset, seqno_val,

db/external_sst_file_ingestion_job.h

@@ -16,13 +16,14 @@
 #include "logging/event_logger.h"
 #include "options/db_options.h"
 #include "rocksdb/db.h"
-#include "rocksdb/env.h"
+#include "rocksdb/file_system.h"
 #include "rocksdb/sst_file_writer.h"
 #include "util/autovector.h"
 
 namespace ROCKSDB_NAMESPACE {
 
 class Directories;
+class SystemClock;
 
 struct IngestedFileInfo {
   // External file path
@@ -73,13 +74,13 @@ struct IngestedFileInfo {
 class ExternalSstFileIngestionJob {
  public:
   ExternalSstFileIngestionJob(
-      Env* env, VersionSet* versions, ColumnFamilyData* cfd,
+      VersionSet* versions, ColumnFamilyData* cfd,
       const ImmutableDBOptions& db_options, const EnvOptions& env_options,
       SnapshotList* db_snapshots,
       const IngestExternalFileOptions& ingestion_options,
       Directories* directories, EventLogger* event_logger,
       const std::shared_ptr<IOTracer>& io_tracer)
-      : env_(env),
+      : clock_(db_options.clock),
         fs_(db_options.fs, io_tracer),
         versions_(versions),
         cfd_(cfd),
@@ -89,7 +90,7 @@ class ExternalSstFileIngestionJob {
         ingestion_options_(ingestion_options),
         directories_(directories),
         event_logger_(event_logger),
-        job_start_time_(env_->NowMicros()),
+        job_start_time_(clock_->NowMicros()),
         consumed_seqno_count_(0),
         io_tracer_(io_tracer) {
     assert(directories != nullptr);
@@ -169,7 +170,7 @@ class ExternalSstFileIngestionJob {
   template <typename TWritableFile>
   Status SyncIngestedFile(TWritableFile* file);
 
-  Env* env_;
+  SystemClock* clock_;
   FileSystemPtr fs_;
   VersionSet* versions_;
   ColumnFamilyData* cfd_;

db/flush_job.cc

@@ -80,24 +80,22 @@ const char* GetFlushReasonString (FlushReason flush_reason) {
   }
 }
 
-FlushJob::FlushJob(const std::string& dbname, ColumnFamilyData* cfd,
-                   const ImmutableDBOptions& db_options,
-                   const MutableCFOptions& mutable_cf_options,
-                   uint64_t max_memtable_id, const FileOptions& file_options,
-                   VersionSet* versions, InstrumentedMutex* db_mutex,
-                   std::atomic<bool>* shutting_down,
-                   std::vector<SequenceNumber> existing_snapshots,
-                   SequenceNumber earliest_write_conflict_snapshot,
-                   SnapshotChecker* snapshot_checker, JobContext* job_context,
-                   LogBuffer* log_buffer, FSDirectory* db_directory,
-                   FSDirectory* output_file_directory,
-                   CompressionType output_compression, Statistics* stats,
-                   EventLogger* event_logger, bool measure_io_stats,
-                   const bool sync_output_directory, const bool write_manifest,
-                   Env::Priority thread_pri,
-                   const std::shared_ptr<IOTracer>& io_tracer,
-                   const std::string& db_id, const std::string& db_session_id,
-                   std::string full_history_ts_low)
+FlushJob::FlushJob(
+    const std::string& dbname, ColumnFamilyData* cfd,
+    const ImmutableDBOptions& db_options,
+    const MutableCFOptions& mutable_cf_options, uint64_t max_memtable_id,
+    const FileOptions& file_options, VersionSet* versions,
+    InstrumentedMutex* db_mutex, std::atomic<bool>* shutting_down,
+    std::vector<SequenceNumber> existing_snapshots,
+    SequenceNumber earliest_write_conflict_snapshot,
+    SnapshotChecker* snapshot_checker, JobContext* job_context,
+    LogBuffer* log_buffer, FSDirectory* db_directory,
+    FSDirectory* output_file_directory, CompressionType output_compression,
+    Statistics* stats, EventLogger* event_logger, bool measure_io_stats,
+    const bool sync_output_directory, const bool write_manifest,
+    Env::Priority thread_pri, const std::shared_ptr<IOTracer>& io_tracer,
+    const std::string& db_id, const std::string& db_session_id,
+    std::string full_history_ts_low, BlobFileCompletionCallback* blob_callback)
     : dbname_(dbname),
       db_id_(db_id),
       db_session_id_(db_session_id),
@@ -127,7 +125,9 @@ FlushJob::FlushJob(const std::string& dbname, ColumnFamilyData* cfd,
       pick_memtable_called(false),
       thread_pri_(thread_pri),
       io_tracer_(io_tracer),
-      full_history_ts_low_(std::move(full_history_ts_low)) {
+      clock_(db_options_.clock),
+      full_history_ts_low_(std::move(full_history_ts_low)),
+      blob_callback_(blob_callback) {
   // Update the thread status to indicate flush.
   ReportStartedFlush();
   TEST_SYNC_POINT("FlushJob::FlushJob()");
@@ -309,8 +309,8 @@ Status FlushJob::WriteLevel0Table() {
   AutoThreadOperationStageUpdater stage_updater(
       ThreadStatus::STAGE_FLUSH_WRITE_L0);
   db_mutex_->AssertHeld();
-  const uint64_t start_micros = db_options_.env->NowMicros();
-  const uint64_t start_cpu_micros = db_options_.env->NowCPUNanos() / 1000;
+  const uint64_t start_micros = clock_->NowMicros();
+  const uint64_t start_cpu_micros = clock_->CPUNanos() / 1000;
   Status s;
 
   std::vector<BlobFileAddition> blob_file_additions;
@@ -371,7 +371,7 @@ Status FlushJob::WriteLevel0Table() {
       TEST_SYNC_POINT_CALLBACK("FlushJob::WriteLevel0Table:output_compression",
                                &output_compression_);
       int64_t _current_time = 0;
-      auto status = db_options_.env->GetCurrentTime(&_current_time);
+      auto status = clock_->GetCurrentTime(&_current_time);
       // Safe to proceed even if GetCurrentTime fails. So, log and proceed.
       if (!status.ok()) {
         ROCKS_LOG_WARN(
@@ -417,7 +417,7 @@ Status FlushJob::WriteLevel0Table() {
           TableFileCreationReason::kFlush, &io_s, io_tracer_, event_logger_,
           job_context_->job_id, Env::IO_HIGH, &table_properties_, 0 /* level */,
           creation_time, oldest_key_time, write_hint, current_time, db_id_,
-          db_session_id_, full_history_ts_low);
+          db_session_id_, full_history_ts_low, blob_callback_);
       if (!io_s.ok()) {
         io_status_ = io_s;
       }
@@ -466,8 +466,8 @@ Status FlushJob::WriteLevel0Table() {
 
   // Note that here we treat flush as level 0 compaction in internal stats
   InternalStats::CompactionStats stats(CompactionReason::kFlush, 1);
-  stats.micros = db_options_.env->NowMicros() - start_micros;
-  stats.cpu_micros = db_options_.env->NowCPUNanos() / 1000 - start_cpu_micros;
+  stats.micros = clock_->NowMicros() - start_micros;
+  stats.cpu_micros = clock_->CPUNanos() / 1000 - start_cpu_micros;
 
   if (has_output) {
     stats.bytes_written = meta_.fd.GetFileSize();
@@ -476,15 +476,16 @@ Status FlushJob::WriteLevel0Table() {
 
   const auto& blobs = edit_->GetBlobFileAdditions();
   for (const auto& blob : blobs) {
-    stats.bytes_written += blob.GetTotalBlobBytes();
+    stats.bytes_written_blob += blob.GetTotalBlobBytes();
   }
 
-  stats.num_output_files += static_cast<int>(blobs.size());
+  stats.num_output_files_blob = static_cast<int>(blobs.size());
 
   RecordTimeToHistogram(stats_, FLUSH_TIME, stats.micros);
   cfd_->internal_stats()->AddCompactionStats(0 /* level */, thread_pri_, stats);
-  cfd_->internal_stats()->AddCFStats(InternalStats::BYTES_FLUSHED,
-                                     stats.bytes_written);
+  cfd_->internal_stats()->AddCFStats(
+      InternalStats::BYTES_FLUSHED,
+      stats.bytes_written + stats.bytes_written_blob);
   RecordFlushIOStats();
   return s;
 }

db/flush_job.h

@@ -17,6 +17,7 @@
 #include <utility>
 #include <vector>
 
+#include "db/blob/blob_file_completion_callback.h"
 #include "db/column_family.h"
 #include "db/dbformat.h"
 #include "db/flush_scheduler.h"
@@ -73,7 +74,8 @@ class FlushJob {
            const bool sync_output_directory, const bool write_manifest,
            Env::Priority thread_pri, const std::shared_ptr<IOTracer>& io_tracer,
            const std::string& db_id = "", const std::string& db_session_id = "",
-           std::string full_history_ts_low = "");
+           std::string full_history_ts_low = "",
+           BlobFileCompletionCallback* blob_callback = nullptr);
 
   ~FlushJob();
 
@@ -162,8 +164,10 @@ class FlushJob {
   IOStatus io_status_;
 
   const std::shared_ptr<IOTracer> io_tracer_;
+  SystemClock* clock_;
 
   const std::string full_history_ts_low_;
+  BlobFileCompletionCallback* blob_callback_;
 };
 
 }  // namespace ROCKSDB_NAMESPACE

db/flush_job_test.cc

@@ -16,6 +16,7 @@
 #include "db/version_set.h"
 #include "file/writable_file_writer.h"
 #include "rocksdb/cache.h"
+#include "rocksdb/file_system.h"
 #include "rocksdb/write_buffer_manager.h"
 #include "table/mock_table.h"
 #include "test_util/testharness.h"
@@ -47,6 +48,8 @@ class FlushJobTestBase : public testing::Test {
     if (getenv("KEEP_DB")) {
       fprintf(stdout, "db is still in %s\n", dbname_.c_str());
     } else {
+      // destroy versions_ to release all file handles
+      versions_.reset();
       EXPECT_OK(DestroyDir(env_, dbname_));
     }
   }
@@ -74,12 +77,13 @@ class FlushJobTestBase : public testing::Test {
     }
 
     const std::string manifest = DescriptorFileName(dbname_, 1);
-    std::unique_ptr<WritableFile> file;
-    Status s = env_->NewWritableFile(
-        manifest, &file, env_->OptimizeForManifestWrite(env_options_));
+    const auto& fs = env_->GetFileSystem();
+    std::unique_ptr<WritableFileWriter> file_writer;
+    Status s = WritableFileWriter::Create(
+        fs, manifest, fs->OptimizeForManifestWrite(env_options_), &file_writer,
+        nullptr);
     ASSERT_OK(s);
-    std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-        NewLegacyWritableFileWrapper(std::move(file)), manifest, EnvOptions()));
+
     {
       log::Writer log(std::move(file_writer), 0, false);
       std::string record;
@@ -188,7 +192,8 @@ TEST_F(FlushJobTest, NonEmpty) {
   for (int i = 1; i < 10000; ++i) {
     std::string key(ToString((i + 1000) % 10000));
     std::string value("value" + key);
-    ASSERT_OK(new_mem->Add(SequenceNumber(i), kTypeValue, key, value));
+    ASSERT_OK(new_mem->Add(SequenceNumber(i), kTypeValue, key, value,
+                           nullptr /* kv_prot_info */));
     if ((i + 1000) % 10000 < 9995) {
       InternalKey internal_key(key, SequenceNumber(i), kTypeValue);
       inserted_keys.push_back({internal_key.Encode().ToString(), value});
@@ -197,7 +202,7 @@ TEST_F(FlushJobTest, NonEmpty) {
 
   {
     ASSERT_OK(new_mem->Add(SequenceNumber(10000), kTypeRangeDeletion, "9995",
-                           "9999a"));
+                           "9999a", nullptr /* kv_prot_info */));
     InternalKey internal_key("9995", SequenceNumber(10000), kTypeRangeDeletion);
     inserted_keys.push_back({internal_key.Encode().ToString(), "9999a"});
   }
@@ -224,7 +229,8 @@ TEST_F(FlushJobTest, NonEmpty) {
     }
 
     const SequenceNumber seq(i + 10001);
-    ASSERT_OK(new_mem->Add(seq, kTypeBlobIndex, key, blob_index));
+    ASSERT_OK(new_mem->Add(seq, kTypeBlobIndex, key, blob_index,
+                           nullptr /* kv_prot_info */));
 
     InternalKey internal_key(key, seq, kTypeBlobIndex);
     inserted_keys.push_back({internal_key.Encode().ToString(), blob_index});
@@ -286,7 +292,7 @@ TEST_F(FlushJobTest, FlushMemTablesSingleColumnFamily) {
       std::string key(ToString(j + i * num_keys_per_table));
       std::string value("value" + key);
       ASSERT_OK(mem->Add(SequenceNumber(j + i * num_keys_per_table), kTypeValue,
-                         key, value));
+                         key, value, nullptr /* kv_prot_info */));
     }
   }
 
@@ -358,7 +364,8 @@ TEST_F(FlushJobTest, FlushMemtablesMultipleColumnFamilies) {
       for (size_t j = 0; j != num_keys_per_memtable; ++j) {
         std::string key(ToString(j + i * num_keys_per_memtable));
         std::string value("value" + key);
-        ASSERT_OK(mem->Add(curr_seqno++, kTypeValue, key, value));
+        ASSERT_OK(mem->Add(curr_seqno++, kTypeValue, key, value,
+                           nullptr /* kv_prot_info */));
       }
 
       cfd->imm()->Add(mem, &to_delete);
@@ -469,7 +476,8 @@ TEST_F(FlushJobTest, Snapshots) {
     for (int j = 0; j < insertions; ++j) {
       std::string value(rnd.HumanReadableString(10));
       auto seqno = ++current_seqno;
-      ASSERT_OK(new_mem->Add(SequenceNumber(seqno), kTypeValue, key, value));
+      ASSERT_OK(new_mem->Add(SequenceNumber(seqno), kTypeValue, key, value,
+                             nullptr /* kv_prot_info */));
       // a key is visible only if:
       // 1. it's the last one written (j == insertions - 1)
       // 2. there's a snapshot pointing at it
@@ -521,7 +529,8 @@ class FlushJobTimestampTest : public FlushJobTestBase {
                              Slice value) {
     std::string key_str(std::move(key));
     PutFixed64(&key_str, ts);
-    ASSERT_OK(memtable->Add(seq, value_type, key_str, value));
+    ASSERT_OK(memtable->Add(seq, value_type, key_str, value,
+                            nullptr /* kv_prot_info */));
   }
 
  protected:

db/flush_scheduler.h

@@ -5,10 +5,11 @@
 
 #pragma once
 
-#include <stdint.h>
 #include <atomic>
+#include <cstdint>
 #include <mutex>
 #include <set>
+
 #include "util/autovector.h"
 
 namespace ROCKSDB_NAMESPACE {

db/forward_iterator.cc

@@ -993,9 +993,9 @@ bool ForwardIterator::TEST_CheckDeletedIters(int* pdeleted_iters,
 uint32_t ForwardIterator::FindFileInRange(
     const std::vector<FileMetaData*>& files, const Slice& internal_key,
     uint32_t left, uint32_t right) {
-  auto cmp = [&](const FileMetaData* f, const Slice& key) -> bool {
+  auto cmp = [&](const FileMetaData* f, const Slice& k) -> bool {
     return cfd_->internal_comparator().InternalKeyComparator::Compare(
-            f->largest.Encode(), key) < 0;
+            f->largest.Encode(), k) < 0;
   };
   const auto &b = files.begin();
   return static_cast<uint32_t>(std::lower_bound(b + left,

db/import_column_family_job.cc

@@ -140,7 +140,7 @@ Status ImportColumnFamilyJob::Run() {
   int64_t temp_current_time = 0;
   uint64_t oldest_ancester_time = kUnknownOldestAncesterTime;
   uint64_t current_time = kUnknownOldestAncesterTime;
-  if (env_->GetCurrentTime(&temp_current_time).ok()) {
+  if (clock_->GetCurrentTime(&temp_current_time).ok()) {
     current_time = oldest_ancester_time =
         static_cast<uint64_t>(temp_current_time);
   }

db/import_column_family_job.h

@@ -9,24 +9,25 @@
 #include "db/snapshot_impl.h"
 #include "options/db_options.h"
 #include "rocksdb/db.h"
-#include "rocksdb/env.h"
 #include "rocksdb/metadata.h"
 #include "rocksdb/sst_file_writer.h"
 #include "util/autovector.h"
 
 namespace ROCKSDB_NAMESPACE {
+struct EnvOptions;
+class SystemClock;
 
 // Imports a set of sst files as is into a new column family. Logic is similar
 // to ExternalSstFileIngestionJob.
 class ImportColumnFamilyJob {
  public:
-  ImportColumnFamilyJob(Env* env, VersionSet* versions, ColumnFamilyData* cfd,
+  ImportColumnFamilyJob(VersionSet* versions, ColumnFamilyData* cfd,
                         const ImmutableDBOptions& db_options,
                         const EnvOptions& env_options,
                         const ImportColumnFamilyOptions& import_options,
                         const std::vector<LiveFileMetaData>& metadata,
                         const std::shared_ptr<IOTracer>& io_tracer)
-      : env_(env),
+      : clock_(db_options.clock),
         versions_(versions),
         cfd_(cfd),
         db_options_(db_options),
@@ -59,7 +60,7 @@ class ImportColumnFamilyJob {
                              IngestedFileInfo* file_to_import,
                              SuperVersion* sv);
 
-  Env* env_;
+  SystemClock* clock_;
   VersionSet* versions_;
   ColumnFamilyData* cfd_;
   const ImmutableDBOptions& db_options_;

db/internal_stats.cc

@@ -50,6 +50,8 @@ const std::map<LevelStatType, LevelStat> InternalStats::compaction_level_stats =
         {LevelStatType::AVG_SEC, LevelStat{"AvgSec", "Avg(sec)"}},
         {LevelStatType::KEY_IN, LevelStat{"KeyIn", "KeyIn"}},
         {LevelStatType::KEY_DROP, LevelStat{"KeyDrop", "KeyDrop"}},
+        {LevelStatType::R_BLOB_GB, LevelStat{"RblobGB", "Rblob(GB)"}},
+        {LevelStatType::W_BLOB_GB, LevelStat{"WblobGB", "Wblob(GB)"}},
 };
 
 namespace {
@@ -67,7 +69,8 @@ void PrintLevelStatsHeader(char* buf, size_t len, const std::string& cf_name,
   };
   int line_size = snprintf(
       buf + written_size, len - written_size,
-      "%s    %s   %s     %s %s  %s %s %s %s %s %s %s %s %s %s %s %s %s %s\n",
+      "%s    %s   %s     %s %s  %s %s %s %s %s %s %s %s %s %s %s %s %s %s %s "
+      "%s\n",
       // Note that we skip COMPACTED_FILES and merge it with Files column
       group_by.c_str(), hdr(LevelStatType::NUM_FILES),
       hdr(LevelStatType::SIZE_BYTES), hdr(LevelStatType::SCORE),
@@ -78,7 +81,8 @@ void PrintLevelStatsHeader(char* buf, size_t len, const std::string& cf_name,
       hdr(LevelStatType::WRITE_MBPS), hdr(LevelStatType::COMP_SEC),
       hdr(LevelStatType::COMP_CPU_SEC), hdr(LevelStatType::COMP_COUNT),
       hdr(LevelStatType::AVG_SEC), hdr(LevelStatType::KEY_IN),
-      hdr(LevelStatType::KEY_DROP));
+      hdr(LevelStatType::KEY_DROP), hdr(LevelStatType::R_BLOB_GB),
+      hdr(LevelStatType::W_BLOB_GB));
 
   written_size += line_size;
   written_size = std::min(written_size, static_cast<int>(len));
@@ -90,10 +94,12 @@ void PrepareLevelStats(std::map<LevelStatType, double>* level_stats,
                        int num_files, int being_compacted,
                        double total_file_size, double score, double w_amp,
                        const InternalStats::CompactionStats& stats) {
-  uint64_t bytes_read =
-      stats.bytes_read_non_output_levels + stats.bytes_read_output_level;
-  int64_t bytes_new = stats.bytes_written - stats.bytes_read_output_level;
-  double elapsed = (stats.micros + 1) / kMicrosInSec;
+  const uint64_t bytes_read = stats.bytes_read_non_output_levels +
+                              stats.bytes_read_output_level +
+                              stats.bytes_read_blob;
+  const uint64_t bytes_written = stats.bytes_written + stats.bytes_written_blob;
+  const int64_t bytes_new = stats.bytes_written - stats.bytes_read_output_level;
+  const double elapsed = (stats.micros + 1) / kMicrosInSec;
 
   (*level_stats)[LevelStatType::NUM_FILES] = num_files;
   (*level_stats)[LevelStatType::COMPACTED_FILES] = being_compacted;
@@ -108,8 +114,7 @@ void PrepareLevelStats(std::map<LevelStatType, double>* level_stats,
   (*level_stats)[LevelStatType::MOVED_GB] = stats.bytes_moved / kGB;
   (*level_stats)[LevelStatType::WRITE_AMP] = w_amp;
   (*level_stats)[LevelStatType::READ_MBPS] = bytes_read / kMB / elapsed;
-  (*level_stats)[LevelStatType::WRITE_MBPS] =
-      stats.bytes_written / kMB / elapsed;
+  (*level_stats)[LevelStatType::WRITE_MBPS] = bytes_written / kMB / elapsed;
   (*level_stats)[LevelStatType::COMP_SEC] = stats.micros / kMicrosInSec;
   (*level_stats)[LevelStatType::COMP_CPU_SEC] = stats.cpu_micros / kMicrosInSec;
   (*level_stats)[LevelStatType::COMP_COUNT] = stats.count;
@@ -119,6 +124,8 @@ void PrepareLevelStats(std::map<LevelStatType, double>* level_stats,
       static_cast<double>(stats.num_input_records);
   (*level_stats)[LevelStatType::KEY_DROP] =
       static_cast<double>(stats.num_dropped_records);
+  (*level_stats)[LevelStatType::R_BLOB_GB] = stats.bytes_read_blob / kGB;
+  (*level_stats)[LevelStatType::W_BLOB_GB] = stats.bytes_written_blob / kGB;
 }
 
 void PrintLevelStats(char* buf, size_t len, const std::string& name,
@@ -143,7 +150,9 @@ void PrintLevelStats(char* buf, size_t len, const std::string& name,
       "%9d "      /*  Comp(cnt) */
       "%8.3f "    /*  Avg(sec) */
       "%7s "      /*  KeyIn */
-      "%6s\n",    /*  KeyDrop */
+      "%6s "      /*  KeyDrop */
+      "%9.1f "    /*  Rblob(GB) */
+      "%9.1f\n",  /*  Wblob(GB) */
       name.c_str(), static_cast<int>(stat_value.at(LevelStatType::NUM_FILES)),
       static_cast<int>(stat_value.at(LevelStatType::COMPACTED_FILES)),
       BytesToHumanString(
@@ -168,7 +177,9 @@ void PrintLevelStats(char* buf, size_t len, const std::string& name,
           .c_str(),
       NumberToHumanString(
           static_cast<std::int64_t>(stat_value.at(LevelStatType::KEY_DROP)))
-          .c_str());
+          .c_str(),
+      stat_value.at(LevelStatType::R_BLOB_GB),
+      stat_value.at(LevelStatType::W_BLOB_GB));
 }
 
 void PrintLevelStats(char* buf, size_t len, const std::string& name,
@@ -990,7 +1001,7 @@ bool InternalStats::HandleBlockCachePinnedUsage(uint64_t* value, DBImpl* /*db*/,
 void InternalStats::DumpDBStats(std::string* value) {
   char buf[1000];
   // DB-level stats, only available from default column family
-  double seconds_up = (env_->NowMicros() - started_at_ + 1) / kMicrosInSec;
+  double seconds_up = (clock_->NowMicros() - started_at_ + 1) / kMicrosInSec;
   double interval_seconds_up = seconds_up - db_stats_snapshot_.seconds_up;
   snprintf(buf, sizeof(buf),
            "\n** DB Stats **\nUptime(secs): %.1f total, %.1f interval\n",
@@ -1112,9 +1123,10 @@ void InternalStats::DumpDBStats(std::string* value) {
  */
 void InternalStats::DumpCFMapStats(
     std::map<std::string, std::string>* cf_stats) {
+  const VersionStorageInfo* vstorage = cfd_->current()->storage_info();
   CompactionStats compaction_stats_sum;
   std::map<int, std::map<LevelStatType, double>> levels_stats;
-  DumpCFMapStats(&levels_stats, &compaction_stats_sum);
+  DumpCFMapStats(vstorage, &levels_stats, &compaction_stats_sum);
   for (auto const& level_ent : levels_stats) {
     auto level_str =
         level_ent.first == -1 ? "Sum" : "L" + ToString(level_ent.first);
@@ -1131,9 +1143,10 @@ void InternalStats::DumpCFMapStats(
 }
 
 void InternalStats::DumpCFMapStats(
+    const VersionStorageInfo* vstorage,
     std::map<int, std::map<LevelStatType, double>>* levels_stats,
     CompactionStats* compaction_stats_sum) {
-  const VersionStorageInfo* vstorage = cfd_->current()->storage_info();
+  assert(vstorage);
 
   int num_levels_to_check =
       (cfd_->ioptions()->compaction_style != kCompactionStyleFIFO)
@@ -1174,12 +1187,14 @@ void InternalStats::DumpCFMapStats(
       if (level == 0) {
         input_bytes = curr_ingest;
       } else {
-        input_bytes = comp_stats_[level].bytes_read_non_output_levels;
+        input_bytes = comp_stats_[level].bytes_read_non_output_levels +
+                      comp_stats_[level].bytes_read_blob;
       }
       double w_amp =
           (input_bytes == 0)
               ? 0.0
-              : static_cast<double>(comp_stats_[level].bytes_written) /
+              : static_cast<double>(comp_stats_[level].bytes_written +
+                                    comp_stats_[level].bytes_written_blob) /
                     input_bytes;
       std::map<LevelStatType, double> level_stats;
       PrepareLevelStats(&level_stats, files, files_being_compacted[level],
@@ -1189,7 +1204,8 @@ void InternalStats::DumpCFMapStats(
     }
   }
   // Cumulative summary
-  double w_amp = compaction_stats_sum->bytes_written /
+  double w_amp = (compaction_stats_sum->bytes_written +
+                  compaction_stats_sum->bytes_written_blob) /
                  static_cast<double>(curr_ingest + 1);
   // Stats summary across levels
   std::map<LevelStatType, double> sum_stats;
@@ -1256,9 +1272,10 @@ void InternalStats::DumpCFStatsNoFileHistogram(std::string* value) {
   value->append(buf);
 
   // Print stats for each level
+  const VersionStorageInfo* vstorage = cfd_->current()->storage_info();
   std::map<int, std::map<LevelStatType, double>> levels_stats;
   CompactionStats compaction_stats_sum;
-  DumpCFMapStats(&levels_stats, &compaction_stats_sum);
+  DumpCFMapStats(vstorage, &levels_stats, &compaction_stats_sum);
   for (int l = 0; l < number_levels_; ++l) {
     if (levels_stats.find(l) != levels_stats.end()) {
       PrintLevelStats(buf, sizeof(buf), "L" + ToString(l), levels_stats[l]);
@@ -1294,7 +1311,8 @@ void InternalStats::DumpCFStatsNoFileHistogram(std::string* value) {
   CompactionStats interval_stats(compaction_stats_sum);
   interval_stats.Subtract(cf_stats_snapshot_.comp_stats);
   double w_amp =
-      interval_stats.bytes_written / static_cast<double>(interval_ingest);
+      (interval_stats.bytes_written + interval_stats.bytes_written_blob) /
+      static_cast<double>(interval_ingest);
   PrintLevelStats(buf, sizeof(buf), "Int", 0, 0, 0, 0, w_amp, interval_stats);
   value->append(buf);
 
@@ -1313,7 +1331,13 @@ void InternalStats::DumpCFStatsNoFileHistogram(std::string* value) {
     }
   }
 
-  double seconds_up = (env_->NowMicros() - started_at_ + 1) / kMicrosInSec;
+  snprintf(buf, sizeof(buf),
+           "\nBlob file count: %" ROCKSDB_PRIszt ", total size: %.1f GB\n\n",
+           vstorage->GetBlobFiles().size(),
+           vstorage->GetTotalBlobFileSize() / kGB);
+  value->append(buf);
+
+  double seconds_up = (clock_->NowMicros() - started_at_ + 1) / kMicrosInSec;
   double interval_seconds_up = seconds_up - cf_stats_snapshot_.seconds_up;
   snprintf(buf, sizeof(buf), "Uptime(secs): %.1f total, %.1f interval\n",
            seconds_up, interval_seconds_up);
@@ -1353,8 +1377,10 @@ void InternalStats::DumpCFStatsNoFileHistogram(std::string* value) {
   uint64_t compact_micros = 0;
   for (int level = 0; level < number_levels_; level++) {
     compact_bytes_read += comp_stats_[level].bytes_read_output_level +
-                          comp_stats_[level].bytes_read_non_output_levels;
-    compact_bytes_write += comp_stats_[level].bytes_written;
+                          comp_stats_[level].bytes_read_non_output_levels +
+                          comp_stats_[level].bytes_read_blob;
+    compact_bytes_write += comp_stats_[level].bytes_written +
+                           comp_stats_[level].bytes_written_blob;
     compact_micros += comp_stats_[level].micros;
   }
 

db/internal_stats.h

@@ -14,6 +14,7 @@
 #include <vector>
 
 #include "db/version_set.h"
+#include "rocksdb/system_clock.h"
 
 class ColumnFamilyData;
 
@@ -78,6 +79,8 @@ enum class LevelStatType {
   AVG_SEC,
   KEY_IN,
   KEY_DROP,
+  R_BLOB_GB,
+  W_BLOB_GB,
   TOTAL  // total number of types
 };
 
@@ -122,7 +125,7 @@ class InternalStats {
     kIntStatsNumMax,
   };
 
-  InternalStats(int num_levels, Env* env, ColumnFamilyData* cfd)
+  InternalStats(int num_levels, SystemClock* clock, ColumnFamilyData* cfd)
       : db_stats_{},
         cf_stats_value_{},
         cf_stats_count_{},
@@ -131,9 +134,9 @@ class InternalStats {
         file_read_latency_(num_levels),
         bg_error_count_(0),
         number_levels_(num_levels),
-        env_(env),
+        clock_(clock),
         cfd_(cfd),
-        started_at_(env->NowMicros()) {}
+        started_at_(clock->NowMicros()) {}
 
   // Per level compaction stats.  comp_stats_[level] stores the stats for
   // compactions that produced data for the specified "level".
@@ -141,32 +144,42 @@ class InternalStats {
     uint64_t micros;
     uint64_t cpu_micros;
 
-    // The number of bytes read from all non-output levels
+    // The number of bytes read from all non-output levels (table files)
     uint64_t bytes_read_non_output_levels;
 
-    // The number of bytes read from the compaction output level.
+    // The number of bytes read from the compaction output level (table files)
     uint64_t bytes_read_output_level;
 
-    // Total number of bytes written during compaction
+    // The number of bytes read from blob files
+    uint64_t bytes_read_blob;
+
+    // Total number of bytes written to table files during compaction
     uint64_t bytes_written;
 
-    // Total number of bytes moved to the output level
+    // Total number of bytes written to blob files during compaction
+    uint64_t bytes_written_blob;
+
+    // Total number of bytes moved to the output level (table files)
     uint64_t bytes_moved;
 
-    // The number of compaction input files in all non-output levels.
+    // The number of compaction input files in all non-output levels (table
+    // files)
     int num_input_files_in_non_output_levels;
 
-    // The number of compaction input files in the output level.
+    // The number of compaction input files in the output level (table files)
     int num_input_files_in_output_level;
 
-    // The number of compaction output files.
+    // The number of compaction output files (table files)
     int num_output_files;
 
+    // The number of compaction output files (blob files)
+    int num_output_files_blob;
+
     // Total incoming entries during compaction between levels N and N+1
     uint64_t num_input_records;
 
     // Accumulated diff number of entries
-    // (num input entries - num output entires) for compaction  levels N and N+1
+    // (num input entries - num output entries) for compaction levels N and N+1
     uint64_t num_dropped_records;
 
     // Number of compactions done
@@ -180,11 +193,14 @@ class InternalStats {
           cpu_micros(0),
           bytes_read_non_output_levels(0),
           bytes_read_output_level(0),
+          bytes_read_blob(0),
           bytes_written(0),
+          bytes_written_blob(0),
           bytes_moved(0),
           num_input_files_in_non_output_levels(0),
           num_input_files_in_output_level(0),
           num_output_files(0),
+          num_output_files_blob(0),
           num_input_records(0),
           num_dropped_records(0),
           count(0) {
@@ -199,11 +215,14 @@ class InternalStats {
           cpu_micros(0),
           bytes_read_non_output_levels(0),
           bytes_read_output_level(0),
+          bytes_read_blob(0),
           bytes_written(0),
+          bytes_written_blob(0),
           bytes_moved(0),
           num_input_files_in_non_output_levels(0),
           num_input_files_in_output_level(0),
           num_output_files(0),
+          num_output_files_blob(0),
           num_input_records(0),
           num_dropped_records(0),
           count(c) {
@@ -224,12 +243,15 @@ class InternalStats {
           cpu_micros(c.cpu_micros),
           bytes_read_non_output_levels(c.bytes_read_non_output_levels),
           bytes_read_output_level(c.bytes_read_output_level),
+          bytes_read_blob(c.bytes_read_blob),
           bytes_written(c.bytes_written),
+          bytes_written_blob(c.bytes_written_blob),
           bytes_moved(c.bytes_moved),
           num_input_files_in_non_output_levels(
               c.num_input_files_in_non_output_levels),
           num_input_files_in_output_level(c.num_input_files_in_output_level),
           num_output_files(c.num_output_files),
+          num_output_files_blob(c.num_output_files_blob),
           num_input_records(c.num_input_records),
           num_dropped_records(c.num_dropped_records),
           count(c.count) {
@@ -244,12 +266,15 @@ class InternalStats {
       cpu_micros = c.cpu_micros;
       bytes_read_non_output_levels = c.bytes_read_non_output_levels;
       bytes_read_output_level = c.bytes_read_output_level;
+      bytes_read_blob = c.bytes_read_blob;
       bytes_written = c.bytes_written;
+      bytes_written_blob = c.bytes_written_blob;
       bytes_moved = c.bytes_moved;
       num_input_files_in_non_output_levels =
           c.num_input_files_in_non_output_levels;
       num_input_files_in_output_level = c.num_input_files_in_output_level;
       num_output_files = c.num_output_files;
+      num_output_files_blob = c.num_output_files_blob;
       num_input_records = c.num_input_records;
       num_dropped_records = c.num_dropped_records;
       count = c.count;
@@ -266,11 +291,14 @@ class InternalStats {
       this->cpu_micros = 0;
       this->bytes_read_non_output_levels = 0;
       this->bytes_read_output_level = 0;
+      this->bytes_read_blob = 0;
       this->bytes_written = 0;
+      this->bytes_written_blob = 0;
       this->bytes_moved = 0;
       this->num_input_files_in_non_output_levels = 0;
       this->num_input_files_in_output_level = 0;
       this->num_output_files = 0;
+      this->num_output_files_blob = 0;
       this->num_input_records = 0;
       this->num_dropped_records = 0;
       this->count = 0;
@@ -285,13 +313,16 @@ class InternalStats {
       this->cpu_micros += c.cpu_micros;
       this->bytes_read_non_output_levels += c.bytes_read_non_output_levels;
       this->bytes_read_output_level += c.bytes_read_output_level;
+      this->bytes_read_blob += c.bytes_read_blob;
       this->bytes_written += c.bytes_written;
+      this->bytes_written_blob += c.bytes_written_blob;
       this->bytes_moved += c.bytes_moved;
       this->num_input_files_in_non_output_levels +=
           c.num_input_files_in_non_output_levels;
       this->num_input_files_in_output_level +=
           c.num_input_files_in_output_level;
       this->num_output_files += c.num_output_files;
+      this->num_output_files_blob += c.num_output_files_blob;
       this->num_input_records += c.num_input_records;
       this->num_dropped_records += c.num_dropped_records;
       this->count += c.count;
@@ -306,13 +337,16 @@ class InternalStats {
       this->cpu_micros -= c.cpu_micros;
       this->bytes_read_non_output_levels -= c.bytes_read_non_output_levels;
       this->bytes_read_output_level -= c.bytes_read_output_level;
+      this->bytes_read_blob -= c.bytes_read_blob;
       this->bytes_written -= c.bytes_written;
+      this->bytes_written_blob -= c.bytes_written_blob;
       this->bytes_moved -= c.bytes_moved;
       this->num_input_files_in_non_output_levels -=
           c.num_input_files_in_non_output_levels;
       this->num_input_files_in_output_level -=
           c.num_input_files_in_output_level;
       this->num_output_files -= c.num_output_files;
+      this->num_output_files_blob -= c.num_output_files_blob;
       this->num_input_records -= c.num_input_records;
       this->num_dropped_records -= c.num_dropped_records;
       this->count -= c.count;
@@ -341,7 +375,7 @@ class InternalStats {
     cf_stats_snapshot_.Clear();
     db_stats_snapshot_.Clear();
     bg_error_count_ = 0;
-    started_at_ = env_->NowMicros();
+    started_at_ = clock_->NowMicros();
   }
 
   void AddCompactionStats(int level, Env::Priority thread_pri,
@@ -411,6 +445,7 @@ class InternalStats {
   void DumpDBStats(std::string* value);
   void DumpCFMapStats(std::map<std::string, std::string>* cf_stats);
   void DumpCFMapStats(
+      const VersionStorageInfo* vstorage,
       std::map<int, std::map<LevelStatType, double>>* level_stats,
       CompactionStats* compaction_stats_sum);
   void DumpCFMapStatsByPriority(
@@ -602,7 +637,7 @@ class InternalStats {
   uint64_t bg_error_count_;
 
   const int number_levels_;
-  Env* env_;
+  SystemClock* clock_;
   ColumnFamilyData* cfd_;
   uint64_t started_at_;
 };
@@ -641,18 +676,22 @@ class InternalStats {
     kIntStatsNumMax,
   };
 
-  InternalStats(int /*num_levels*/, Env* /*env*/, ColumnFamilyData* /*cfd*/) {}
+  InternalStats(int /*num_levels*/, SystemClock* /*clock*/,
+                ColumnFamilyData* /*cfd*/) {}
 
   struct CompactionStats {
     uint64_t micros;
     uint64_t cpu_micros;
     uint64_t bytes_read_non_output_levels;
     uint64_t bytes_read_output_level;
+    uint64_t bytes_read_blob;
     uint64_t bytes_written;
+    uint64_t bytes_written_blob;
     uint64_t bytes_moved;
     int num_input_files_in_non_output_levels;
     int num_input_files_in_output_level;
     int num_output_files;
+    int num_output_files_blob;
     uint64_t num_input_records;
     uint64_t num_dropped_records;
     int count;

db/kv_checksum.h

@@ -0,0 +1,424 @@
+//  Copyright (c) 2020-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).
+//
+// This file contains classes containing fields to protect individual entries.
+// The classes are named "ProtectionInfo<suffix>", where <suffix> indicates the
+// combination of fields that are covered. Each field has a single letter
+// abbreviation as follows.
+//
+// K = key
+// V = value
+// O = optype aka value type
+// T = timestamp
+// S = seqno
+// C = CF ID
+//
+// Then, for example, a class that protects an entry consisting of key, value,
+// optype, timestamp, and CF ID (i.e., a `WriteBatch` entry) would be named
+// `ProtectionInfoKVOTC`.
+//
+// The `ProtectionInfo.*` classes are templated on the integer type used to hold
+// the XOR of hashes for each field. Only unsigned integer types are supported,
+// and the maximum supported integer width is 64 bits. When the integer type is
+// narrower than the hash values, we lop off the most significant bits to make
+// them fit.
+//
+// The `ProtectionInfo.*` classes are all intended to be non-persistent. We do
+// not currently make the byte order consistent for integer fields before
+// hashing them, so the resulting values are endianness-dependent.
+
+#pragma once
+
+#include <type_traits>
+
+#include "db/dbformat.h"
+#include "rocksdb/types.h"
+#include "util/hash.h"
+
+namespace ROCKSDB_NAMESPACE {
+
+template <typename T>
+class ProtectionInfo;
+template <typename T>
+class ProtectionInfoKVOT;
+template <typename T>
+class ProtectionInfoKVOTC;
+template <typename T>
+class ProtectionInfoKVOTS;
+
+// Aliases for 64-bit protection infos.
+typedef ProtectionInfo<uint64_t> ProtectionInfo64;
+typedef ProtectionInfoKVOT<uint64_t> ProtectionInfoKVOT64;
+typedef ProtectionInfoKVOTC<uint64_t> ProtectionInfoKVOTC64;
+typedef ProtectionInfoKVOTS<uint64_t> ProtectionInfoKVOTS64;
+
+template <typename T>
+class ProtectionInfo {
+ public:
+  ProtectionInfo<T>() = default;
+
+  Status GetStatus() const;
+  ProtectionInfoKVOT<T> ProtectKVOT(const Slice& key, const Slice& value,
+                                    ValueType op_type,
+                                    const Slice& timestamp) const;
+  ProtectionInfoKVOT<T> ProtectKVOT(const SliceParts& key,
+                                    const SliceParts& value, ValueType op_type,
+                                    const Slice& timestamp) const;
+
+ private:
+  friend class ProtectionInfoKVOT<T>;
+  friend class ProtectionInfoKVOTS<T>;
+  friend class ProtectionInfoKVOTC<T>;
+
+  // Each field is hashed with an independent value so we can catch fields being
+  // swapped. Per the `NPHash64()` docs, using consecutive seeds is a pitfall,
+  // and we should instead vary our seeds by a large odd number. This value by
+  // which we increment (0xD28AAD72F49BD50B) was taken from
+  // `head -c8 /dev/urandom | hexdump`, run repeatedly until it yielded an odd
+  // number. The values are computed manually since the Windows C++ compiler
+  // complains about the overflow when adding constants.
+  static const uint64_t kSeedK = 0;
+  static const uint64_t kSeedV = 0xD28AAD72F49BD50B;
+  static const uint64_t kSeedO = 0xA5155AE5E937AA16;
+  static const uint64_t kSeedT = 0x77A00858DDD37F21;
+  static const uint64_t kSeedS = 0x4A2AB5CBD26F542C;
+  static const uint64_t kSeedC = 0x1CB5633EC70B2937;
+
+  ProtectionInfo<T>(T val) : val_(val) {
+    static_assert(sizeof(ProtectionInfo<T>) == sizeof(T), "");
+  }
+
+  T GetVal() const { return val_; }
+  void SetVal(T val) { val_ = val; }
+
+  T val_ = 0;
+};
+
+template <typename T>
+class ProtectionInfoKVOT {
+ public:
+  ProtectionInfoKVOT<T>() = default;
+
+  ProtectionInfo<T> StripKVOT(const Slice& key, const Slice& value,
+                              ValueType op_type, const Slice& timestamp) const;
+  ProtectionInfo<T> StripKVOT(const SliceParts& key, const SliceParts& value,
+                              ValueType op_type, const Slice& timestamp) const;
+
+  ProtectionInfoKVOTC<T> ProtectC(ColumnFamilyId column_family_id) const;
+  ProtectionInfoKVOTS<T> ProtectS(SequenceNumber sequence_number) const;
+
+  void UpdateK(const Slice& old_key, const Slice& new_key);
+  void UpdateK(const SliceParts& old_key, const SliceParts& new_key);
+  void UpdateV(const Slice& old_value, const Slice& new_value);
+  void UpdateV(const SliceParts& old_value, const SliceParts& new_value);
+  void UpdateO(ValueType old_op_type, ValueType new_op_type);
+  void UpdateT(const Slice& old_timestamp, const Slice& new_timestamp);
+
+ private:
+  friend class ProtectionInfo<T>;
+  friend class ProtectionInfoKVOTS<T>;
+  friend class ProtectionInfoKVOTC<T>;
+
+  ProtectionInfoKVOT<T>(T val) : info_(val) {
+    static_assert(sizeof(ProtectionInfoKVOT<T>) == sizeof(T), "");
+  }
+
+  T GetVal() const { return info_.GetVal(); }
+  void SetVal(T val) { info_.SetVal(val); }
+
+  ProtectionInfo<T> info_;
+};
+
+template <typename T>
+class ProtectionInfoKVOTC {
+ public:
+  ProtectionInfoKVOTC<T>() = default;
+
+  ProtectionInfoKVOT<T> StripC(ColumnFamilyId column_family_id) const;
+
+  void UpdateK(const Slice& old_key, const Slice& new_key) {
+    kvot_.UpdateK(old_key, new_key);
+  }
+  void UpdateK(const SliceParts& old_key, const SliceParts& new_key) {
+    kvot_.UpdateK(old_key, new_key);
+  }
+  void UpdateV(const Slice& old_value, const Slice& new_value) {
+    kvot_.UpdateV(old_value, new_value);
+  }
+  void UpdateV(const SliceParts& old_value, const SliceParts& new_value) {
+    kvot_.UpdateV(old_value, new_value);
+  }
+  void UpdateO(ValueType old_op_type, ValueType new_op_type) {
+    kvot_.UpdateO(old_op_type, new_op_type);
+  }
+  void UpdateT(const Slice& old_timestamp, const Slice& new_timestamp) {
+    kvot_.UpdateT(old_timestamp, new_timestamp);
+  }
+  void UpdateC(ColumnFamilyId old_column_family_id,
+               ColumnFamilyId new_column_family_id);
+
+ private:
+  friend class ProtectionInfoKVOT<T>;
+
+  ProtectionInfoKVOTC<T>(T val) : kvot_(val) {
+    static_assert(sizeof(ProtectionInfoKVOTC<T>) == sizeof(T), "");
+  }
+
+  T GetVal() const { return kvot_.GetVal(); }
+  void SetVal(T val) { kvot_.SetVal(val); }
+
+  ProtectionInfoKVOT<T> kvot_;
+};
+
+template <typename T>
+class ProtectionInfoKVOTS {
+ public:
+  ProtectionInfoKVOTS<T>() = default;
+
+  ProtectionInfoKVOT<T> StripS(SequenceNumber sequence_number) const;
+
+  void UpdateK(const Slice& old_key, const Slice& new_key) {
+    kvot_.UpdateK(old_key, new_key);
+  }
+  void UpdateK(const SliceParts& old_key, const SliceParts& new_key) {
+    kvot_.UpdateK(old_key, new_key);
+  }
+  void UpdateV(const Slice& old_value, const Slice& new_value) {
+    kvot_.UpdateV(old_value, new_value);
+  }
+  void UpdateV(const SliceParts& old_value, const SliceParts& new_value) {
+    kvot_.UpdateV(old_value, new_value);
+  }
+  void UpdateO(ValueType old_op_type, ValueType new_op_type) {
+    kvot_.UpdateO(old_op_type, new_op_type);
+  }
+  void UpdateT(const Slice& old_timestamp, const Slice& new_timestamp) {
+    kvot_.UpdateT(old_timestamp, new_timestamp);
+  }
+  void UpdateS(SequenceNumber old_sequence_number,
+               SequenceNumber new_sequence_number);
+
+ private:
+  friend class ProtectionInfoKVOT<T>;
+
+  ProtectionInfoKVOTS<T>(T val) : kvot_(val) {
+    static_assert(sizeof(ProtectionInfoKVOTS<T>) == sizeof(T), "");
+  }
+
+  T GetVal() const { return kvot_.GetVal(); }
+  void SetVal(T val) { kvot_.SetVal(val); }
+
+  ProtectionInfoKVOT<T> kvot_;
+};
+
+template <typename T>
+Status ProtectionInfo<T>::GetStatus() const {
+  if (val_ != 0) {
+    return Status::Corruption("ProtectionInfo mismatch");
+  }
+  return Status::OK();
+}
+
+template <typename T>
+ProtectionInfoKVOT<T> ProtectionInfo<T>::ProtectKVOT(
+    const Slice& key, const Slice& value, ValueType op_type,
+    const Slice& timestamp) const {
+  T val = GetVal();
+  val = val ^ static_cast<T>(GetSliceNPHash64(key, ProtectionInfo<T>::kSeedK));
+  val =
+      val ^ static_cast<T>(GetSliceNPHash64(value, ProtectionInfo<T>::kSeedV));
+  val = val ^
+        static_cast<T>(NPHash64(reinterpret_cast<char*>(&op_type),
+                                sizeof(op_type), ProtectionInfo<T>::kSeedO));
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(timestamp, ProtectionInfo<T>::kSeedT));
+  return ProtectionInfoKVOT<T>(val);
+}
+
+template <typename T>
+ProtectionInfoKVOT<T> ProtectionInfo<T>::ProtectKVOT(
+    const SliceParts& key, const SliceParts& value, ValueType op_type,
+    const Slice& timestamp) const {
+  T val = GetVal();
+  val = val ^
+        static_cast<T>(GetSlicePartsNPHash64(key, ProtectionInfo<T>::kSeedK));
+  val = val ^
+        static_cast<T>(GetSlicePartsNPHash64(value, ProtectionInfo<T>::kSeedV));
+  val = val ^
+        static_cast<T>(NPHash64(reinterpret_cast<char*>(&op_type),
+                                sizeof(op_type), ProtectionInfo<T>::kSeedO));
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(timestamp, ProtectionInfo<T>::kSeedT));
+  return ProtectionInfoKVOT<T>(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOT<T>::UpdateK(const Slice& old_key,
+                                    const Slice& new_key) {
+  T val = GetVal();
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(old_key, ProtectionInfo<T>::kSeedK));
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(new_key, ProtectionInfo<T>::kSeedK));
+  SetVal(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOT<T>::UpdateK(const SliceParts& old_key,
+                                    const SliceParts& new_key) {
+  T val = GetVal();
+  val = val ^ static_cast<T>(
+                  GetSlicePartsNPHash64(old_key, ProtectionInfo<T>::kSeedK));
+  val = val ^ static_cast<T>(
+                  GetSlicePartsNPHash64(new_key, ProtectionInfo<T>::kSeedK));
+  SetVal(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOT<T>::UpdateV(const Slice& old_value,
+                                    const Slice& new_value) {
+  T val = GetVal();
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(old_value, ProtectionInfo<T>::kSeedV));
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(new_value, ProtectionInfo<T>::kSeedV));
+  SetVal(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOT<T>::UpdateV(const SliceParts& old_value,
+                                    const SliceParts& new_value) {
+  T val = GetVal();
+  val = val ^ static_cast<T>(
+                  GetSlicePartsNPHash64(old_value, ProtectionInfo<T>::kSeedV));
+  val = val ^ static_cast<T>(
+                  GetSlicePartsNPHash64(new_value, ProtectionInfo<T>::kSeedV));
+  SetVal(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOT<T>::UpdateO(ValueType old_op_type,
+                                    ValueType new_op_type) {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(reinterpret_cast<char*>(&old_op_type),
+                                      sizeof(old_op_type),
+                                      ProtectionInfo<T>::kSeedO));
+  val = val ^ static_cast<T>(NPHash64(reinterpret_cast<char*>(&new_op_type),
+                                      sizeof(new_op_type),
+                                      ProtectionInfo<T>::kSeedO));
+  SetVal(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOT<T>::UpdateT(const Slice& old_timestamp,
+                                    const Slice& new_timestamp) {
+  T val = GetVal();
+  val = val ^ static_cast<T>(
+                  GetSliceNPHash64(old_timestamp, ProtectionInfo<T>::kSeedT));
+  val = val ^ static_cast<T>(
+                  GetSliceNPHash64(new_timestamp, ProtectionInfo<T>::kSeedT));
+  SetVal(val);
+}
+
+template <typename T>
+ProtectionInfo<T> ProtectionInfoKVOT<T>::StripKVOT(
+    const Slice& key, const Slice& value, ValueType op_type,
+    const Slice& timestamp) const {
+  T val = GetVal();
+  val = val ^ static_cast<T>(GetSliceNPHash64(key, ProtectionInfo<T>::kSeedK));
+  val =
+      val ^ static_cast<T>(GetSliceNPHash64(value, ProtectionInfo<T>::kSeedV));
+  val = val ^
+        static_cast<T>(NPHash64(reinterpret_cast<char*>(&op_type),
+                                sizeof(op_type), ProtectionInfo<T>::kSeedO));
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(timestamp, ProtectionInfo<T>::kSeedT));
+  return ProtectionInfo<T>(val);
+}
+
+template <typename T>
+ProtectionInfo<T> ProtectionInfoKVOT<T>::StripKVOT(
+    const SliceParts& key, const SliceParts& value, ValueType op_type,
+    const Slice& timestamp) const {
+  T val = GetVal();
+  val = val ^
+        static_cast<T>(GetSlicePartsNPHash64(key, ProtectionInfo<T>::kSeedK));
+  val = val ^
+        static_cast<T>(GetSlicePartsNPHash64(value, ProtectionInfo<T>::kSeedV));
+  val = val ^
+        static_cast<T>(NPHash64(reinterpret_cast<char*>(&op_type),
+                                sizeof(op_type), ProtectionInfo<T>::kSeedO));
+  val = val ^
+        static_cast<T>(GetSliceNPHash64(timestamp, ProtectionInfo<T>::kSeedT));
+  return ProtectionInfo<T>(val);
+}
+
+template <typename T>
+ProtectionInfoKVOTC<T> ProtectionInfoKVOT<T>::ProtectC(
+    ColumnFamilyId column_family_id) const {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(
+                  reinterpret_cast<char*>(&column_family_id),
+                  sizeof(column_family_id), ProtectionInfo<T>::kSeedC));
+  return ProtectionInfoKVOTC<T>(val);
+}
+
+template <typename T>
+ProtectionInfoKVOT<T> ProtectionInfoKVOTC<T>::StripC(
+    ColumnFamilyId column_family_id) const {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(
+                  reinterpret_cast<char*>(&column_family_id),
+                  sizeof(column_family_id), ProtectionInfo<T>::kSeedC));
+  return ProtectionInfoKVOT<T>(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOTC<T>::UpdateC(ColumnFamilyId old_column_family_id,
+                                     ColumnFamilyId new_column_family_id) {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(
+                  reinterpret_cast<char*>(&old_column_family_id),
+                  sizeof(old_column_family_id), ProtectionInfo<T>::kSeedC));
+  val = val ^ static_cast<T>(NPHash64(
+                  reinterpret_cast<char*>(&new_column_family_id),
+                  sizeof(new_column_family_id), ProtectionInfo<T>::kSeedC));
+  SetVal(val);
+}
+
+template <typename T>
+ProtectionInfoKVOTS<T> ProtectionInfoKVOT<T>::ProtectS(
+    SequenceNumber sequence_number) const {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(reinterpret_cast<char*>(&sequence_number),
+                                      sizeof(sequence_number),
+                                      ProtectionInfo<T>::kSeedS));
+  return ProtectionInfoKVOTS<T>(val);
+}
+
+template <typename T>
+ProtectionInfoKVOT<T> ProtectionInfoKVOTS<T>::StripS(
+    SequenceNumber sequence_number) const {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(reinterpret_cast<char*>(&sequence_number),
+                                      sizeof(sequence_number),
+                                      ProtectionInfo<T>::kSeedS));
+  return ProtectionInfoKVOT<T>(val);
+}
+
+template <typename T>
+void ProtectionInfoKVOTS<T>::UpdateS(SequenceNumber old_sequence_number,
+                                     SequenceNumber new_sequence_number) {
+  T val = GetVal();
+  val = val ^ static_cast<T>(NPHash64(
+                  reinterpret_cast<char*>(&old_sequence_number),
+                  sizeof(old_sequence_number), ProtectionInfo<T>::kSeedS));
+  val = val ^ static_cast<T>(NPHash64(
+                  reinterpret_cast<char*>(&new_sequence_number),
+                  sizeof(new_sequence_number), ProtectionInfo<T>::kSeedS));
+  SetVal(val);
+}
+
+}  // namespace ROCKSDB_NAMESPACE

db/log_writer.h

@@ -8,8 +8,7 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 #pragma once
 
-#include <stdint.h>
-
+#include <cstdint>
 #include <memory>
 
 #include "db/log_format.h"

db/logs_with_prep_tracker.h

@@ -5,8 +5,8 @@
 //
 #pragma once
 
-#include <stdint.h>
 #include <cassert>
+#include <cstdint>
 #include <cstdlib>
 #include <mutex>
 #include <unordered_map>

db/memtable.cc

@@ -13,7 +13,9 @@
 #include <array>
 #include <limits>
 #include <memory>
+
 #include "db/dbformat.h"
+#include "db/kv_checksum.h"
 #include "db/merge_context.h"
 #include "db/merge_helper.h"
 #include "db/pinned_iterators_manager.h"
@@ -102,7 +104,7 @@ MemTable::MemTable(const InternalKeyComparator& cmp,
                  : 0),
       prefix_extractor_(mutable_cf_options.prefix_extractor.get()),
       flush_state_(FLUSH_NOT_REQUESTED),
-      env_(ioptions.env),
+      clock_(ioptions.clock),
       insert_with_hint_prefix_extractor_(
           ioptions.memtable_insert_with_hint_prefix_extractor),
       oldest_key_time_(std::numeric_limits<uint64_t>::max()),
@@ -221,7 +223,7 @@ void MemTable::UpdateOldestKeyTime() {
   uint64_t oldest_key_time = oldest_key_time_.load(std::memory_order_relaxed);
   if (oldest_key_time == std::numeric_limits<uint64_t>::max()) {
     int64_t current_time = 0;
-    auto s = env_->GetCurrentTime(&current_time);
+    auto s = clock_->GetCurrentTime(&current_time);
     if (s.ok()) {
       assert(current_time >= 0);
       // If fail, the timestamp is already set.
@@ -484,9 +486,54 @@ MemTable::MemTableStats MemTable::ApproximateStats(const Slice& start_ikey,
   return {entry_count * (data_size / n), entry_count};
 }
 
+Status MemTable::VerifyEncodedEntry(Slice encoded,
+                                    const ProtectionInfoKVOTS64& kv_prot_info) {
+  uint32_t ikey_len = 0;
+  if (!GetVarint32(&encoded, &ikey_len)) {
+    return Status::Corruption("Unable to parse internal key length");
+  }
+  size_t ts_sz = GetInternalKeyComparator().user_comparator()->timestamp_size();
+  if (ikey_len < 8 + ts_sz) {
+    return Status::Corruption("Internal key length too short");
+  }
+  if (ikey_len > encoded.size()) {
+    return Status::Corruption("Internal key length too long");
+  }
+  uint32_t value_len = 0;
+  const size_t key_without_ts_len = ikey_len - ts_sz - 8;
+  Slice key(encoded.data(), key_without_ts_len);
+  encoded.remove_prefix(key_without_ts_len);
+
+  Slice timestamp(encoded.data(), ts_sz);
+  encoded.remove_prefix(ts_sz);
+
+  uint64_t packed = DecodeFixed64(encoded.data());
+  ValueType value_type = kMaxValue;
+  SequenceNumber sequence_number = kMaxSequenceNumber;
+  UnPackSequenceAndType(packed, &sequence_number, &value_type);
+  encoded.remove_prefix(8);
+
+  if (!GetVarint32(&encoded, &value_len)) {
+    return Status::Corruption("Unable to parse value length");
+  }
+  if (value_len < encoded.size()) {
+    return Status::Corruption("Value length too short");
+  }
+  if (value_len > encoded.size()) {
+    return Status::Corruption("Value length too long");
+  }
+  Slice value(encoded.data(), value_len);
+
+  return kv_prot_info.StripS(sequence_number)
+      .StripKVOT(key, value, value_type, timestamp)
+      .GetStatus();
+}
+
 Status MemTable::Add(SequenceNumber s, ValueType type,
                      const Slice& key, /* user key */
-                     const Slice& value, bool allow_concurrent,
+                     const Slice& value,
+                     const ProtectionInfoKVOTS64* kv_prot_info,
+                     bool allow_concurrent,
                      MemTablePostProcessInfo* post_process_info, void** hint) {
   // Format of an entry is concatenation of:
   //  key_size     : varint32 of internal_key.size()
@@ -514,6 +561,15 @@ Status MemTable::Add(SequenceNumber s, ValueType type,
   p = EncodeVarint32(p, val_size);
   memcpy(p, value.data(), val_size);
   assert((unsigned)(p + val_size - buf) == (unsigned)encoded_len);
+  if (kv_prot_info != nullptr) {
+    Slice encoded(buf, encoded_len);
+    TEST_SYNC_POINT_CALLBACK("MemTable::Add:Encoded", &encoded);
+    Status status = VerifyEncodedEntry(encoded, *kv_prot_info);
+    if (!status.ok()) {
+      return status;
+    }
+  }
+
   size_t ts_sz = GetInternalKeyComparator().user_comparator()->timestamp_size();
   Slice key_without_ts = StripTimestampFromUserKey(key, ts_sz);
 
@@ -628,7 +684,8 @@ struct Saver {
   Statistics* statistics;
   bool inplace_update_support;
   bool do_merge;
-  Env* env_;
+  SystemClock* clock;
+
   ReadCallback* callback_;
   bool* is_blob_index;
   bool allow_data_in_errors;
@@ -666,8 +723,8 @@ static bool SaveValue(void* arg, const char* entry) {
   const Comparator* user_comparator =
       s->mem->GetInternalKeyComparator().user_comparator();
   size_t ts_sz = user_comparator->timestamp_size();
-  if (user_comparator->CompareWithoutTimestamp(user_key_slice,
-                                               s->key->user_key()) == 0) {
+  if (user_comparator->EqualWithoutTimestamp(user_key_slice,
+                                             s->key->user_key())) {
     // Correct user key
     const uint64_t tag = DecodeFixed64(key_ptr + key_length - 8);
     ValueType type;
@@ -712,7 +769,7 @@ static bool SaveValue(void* arg, const char* entry) {
               *(s->status) = MergeHelper::TimedFullMerge(
                   merge_operator, s->key->user_key(), &v,
                   merge_context->GetOperands(), s->value, s->logger,
-                  s->statistics, s->env_, nullptr /* result_operand */, true);
+                  s->statistics, s->clock, nullptr /* result_operand */, true);
             }
           } else {
             // Preserve the value with the goal of returning it as part of
@@ -751,7 +808,7 @@ static bool SaveValue(void* arg, const char* entry) {
             *(s->status) = MergeHelper::TimedFullMerge(
                 merge_operator, s->key->user_key(), nullptr,
                 merge_context->GetOperands(), s->value, s->logger,
-                s->statistics, s->env_, nullptr /* result_operand */, true);
+                s->statistics, s->clock, nullptr /* result_operand */, true);
           }
         } else {
           *(s->status) = Status::NotFound();
@@ -779,7 +836,7 @@ static bool SaveValue(void* arg, const char* entry) {
           *(s->status) = MergeHelper::TimedFullMerge(
               merge_operator, s->key->user_key(), nullptr,
               merge_context->GetOperands(), s->value, s->logger, s->statistics,
-              s->env_, nullptr /* result_operand */, true);
+              s->clock, nullptr /* result_operand */, true);
           *(s->found_final_value) = true;
           return false;
         }
@@ -887,7 +944,7 @@ void MemTable::GetFromTable(const LookupKey& key,
   saver.logger = moptions_.info_log;
   saver.inplace_update_support = moptions_.inplace_update_support;
   saver.statistics = moptions_.statistics;
-  saver.env_ = env_;
+  saver.clock = clock_;
   saver.callback_ = callback;
   saver.is_blob_index = is_blob_index;
   saver.do_merge = do_merge;
@@ -978,7 +1035,8 @@ void MemTable::MultiGet(const ReadOptions& read_options, MultiGetRange* range,
 }
 
 Status MemTable::Update(SequenceNumber seq, const Slice& key,
-                        const Slice& value) {
+                        const Slice& value,
+                        const ProtectionInfoKVOTS64* kv_prot_info) {
   LookupKey lkey(key, seq);
   Slice mem_key = lkey.memtable_key();
 
@@ -1022,6 +1080,13 @@ Status MemTable::Update(SequenceNumber seq, const Slice& key,
                  (unsigned)(VarintLength(key_length) + key_length +
                             VarintLength(value.size()) + value.size()));
           RecordTick(moptions_.statistics, NUMBER_KEYS_UPDATED);
+          if (kv_prot_info != nullptr) {
+            ProtectionInfoKVOTS64 updated_kv_prot_info(*kv_prot_info);
+            // `seq` is swallowed and `existing_seq` prevails.
+            updated_kv_prot_info.UpdateS(seq, existing_seq);
+            Slice encoded(entry, p + value.size() - entry);
+            return VerifyEncodedEntry(encoded, updated_kv_prot_info);
+          }
           return Status::OK();
         }
       }
@@ -1029,11 +1094,12 @@ Status MemTable::Update(SequenceNumber seq, const Slice& key,
   }
 
   // The latest value is not `kTypeValue` or key doesn't exist
-  return Add(seq, kTypeValue, key, value);
+  return Add(seq, kTypeValue, key, value, kv_prot_info);
 }
 
 Status MemTable::UpdateCallback(SequenceNumber seq, const Slice& key,
-                                const Slice& delta) {
+                                const Slice& delta,
+                                const ProtectionInfoKVOTS64* kv_prot_info) {
   LookupKey lkey(key, seq);
   Slice memkey = lkey.memtable_key();
 
@@ -1059,8 +1125,8 @@ Status MemTable::UpdateCallback(SequenceNumber seq, const Slice& key,
       // Correct user key
       const uint64_t tag = DecodeFixed64(key_ptr + key_length - 8);
       ValueType type;
-      uint64_t unused;
-      UnPackSequenceAndType(tag, &unused, &type);
+      uint64_t existing_seq;
+      UnPackSequenceAndType(tag, &existing_seq, &type);
       switch (type) {
         case kTypeValue: {
           Slice prev_value = GetLengthPrefixedSlice(key_ptr + key_length);
@@ -1087,9 +1153,27 @@ Status MemTable::UpdateCallback(SequenceNumber seq, const Slice& key,
             }
             RecordTick(moptions_.statistics, NUMBER_KEYS_UPDATED);
             UpdateFlushState();
+            if (kv_prot_info != nullptr) {
+              ProtectionInfoKVOTS64 updated_kv_prot_info(*kv_prot_info);
+              // `seq` is swallowed and `existing_seq` prevails.
+              updated_kv_prot_info.UpdateS(seq, existing_seq);
+              updated_kv_prot_info.UpdateV(delta,
+                                           Slice(prev_buffer, new_prev_size));
+              Slice encoded(entry, prev_buffer + new_prev_size - entry);
+              return VerifyEncodedEntry(encoded, updated_kv_prot_info);
+            }
             return Status::OK();
           } else if (status == UpdateStatus::UPDATED) {
-            Status s = Add(seq, kTypeValue, key, Slice(str_value));
+            Status s;
+            if (kv_prot_info != nullptr) {
+              ProtectionInfoKVOTS64 updated_kv_prot_info(*kv_prot_info);
+              updated_kv_prot_info.UpdateV(delta, str_value);
+              s = Add(seq, kTypeValue, key, Slice(str_value),
+                      &updated_kv_prot_info);
+            } else {
+              s = Add(seq, kTypeValue, key, Slice(str_value),
+                      nullptr /* kv_prot_info */);
+            }
             RecordTick(moptions_.statistics, NUMBER_KEYS_WRITTEN);
             UpdateFlushState();
             return s;

db/memtable.h

@@ -15,7 +15,9 @@
 #include <string>
 #include <unordered_map>
 #include <vector>
+
 #include "db/dbformat.h"
+#include "db/kv_checksum.h"
 #include "db/range_tombstone_fragmenter.h"
 #include "db/read_callback.h"
 #include "db/version_edit.h"
@@ -24,7 +26,6 @@
 #include "monitoring/instrumented_mutex.h"
 #include "options/cf_options.h"
 #include "rocksdb/db.h"
-#include "rocksdb/env.h"
 #include "rocksdb/memtablerep.h"
 #include "table/multiget_context.h"
 #include "util/dynamic_bloom.h"
@@ -36,6 +37,7 @@ struct FlushJobInfo;
 class Mutex;
 class MemTableIterator;
 class MergeContext;
+class SystemClock;
 
 struct ImmutableMemTableOptions {
   explicit ImmutableMemTableOptions(const ImmutableCFOptions& ioptions,
@@ -175,6 +177,9 @@ class MemTable {
   FragmentedRangeTombstoneIterator* NewRangeTombstoneIterator(
       const ReadOptions& read_options, SequenceNumber read_seq);
 
+  Status VerifyEncodedEntry(Slice encoded,
+                            const ProtectionInfoKVOTS64& kv_prot_info);
+
   // Add an entry into memtable that maps key to value at the
   // specified sequence number and with the specified type.
   // Typically value will be empty if type==kTypeDeletion.
@@ -186,7 +191,8 @@ class MemTable {
   // in the memtable and `MemTableRepFactory::CanHandleDuplicatedKey()` is true.
   // The next attempt should try a larger value for `seq`.
   Status Add(SequenceNumber seq, ValueType type, const Slice& key,
-             const Slice& value, bool allow_concurrent = false,
+             const Slice& value, const ProtectionInfoKVOTS64* kv_prot_info,
+             bool allow_concurrent = false,
              MemTablePostProcessInfo* post_process_info = nullptr,
              void** hint = nullptr);
 
@@ -250,7 +256,8 @@ class MemTable {
   //
   // REQUIRES: external synchronization to prevent simultaneous
   // operations on the same MemTable.
-  Status Update(SequenceNumber seq, const Slice& key, const Slice& value);
+  Status Update(SequenceNumber seq, const Slice& key, const Slice& value,
+                const ProtectionInfoKVOTS64* kv_prot_info);
 
   // If `key` exists in current memtable with type `kTypeValue` and the existing
   // value is at least as large as the new value, updates it in-place. Otherwise
@@ -267,7 +274,8 @@ class MemTable {
   // REQUIRES: external synchronization to prevent simultaneous
   // operations on the same MemTable.
   Status UpdateCallback(SequenceNumber seq, const Slice& key,
-                        const Slice& delta);
+                        const Slice& delta,
+                        const ProtectionInfoKVOTS64* kv_prot_info);
 
   // Returns the number of successive merge entries starting from the newest
   // entry for the key up to the last non-merge entry or last entry for the
@@ -504,7 +512,7 @@ class MemTable {
 
   std::atomic<FlushStateEnum> flush_state_;
 
-  Env* env_;
+  SystemClock* clock_;
 
   // Extract sequential insert prefixes.
   const SliceTransform* insert_with_hint_prefix_extractor_;