Making platform 007 (gcc 7) default in build_detect_platform.sh (#5947 )

Summary: Making platform 007 (gcc 7) default in build_detect_platform.sh. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5947 Differential Revision: D18038837 Pulled By: vjnadimpalli fbshipit-source-id: 9ac2ddaa93bf328a416faec028970e039886378e
Add latest toolchain (gcc-8, etc.) build support for fbcode users (#4923 )
2026-11-07 03:22:34 -05:00 · 2019-10-30 09:46:28 -07:00 · 2019-10-30 09:46:07 -07:00 · 2018-09-13 10:25:20 -07:00 · 2018-09-11 16:58:49 -07:00 · 2018-09-05 13:46:30 -07:00
547 changed files with 7245 additions and 33872 deletions
@@ -45,8 +45,6 @@ etags
 rocksdb_dump
 rocksdb_undump
 db_test2
-trace_analyzer
-trace_analyzer_test

 java/out
 java/target
@@ -71,10 +71,7 @@ install:
      CC=gcc-8 && CXX=g++-8;
    fi
  - if [[ "${JOB_NAME}" == cmake* ]] && [ "${TRAVIS_OS_NAME}" == linux ]; then
-      mkdir cmake-dist && curl --silent --fail --show-error --location https://github.com/Kitware/CMake/releases/download/v3.14.5/cmake-3.14.5-Linux-x86_64.tar.gz | tar --strip-components=1 -C cmake-dist -xz && export PATH=$PWD/cmake-dist/bin:$PATH;
-    fi
-  - if [[ "${JOB_NAME}" == java_test ]]; then
-      java -version && echo "JAVA_HOME=${JAVA_HOME}";
+      mkdir cmake-dist && curl -sfSL https://cmake.org/files/v3.8/cmake-3.8.1-Linux-x86_64.tar.gz | tar --strip-components=1 -C cmake-dist -xz && export PATH=$PWD/cmake-dist/bin:$PATH;
    fi

 before_script:
@@ -93,7 +90,7 @@ script:
      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=db_block_cache_test ROCKSDBTESTS_END=full_filter_block_test make -j4 check_some
      ;;
    2)
-      OPT=-DTRAVIS V=1 make -j4 tools && OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=full_filter_block_test ROCKSDBTESTS_END=write_batch_with_index_test make -j4 check_some
+      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=full_filter_block_test ROCKSDBTESTS_END=write_batch_with_index_test make -j4 check_some
      ;;
    3)
      OPT=-DTRAVIS V=1 ROCKSDBTESTS_START=write_batch_with_index_test ROCKSDBTESTS_END=write_prepared_transaction_test make -j4 check_some
@@ -49,10 +49,6 @@ option(WITH_SNAPPY "build with SNAPPY" OFF)
 option(WITH_LZ4 "build with lz4" OFF)
 option(WITH_ZLIB "build with zlib" OFF)
 option(WITH_ZSTD "build with zstd" OFF)
-option(WITH_WINDOWS_UTF8_FILENAMES "use UTF8 as characterset for opening files, regardles of the system code page" OFF)
-if (WITH_WINDOWS_UTF8_FILENAMES)
-  add_definitions(-DROCKSDB_WINDOWS_UTF8_FILENAMES)
-endif()
 if(MSVC)
  # Defaults currently different for GFLAGS.
  #  We will address find_package work a little later
@@ -340,18 +336,12 @@ endif()

 # Used to run CI build and tests so we can run faster
 option(OPTDBG "Build optimized debug build with MSVC" OFF)
-option(WITH_RUNTIME_DEBUG "build with debug version of runtime library" ON)
 if(MSVC)
  if(OPTDBG)
    message(STATUS "Debug optimization is enabled")
-    set(CMAKE_CXX_FLAGS_DEBUG "/Oxt")
+    set(CMAKE_CXX_FLAGS_DEBUG "/Oxt /${RUNTIME_LIBRARY}d")
  else()
-    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm")
-  endif()
-  if(WITH_RUNTIME_DEBUG)
-    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /${RUNTIME_LIBRARY}d")
-  else()
-    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /${RUNTIME_LIBRARY}")
+    set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm /${RUNTIME_LIBRARY}d")
  endif()
  set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /Oxt /Zp8 /Gm- /Gy /${RUNTIME_LIBRARY}")

@@ -366,7 +356,7 @@ endif()
 option(ROCKSDB_LITE "Build RocksDBLite version" OFF)
 if(ROCKSDB_LITE)
  add_definitions(-DROCKSDB_LITE)
-  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-exceptions -Os")
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-exceptions")
 endif()

 if(CMAKE_SYSTEM_NAME MATCHES "Cygwin")
@@ -471,7 +461,6 @@ set(SOURCES
        db/compaction_iterator.cc
        db/compaction_job.cc
        db/compaction_picker.cc
-        db/compaction_picker_fifo.cc
        db/compaction_picker_universal.cc
        db/convenience.cc
        db/db_filesnapshot.cc
@@ -504,7 +493,6 @@ set(SOURCES
        db/merge_helper.cc
        db/merge_operator.cc
        db/range_del_aggregator.cc
-        db/range_tombstone_fragmenter.cc
        db/repair.cc
        db/snapshot_impl.cc
        db/table_cache.cc
@@ -561,8 +549,6 @@ set(SOURCES
        table/cuckoo_table_builder.cc
        table/cuckoo_table_factory.cc
        table/cuckoo_table_reader.cc
-        table/data_block_hash_index.cc
-        table/data_block_footer.cc
        table/flush_block_policy.cc
        table/format.cc
        table/full_filter_block.cc
@@ -578,7 +564,6 @@ set(SOURCES
        table/plain_table_index.cc
        table/plain_table_key_coding.cc
        table/plain_table_reader.cc
-        table/sst_file_reader.cc
        table/sst_file_writer.cc
        table/table_properties.cc
        table/two_level_iterator.cc
@@ -587,7 +572,6 @@ set(SOURCES
        tools/ldb_cmd.cc
        tools/ldb_tool.cc
        tools/sst_dump_tool.cc
-        tools/trace_analyzer_tool.cc
        util/arena.cc
        util/auto_roll_logger.cc
        util/bloom.cc
@@ -605,7 +589,6 @@ set(SOURCES
        util/filename.cc
        util/filter_policy.cc
        util/hash.cc
-        util/jemalloc_nodump_allocator.cc
        util/log_buffer.cc
        util/murmurhash.cc
        util/random.cc
@@ -613,20 +596,19 @@ set(SOURCES
        util/slice.cc
        util/sst_file_manager_impl.cc
        util/status.cc
+        util/status_message.cc
        util/string_util.cc
        util/sync_point.cc
        util/sync_point_impl.cc
        util/testutil.cc
        util/thread_local.cc
        util/threadpool_imp.cc
-        util/trace_replay.cc
        util/transaction_test_util.cc
        util/xxhash.cc
        utilities/backupable/backupable_db.cc
        utilities/blob_db/blob_compaction_filter.cc
        utilities/blob_db/blob_db.cc
        utilities/blob_db/blob_db_impl.cc
-        utilities/blob_db/blob_db_impl_filesnapshot.cc
        utilities/blob_db/blob_dump_tool.cc
        utilities/blob_db/blob_file.cc
        utilities/blob_db/blob_log_reader.cc
@@ -668,7 +650,6 @@ set(SOURCES
        utilities/simulator_cache/sim_cache.cc
        utilities/spatialdb/spatial_db.cc
        utilities/table_properties_collectors/compact_on_deletion_collector.cc
-        utilities/trace/file_trace_reader_writer.cc
        utilities/transactions/optimistic_transaction_db_impl.cc
        utilities/transactions/optimistic_transaction.cc
        utilities/transactions/pessimistic_transaction.cc
@@ -687,10 +668,12 @@ set(SOURCES
        utilities/write_batch_with_index/write_batch_with_index_internal.cc
        $<TARGET_OBJECTS:build_version>)

-if(HAVE_SSE42 AND NOT MSVC)
-  set_source_files_properties(
-    util/crc32c.cc
-    PROPERTIES COMPILE_FLAGS "-msse4.2 -mpclmul")
+if(HAVE_SSE42 AND NOT FORCE_SSE42)
+  if(NOT MSVC)
+set_source_files_properties(
+  util/crc32c.cc
+      PROPERTIES COMPILE_FLAGS "-msse4.2 -mpclmul")
+  endif()
 endif()

 if(HAVE_POWER8)
@@ -905,8 +888,6 @@ if(WITH_TESTS)
        db/perf_context_test.cc
        db/plain_table_db_test.cc
        db/prefix_test.cc
-        db/range_del_aggregator_test.cc
-        db/range_tombstone_fragmenter_test.cc
        db/repair_test.cc
        db/table_properties_collector_test.cc
        db/version_builder_test.cc
@@ -932,15 +913,12 @@ if(WITH_TESTS)
        table/cleanable_test.cc
        table/cuckoo_table_builder_test.cc
        table/cuckoo_table_reader_test.cc
-        table/data_block_hash_index_test.cc
        table/full_filter_block_test.cc
        table/merger_test.cc
-        table/sst_file_reader_test.cc
        table/table_test.cc
        tools/ldb_cmd_test.cc
        tools/reduce_levels_test.cc
        tools/sst_dump_test.cc
-        tools/trace_analyzer_test.cc
        util/arena_test.cc
        util/auto_roll_logger_test.cc
        util/autovector_test.cc
@@ -955,7 +933,6 @@ if(WITH_TESTS)
        util/hash_test.cc
        util/heap_test.cc
        util/rate_limiter_test.cc
-        util/repeatable_thread_test.cc
        util/slice_transform_test.cc
        util/timer_queue_test.cc
        util/thread_list_test.cc
@@ -998,7 +975,6 @@ if(WITH_TESTS)
  set(BENCHMARKS
    cache/cache_bench.cc
    memtable/memtablerep_bench.cc
-    db/range_del_aggregator_bench.cc
    tools/db_bench.cc
    table/table_reader_bench.cc
    utilities/column_aware_encoding_exp.cc
@@ -1,91 +1,44 @@
 # Rocksdb Change Log
-## 5.18.4 (3/3/2020)
-* Various fixes for ARM64 support (#6250)
-* Fix JEMALLOC_CXX_THROW macro missing from older Jemalloc versions, causing build failures on some platforms.
-
-## 5.18.3 (2/11/2019)
+### 5.15.10 (9/13/2018)
 ### Bug Fixes
-* Fix possible LSM corruption when both range deletions and subcompactions are used. The symptom of this corruption is L1+ files overlapping in the user key space.
-
-## 5.18.2 (01/31/2019)
-
-### Public API Change
-* Change time resolution in FileOperationInfo.
-* Deleting Blob files also go through SStFileManager.
-
-## 5.18.0 (11/30/2018)
-### New Features
-* Introduced `JemallocNodumpAllocator` memory allocator. When being use, block cache will be excluded from core dump.
-* Introduced `PerfContextByLevel` as part of `PerfContext` which allows storing perf context at each level. Also replaced `__thread` with `thread_local` keyword for perf_context. Added per-level perf context for bloom filter and `Get` query.
-* With level_compaction_dynamic_level_bytes = true, level multiplier may be adjusted automatically when Level 0 to 1 compaction is lagged behind.
-* Introduced DB option `atomic_flush`. If true, RocksDB supports flushing multiple column families and atomically committing the result to MANIFEST. Useful when WAL is disabled.
-* Added `num_deletions` and `num_merge_operands` members to `TableProperties`.
-* Added "rocksdb.min-obsolete-sst-number-to-keep" DB property that reports the lower bound on SST file numbers that are being kept from deletion, even if the SSTs are obsolete.
-* Add xxhash64 checksum support
-* Introduced `MemoryAllocator`, which lets the user specify custom memory allocator for block based table.
-* Improved `DeleteRange` to prevent read performance degradation. The feature is no longer marked as experimental.
-* Enabled checkpoint on readonly db (DBImplReadOnly).
-
-### Public API Change
-* `DBOptions::use_direct_reads` now affects reads issued by `BackupEngine` on the database's SSTs.
-* `NO_ITERATORS` is divided into two counters `NO_ITERATOR_CREATED` and `NO_ITERATOR_DELETE`. Both of them are only increasing now, just as other counters.
+* Fix RocksDB Java build and tests.

+### 5.15.9 (9/4/2018)
 ### Bug Fixes
-* Fix corner case where a write group leader blocked due to write stall blocks other writers in queue with WriteOptions::no_slowdown set.
-* Fix in-memory range tombstone truncation to avoid erroneously covering newer keys at a lower level, and include range tombstones in compacted files whose largest key is the range tombstone's start key.
-* Properly set the stop key for a truncated manual CompactRange
-* Fix slow flush/compaction when DB contains many snapshots. The problem became noticeable to us in DBs with 100,000+ snapshots, though it will affect others at different thresholds.
-* Fix the bug that WriteBatchWithIndex's SeekForPrev() doesn't see the entries with the same key.
-* Fix the bug where user comparator was sometimes fed with InternalKey instead of the user key. The bug manifests when during GenerateBottommostFiles.
-* Fix a bug in WritePrepared txns where if the number of old snapshots goes beyond the snapshot cache size (128 default) the rest will not be checked when evicting a commit entry from the commit cache.
-* Fixed Get correctness bug in the presence of range tombstones where merge operands covered by a range tombstone always result in NotFound.
-* Start populating `NO_FILE_CLOSES` ticker statistic, which was always zero previously.
-* The default value of NewBloomFilterPolicy()'s argument use_block_based_builder is changed to false. Note that this new default may cause large temp memory usage when building very large SST files.
-* Fix a deadlock caused by compaction and file ingestion waiting for each other in the event of write stalls.
-* Make DB ignore dropped column families while committing results of atomic flush.
+* Fix compilation errors on OS X clang due to '-Wsuggest-override'.

-## 5.17.0 (10/05/2018)
-### Public API Change
-* `OnTableFileCreated` will now be called for empty files generated during compaction. In that case, `TableFileCreationInfo::file_path` will be "(nil)" and `TableFileCreationInfo::file_size` will be zero.
-* Add `FlushOptions::allow_write_stall`, which controls whether Flush calls start working immediately, even if it causes user writes to stall, or will wait until flush can be performed without causing write stall (similar to `CompactRangeOptions::allow_write_stall`). Note that the default value is false, meaning we add delay to Flush calls until stalling can be avoided when possible. This is behavior change compared to previous RocksDB versions, where Flush calls didn't check if they might cause stall or not.
-* Application using PessimisticTransactionDB is expected to rollback/commit recovered transactions before starting new ones. This assumption is used to skip concurrency control during recovery.
-* Expose column family id to `OnCompactionCompleted`.
-
-### New Features
-* TransactionOptions::skip_concurrency_control allows pessimistic transactions to skip the overhead of concurrency control. Could be used for optimizing certain transactions or during recovery.
+## 5.15.8 (8/31/2018)
+### Bug Fixes
+* Further avoid creating empty SSTs and subsequently deleting them during compaction.

+## 5.15.7 (8/24/2018)
 ### Bug Fixes
 * Avoid creating empty SSTs and subsequently deleting them in certain cases during compaction.
-* Sync CURRENT file contents during checkpoint.

-## 5.16.3 (10/1/2018)
-### Bug Fixes
-* Fix crash caused when `CompactFiles` run with `CompactionOptions::compression == CompressionType::kDisableCompressionOption`. Now that setting causes the compression type to be chosen according to the column family-wide compression options.
-
-## 5.16.2 (9/21/2018)
-### Bug Fixes
-* Fix bug in partition filters with format_version=4.
-
-## 5.16.1 (9/17/2018)
-### Bug Fixes
-* Remove trace_analyzer_tool from rocksdb_lib target in TARGETS file.
-* Fix RocksDB Java build and tests.
-* Remove sync point in Block destructor.
-
-## 5.16.0 (8/21/2018)
+## 5.15.6 (8/21/2018)
 ### Public API Change
 * The merge operands are passed to `MergeOperator::ShouldMerge` in the reversed order relative to how they were merged (passed to FullMerge or FullMergeV2) for performance reasons
-* GetAllKeyVersions() to take an extra argument of `max_num_ikeys`.
-* Using ZSTD dictionary trainer (i.e., setting `CompressionOptions::zstd_max_train_bytes` to a nonzero value) now requires ZSTD version 1.1.3 or later.

-### New Features
-* Changes the format of index blocks by delta encoding the index values, which are the block handles. This saves the encoding of BlockHandle::offset of the non-head index entries in each restart interval. The feature is backward compatible but not forward compatible. It is disabled by default unless format_version 4 or above is used.
-* Add a new tool: trace_analyzer. Trace_analyzer analyzes the trace file generated by using trace_replay API. It can convert the binary format trace file to a human readable txt file, output the statistics of the analyzed query types such as access statistics and size statistics, combining the dumped whole key space file to analyze, support query correlation analyzing, and etc. Current supported query types are: Get, Put, Delete, SingleDelete, DeleteRange, Merge, Iterator (Seek, SeekForPrev only).
-* Add hash index support to data blocks, which helps reducing the cpu utilization of point-lookup operations. This feature is backward compatible with the data block created without the hash index. It is disabled by default unless BlockBasedTableOptions::data_block_index_type is set to data_block_index_type = kDataBlockBinaryAndHash.
+## 5.15.5 (8/16/2018)
+### Bug Fixes
+* Fix VerifyChecksum() API not preserving options

+## 5.15.4 (8/11/2018)
+### Bug Fixes
+* Fix a bug caused by not generating OnTableFileCreated() notification for a 0-byte SST.
+
+## 5.15.3 (8/10/2018)
 ### Bug Fixes
 * Fix a bug in misreporting the estimated partition index size in properties block.

+## 5.15.2 (8/9/2018)
+### Bug Fixes
+* Return correct usable_size for BlockContents.
+
+## 5.15.1 (8/1/2018)
+### Bug Fixes
+* Prevent dereferencing invalid STL iterators when there are range tombstones in ingested files.
+
 ## 5.15.0 (7/17/2018)
 ### Public API Change
 * Remove managed iterator. ReadOptions.managed is not effective anymore.
@@ -95,13 +48,12 @@
 * The "rocksdb.num.entries" table property no longer counts range deletion tombstones as entries.

 ### New Features
-* Changes the format of index blocks by storing the key in their raw form rather than converting them to InternalKey. This saves 8 bytes per index key. The feature is backward compatible but not forward compatible. It is disabled by default unless format_version 3 or above is used.
+* Changes the format of index blocks by storing the key in their raw form rather than converting them to InternalKey. This saves 8 bytes per index key. The feature is backward compatbile but not forward compatible. It is disabled by default unless format_version 3 or above is used.
 * Avoid memcpy when reading mmap files with OpenReadOnly and max_open_files==-1.
 * Support dynamically changing `ColumnFamilyOptions::ttl` via `SetOptions()`.
 * Add a new table property, "rocksdb.num.range-deletions", which counts the number of range deletion tombstones in the table.
 * Improve the performance of iterators doing long range scans by using readahead, when using direct IO.
 * pin_top_level_index_and_filter (default true) in BlockBasedTableOptions can be used in combination with cache_index_and_filter_blocks to prefetch and pin the top-level index of partitioned index and filter blocks in cache. It has no impact when cache_index_and_filter_blocks is false.
-* Write properties meta-block at the end of block-based table to save read-ahead IO.

 ### Bug Fixes
 * Fix deadlock with enable_pipelined_write=true and max_successive_merges > 0
@@ -220,8 +172,7 @@
 * `BackupableDBOptions::max_valid_backups_to_open == 0` now means no backups will be opened during BackupEngine initialization. Previously this condition disabled limiting backups opened.
 * `DBOptions::preserve_deletes` is a new option that allows one to specify that DB should not drop tombstones for regular deletes if they have sequence number larger than what was set by the new API call `DB::SetPreserveDeletesSequenceNumber(SequenceNumber seqnum)`. Disabled by default.
 * API call `DB::SetPreserveDeletesSequenceNumber(SequenceNumber seqnum)` was added, users who wish to preserve deletes are expected to periodically call this function to advance the cutoff seqnum (all deletes made before this seqnum can be dropped by DB). It's user responsibility to figure out how to advance the seqnum in the way so the tombstones are kept for the desired period of time, yet are eventually processed in time and don't eat up too much space.
-* `ReadOptions::iter_start_seqnum` was added;
-if set to something > 0 user will see 2 changes in iterators behavior 1) only keys written with sequence larger than this parameter would be returned and 2) the `Slice` returned by iter->key() now points to the memory that keep User-oriented representation of the internal key, rather than user key. New struct `FullKey` was added to represent internal keys, along with a new helper function `ParseFullKey(const Slice& internal_key, FullKey* result);`.
+* `ReadOptions::iter_start_seqnum` was added; if set to something > 0 user will see 2 changes in iterators behavior 1) only keys written with sequence larger than this parameter would be returned and 2) the `Slice` returned by iter->key() now points to the memory that keep User-oriented representation of the internal key, rather than user key. New struct `FullKey` was added to represent internal keys, along with a new helper function `ParseFullKey(const Slice& internal_key, FullKey* result);`.
 * Deprecate trash_dir param in NewSstFileManager, right now we will rename deleted files to <name>.trash instead of moving them to trash directory
 * Allow setting a custom trash/DB size ratio limit in the SstFileManager, after which files that are to be scheduled for deletion are deleted immediately, regardless of any delete ratelimit.
 * Return an error on write if write_options.sync = true and write_options.disableWAL = true to warn user of inconsistent options. Previously we will not write to WAL and not respecting the sync options in this case.
@@ -43,8 +43,6 @@ to build a portable binary, add `PORTABLE=1` before your make commands, like thi
      command line flags processing. You can compile rocksdb library even
      if you don't have gflags installed.

-* If you wish to build the RocksJava static target, then cmake is required for building Snappy.
-
 ## Supported platforms

 * **Linux - Ubuntu**
@@ -93,38 +93,19 @@ ifeq ($(MAKECMDGOALS),rocksdbjavastaticpublish)
 	DEBUG_LEVEL=0
 endif

-# Lite build flag.
-LITE ?= 0
-ifeq ($(LITE), 0)
-ifneq ($(filter -DROCKSDB_LITE,$(OPT)),)
-  # Be backward compatible and support older format where OPT=-DROCKSDB_LITE is
-  # specified instead of LITE=1 on the command line.
-  LITE=1
-endif
-else ifeq ($(LITE), 1)
-ifeq ($(filter -DROCKSDB_LITE,$(OPT)),)
-	OPT += -DROCKSDB_LITE
-endif
-endif
-
-# Figure out optimize level.
-ifneq ($(DEBUG_LEVEL), 2)
-ifeq ($(LITE), 0)
-	OPT += -O2
-else
-	OPT += -Os
-endif
-endif
-
 # compile with -O2 if debug level is not 2
 ifneq ($(DEBUG_LEVEL), 2)
-OPT += -fno-omit-frame-pointer
+OPT += -O2 -fno-omit-frame-pointer
 # Skip for archs that don't support -momit-leaf-frame-pointer
 ifeq (,$(shell $(CXX) -fsyntax-only -momit-leaf-frame-pointer -xc /dev/null 2>&1))
 OPT += -momit-leaf-frame-pointer
 endif
 endif

+ifeq (,$(shell $(CXX) -fsyntax-only -faligned-new -xc++ /dev/null 2>&1))
+CXXFLAGS += -faligned-new -DHAVE_ALIGNED_NEW
+endif
+
 ifeq (,$(shell $(CXX) -fsyntax-only -maltivec -xc /dev/null 2>&1))
 CXXFLAGS += -DHAS_ALTIVEC
 CFLAGS += -DHAS_ALTIVEC
@@ -137,12 +118,6 @@ CFLAGS +=  -DHAVE_POWER8
 HAVE_POWER8=1
 endif

-ifeq (,$(shell $(CXX) -fsyntax-only -march=armv8-a+crc -xc /dev/null 2>&1))
-CXXFLAGS += -march=armv8-a+crc
-CFLAGS += -march=armv8-a+crc
-ARMCRC_SOURCE=1
-endif
-
 # if we're compiling for release, compile without debug code (-DNDEBUG)
 ifeq ($(DEBUG_LEVEL),0)
 OPT += -DNDEBUG
@@ -347,7 +322,7 @@ endif
 ifeq ("$(wildcard $(LUA_LIB))", "") # LUA_LIB does not exist
 $(error $(LUA_LIB) does not exist.  Try to specify both LUA_PATH and LUA_LIB manually)
 endif
-EXEC_LDFLAGS += $(LUA_LIB)
+LDFLAGS += $(LUA_LIB)

 endif

@@ -407,8 +382,6 @@ VALGRIND_OPTS = --error-exitcode=$(VALGRIND_ERROR) --leak-check=full

 BENCHTOOLOBJECTS = $(BENCH_LIB_SOURCES:.cc=.o) $(LIBOBJECTS) $(TESTUTIL)

-ANALYZETOOLOBJECTS = $(ANALYZER_LIB_SOURCES:.cc=.o)
-
 EXPOBJECTS = $(EXP_LIB_SOURCES:.cc=.o) $(LIBOBJECTS) $(TESTUTIL)

 TESTS = \
@@ -464,7 +437,6 @@ TESTS = \
 	table_properties_collector_test \
 	arena_test \
 	block_test \
-	data_block_hash_index_test \
 	cache_test \
 	corruption_test \
 	slice_transform_test \
@@ -549,6 +521,7 @@ TESTS = \
 	persistent_cache_test \
 	statistics_test \
 	lua_test \
+	range_del_aggregator_test \
 	lru_cache_test \
 	object_registry_test \
 	repair_test \
@@ -556,11 +529,6 @@ TESTS = \
 	write_prepared_transaction_test \
 	write_unprepared_transaction_test \
 	db_universal_compaction_test \
-	trace_analyzer_test \
-	repeatable_thread_test \
-	range_tombstone_fragmenter_test \
-	range_del_aggregator_test \
-	sst_file_reader_test \

 PARALLEL_TEST = \
 	backupable_db_test \
@@ -604,13 +572,12 @@ TOOLS = \
 	rocksdb_dump \
 	rocksdb_undump \
 	blob_dump \
-	trace_analyzer \

 TEST_LIBS = \
 	librocksdb_env_basic_test.a

 # TODO: add back forward_iterator_bench, after making it build in all environemnts.
-BENCHMARKS = db_bench table_reader_bench cache_bench memtablerep_bench column_aware_encoding_exp persistent_cache_bench range_del_aggregator_bench
+BENCHMARKS = db_bench table_reader_bench cache_bench memtablerep_bench column_aware_encoding_exp persistent_cache_bench

 # if user didn't config LIBNAME, set the default
 ifeq ($(LIBNAME),)
@@ -698,7 +665,7 @@ $(SHARED4): $(shared_all_libobjects)
 endif  # PLATFORM_SHARED_EXT

 .PHONY: blackbox_crash_test check clean coverage crash_test ldb_tests package \
-	release tags tags0 valgrind_check whitebox_crash_test format static_lib shared_lib all \
+	release tags valgrind_check whitebox_crash_test format static_lib shared_lib all \
 	dbg rocksdbjavastatic rocksdbjava install install-static install-shared uninstall \
 	analyze tools tools_lib

@@ -911,7 +878,6 @@ crash_test: whitebox_crash_test blackbox_crash_test

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

 ifeq ($(CRASH_TEST_KILL_ODD),)
@@ -920,8 +886,6 @@ endif

 whitebox_crash_test: db_stress
 	python -u tools/db_crashtest.py --simple whitebox --random_kill_odd \
-      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)
-	python -u tools/db_crashtest.py --enable_atomic_flush whitebox  --random_kill_odd \
      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)
 	python -u tools/db_crashtest.py whitebox  --random_kill_odd \
      $(CRASH_TEST_KILL_ODD) $(CRASH_TEST_EXT_ARGS)
@@ -1033,10 +997,8 @@ unity.a: unity.o
 	$(AM_V_AR)rm -f $@
 	$(AM_V_at)$(AR) $(ARFLAGS) $@ unity.o

-
-TOOLLIBOBJECTS = $(TOOL_LIB_SOURCES:.cc=.o)
 # try compiling db_test with unity
-unity_test: db/db_test.o db/db_test_util.o $(TESTHARNESS) $(TOOLLIBOBJECTS) unity.a
+unity_test: db/db_test.o db/db_test_util.o $(TESTHARNESS) unity.a
 	$(AM_LINK)
 	./unity_test

@@ -1056,13 +1018,6 @@ tags:
 	cscope -b `$(FIND) . -name '*.cc'` `$(FIND) . -name '*.h'` `$(FIND) . -name '*.c'`
 	ctags -e -R -o etags *

-tags0:
-	ctags -R .
-	cscope -b `$(FIND) . -name '*.cc' -and ! -name '*_test.cc'` \
-		  `$(FIND) . -name '*.c' -and ! -name '*_test.c'` \
-		  `$(FIND) . -name '*.h' -and ! -name '*_test.h'`
-	ctags -e -R -o etags *
-
 format:
 	build_tools/format-diff.sh

@@ -1076,7 +1031,7 @@ $(LIBRARY): $(LIBOBJECTS)
 	$(AM_V_AR)rm -f $@
 	$(AM_V_at)$(AR) $(ARFLAGS) $@ $(LIBOBJECTS)

-$(TOOLS_LIBRARY): $(BENCH_LIB_SOURCES:.cc=.o) $(TOOL_LIB_SOURCES:.cc=.o) $(LIB_SOURCES:.cc=.o) $(TESTUTIL) $(ANALYZER_LIB_SOURCES:.cc=.o)
+$(TOOLS_LIBRARY): $(BENCH_LIB_SOURCES:.cc=.o) $(TOOL_LIB_SOURCES:.cc=.o) $(LIB_SOURCES:.cc=.o) $(TESTUTIL)
 	$(AM_V_AR)rm -f $@
 	$(AM_V_at)$(AR) $(ARFLAGS) $@ $^

@@ -1087,9 +1042,6 @@ librocksdb_env_basic_test.a: env/env_basic_test.o $(LIBOBJECTS) $(TESTHARNESS)
 db_bench: tools/db_bench.o $(BENCHTOOLOBJECTS)
 	$(AM_LINK)

-trace_analyzer: tools/trace_analyzer.o $(ANALYZETOOLOBJECTS) $(LIBOBJECTS)
-	$(AM_LINK)
-
 cache_bench: cache/cache_bench.o $(LIBOBJECTS) $(TESTUTIL)
 	$(AM_LINK)

@@ -1398,9 +1350,6 @@ table_test: table/table_test.o $(LIBOBJECTS) $(TESTHARNESS)
 block_test: table/block_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-data_block_hash_index_test: table/data_block_hash_index_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 inlineskiplist_test: memtable/inlineskiplist_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1497,9 +1446,6 @@ options_util_test: utilities/options/options_util_test.o $(LIBOBJECTS) $(TESTHAR
 db_bench_tool_test: tools/db_bench_tool_test.o $(BENCHTOOLOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-trace_analyzer_test: tools/trace_analyzer_test.o $(LIBOBJECTS) $(ANALYZETOOLOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 event_logger_test: util/event_logger_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

@@ -1581,21 +1527,9 @@ lua_test: utilities/lua/rocks_lua_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTH
 range_del_aggregator_test: db/range_del_aggregator_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-range_del_aggregator_bench: db/range_del_aggregator_bench.o $(LIBOBJECTS) $(TESTUTIL)
-	$(AM_LINK)
-
 blob_db_test: utilities/blob_db/blob_db_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(AM_LINK)

-repeatable_thread_test: util/repeatable_thread_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
-range_tombstone_fragmenter_test: db/range_tombstone_fragmenter_test.o db/db_test_util.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
-sst_file_reader_test: table/sst_file_reader_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(AM_LINK)
-
 #-------------------------------------------------
 # make install related stuff
 INSTALL_PATH ?= /usr/local
@@ -1641,7 +1575,7 @@ JAVA_INCLUDE = -I$(JAVA_HOME)/include/ -I$(JAVA_HOME)/include/linux
 ifeq ($(PLATFORM), OS_SOLARIS)
 	ARCH := $(shell isainfo -b)
 else ifeq ($(PLATFORM), OS_OPENBSD)
-	ifneq (,$(filter amd64 ppc64 ppc64le arm64 aarch64 sparc64, $(MACHINE)))
+	ifneq (,$(filter $(MACHINE), amd64 arm64 sparc64))
 		ARCH := 64
 	else
 		ARCH := 32
@@ -1650,10 +1584,10 @@ else
 	ARCH := $(shell getconf LONG_BIT)
 endif

-ifneq (,$(filter ppc% arm64 aarch64 sparc64, $(MACHINE)))
-	ROCKSDBJNILIB = librocksdbjni-linux-$(MACHINE).so
+ifeq (,$(findstring ppc,$(MACHINE)))
+        ROCKSDBJNILIB = librocksdbjni-linux$(ARCH).so
 else
-	ROCKSDBJNILIB = librocksdbjni-linux$(ARCH).so
+        ROCKSDBJNILIB = librocksdbjni-linux-$(MACHINE).so
 endif
 ROCKSDB_JAR = rocksdbjni-$(ROCKSDB_MAJOR).$(ROCKSDB_MINOR).$(ROCKSDB_PATCH)-linux$(ARCH).jar
 ROCKSDB_JAR_ALL = rocksdbjni-$(ROCKSDB_MAJOR).$(ROCKSDB_MINOR).$(ROCKSDB_PATCH).jar
@@ -1666,15 +1600,15 @@ ZLIB_SHA256 ?= c3e5e9fdd5004dcb542feda5ee4f0ff0744628baf8ed2dd5d66f8ca1197cb1a1
 ZLIB_DOWNLOAD_BASE ?= http://zlib.net
 BZIP2_VER ?= 1.0.6
 BZIP2_SHA256 ?= a2848f34fcd5d6cf47def00461fcb528a0484d8edef8208d6d2e2909dc61d9cd
-BZIP2_DOWNLOAD_BASE ?= https://downloads.sourceforge.net/project/bzip2
-SNAPPY_VER ?= 1.1.8
-SNAPPY_SHA256 ?= 16b677f07832a612b0836178db7f374e414f94657c138e6993cbfc5dcc58651f
-SNAPPY_DOWNLOAD_BASE ?= https://github.com/google/snappy/archive
-LZ4_VER ?= 1.9.2
-LZ4_SHA256 ?= 658ba6191fa44c92280d4aa2c271b0f4fbc0e34d249578dd05e50e76d0e5efcc
+BZIP2_DOWNLOAD_BASE ?= https://web.archive.org/web/20180624184835/http://www.bzip.org
+SNAPPY_VER ?= 1.1.4
+SNAPPY_SHA256 ?= 134bfe122fd25599bb807bb8130e7ba6d9bdb851e0b16efcb83ac4f5d0b70057
+SNAPPY_DOWNLOAD_BASE ?= https://github.com/google/snappy/releases/download
+LZ4_VER ?= 1.8.0
+LZ4_SHA256 ?= 2ca482ea7a9bb103603108b5a7510b7592b90158c151ff50a28f1ca8389fccf6
 LZ4_DOWNLOAD_BASE ?= https://github.com/lz4/lz4/archive
-ZSTD_VER ?= 1.4.4
-ZSTD_SHA256 ?= a364f5162c7d1a455cc915e8e3cf5f4bd8b75d09bc0f53965b0c9ca1383c52c8
+ZSTD_VER ?= 1.3.3
+ZSTD_SHA256 ?= a77c47153ee7de02626c5b2a097005786b71688be61e9fb81806a011f90b297b
 ZSTD_DOWNLOAD_BASE ?= https://github.com/facebook/zstd/archive
 CURL_SSL_OPTS ?= --tlsv1

@@ -1713,9 +1647,7 @@ endif

 libz.a:
 	-rm -rf zlib-$(ZLIB_VER)
-ifeq (,$(wildcard ./zlib-$(ZLIB_VER).tar.gz))
-	curl --fail --output zlib-$(ZLIB_VER).tar.gz --location ${ZLIB_DOWNLOAD_BASE}/zlib-$(ZLIB_VER).tar.gz
-endif
+	curl -O -L ${ZLIB_DOWNLOAD_BASE}/zlib-$(ZLIB_VER).tar.gz
 	ZLIB_SHA256_ACTUAL=`$(SHA256_CMD) zlib-$(ZLIB_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(ZLIB_SHA256)" != "$$ZLIB_SHA256_ACTUAL" ]; then \
 		echo zlib-$(ZLIB_VER).tar.gz checksum mismatch, expected=\"$(ZLIB_SHA256)\" actual=\"$$ZLIB_SHA256_ACTUAL\"; \
@@ -1727,9 +1659,7 @@ endif

 libbz2.a:
 	-rm -rf bzip2-$(BZIP2_VER)
-ifeq (,$(wildcard ./bzip2-$(BZIP2_VER).tar.gz))
-	curl --fail --output bzip2-$(BZIP2_VER).tar.gz --location ${CURL_SSL_OPTS} ${BZIP2_DOWNLOAD_BASE}/bzip2-$(BZIP2_VER).tar.gz
-endif
+	curl -O -L ${BZIP2_DOWNLOAD_BASE}/$(BZIP2_VER)/bzip2-$(BZIP2_VER).tar.gz
 	BZIP2_SHA256_ACTUAL=`$(SHA256_CMD) bzip2-$(BZIP2_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(BZIP2_SHA256)" != "$$BZIP2_SHA256_ACTUAL" ]; then \
 		echo bzip2-$(BZIP2_VER).tar.gz checksum mismatch, expected=\"$(BZIP2_SHA256)\" actual=\"$$BZIP2_SHA256_ACTUAL\"; \
@@ -1741,24 +1671,21 @@ endif

 libsnappy.a:
 	-rm -rf snappy-$(SNAPPY_VER)
-ifeq (,$(wildcard ./snappy-$(SNAPPY_VER).tar.gz))
-	curl --fail --output snappy-$(SNAPPY_VER).tar.gz --location ${CURL_SSL_OPTS} ${SNAPPY_DOWNLOAD_BASE}/$(SNAPPY_VER).tar.gz
-endif
+	curl -O -L ${CURL_SSL_OPTS} ${SNAPPY_DOWNLOAD_BASE}/$(SNAPPY_VER)/snappy-$(SNAPPY_VER).tar.gz
 	SNAPPY_SHA256_ACTUAL=`$(SHA256_CMD) snappy-$(SNAPPY_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(SNAPPY_SHA256)" != "$$SNAPPY_SHA256_ACTUAL" ]; then \
 		echo snappy-$(SNAPPY_VER).tar.gz checksum mismatch, expected=\"$(SNAPPY_SHA256)\" actual=\"$$SNAPPY_SHA256_ACTUAL\"; \
 		exit 1; \
 	fi
 	tar xvzf snappy-$(SNAPPY_VER).tar.gz
-	mkdir snappy-$(SNAPPY_VER)/build
-	cd snappy-$(SNAPPY_VER)/build && CFLAGS='${EXTRA_CFLAGS}' CXXFLAGS='${EXTRA_CXXFLAGS}' LDFLAGS='${EXTRA_LDFLAGS}' cmake -DCMAKE_POSITION_INDEPENDENT_CODE=ON .. && $(MAKE) ${SNAPPY_MAKE_TARGET}
-	cp snappy-$(SNAPPY_VER)/build/libsnappy.a .
+	cd snappy-$(SNAPPY_VER) && CFLAGS='${EXTRA_CFLAGS}' CXXFLAGS='${EXTRA_CXXFLAGS}' LDFLAGS='${EXTRA_LDFLAGS}' ./configure --with-pic --enable-static --disable-shared
+	cd snappy-$(SNAPPY_VER) && $(MAKE) ${SNAPPY_MAKE_TARGET}
+	cp snappy-$(SNAPPY_VER)/.libs/libsnappy.a .

 liblz4.a:
 	-rm -rf lz4-$(LZ4_VER)
-ifeq (,$(wildcard ./lz4-$(LZ4_VER).tar.gz))
-	curl --fail --output lz4-$(LZ4_VER).tar.gz --location ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
-endif
+	curl -O -L ${CURL_SSL_OPTS} ${LZ4_DOWNLOAD_BASE}/v$(LZ4_VER).tar.gz
+	mv v$(LZ4_VER).tar.gz lz4-$(LZ4_VER).tar.gz
 	LZ4_SHA256_ACTUAL=`$(SHA256_CMD) lz4-$(LZ4_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(LZ4_SHA256)" != "$$LZ4_SHA256_ACTUAL" ]; then \
 		echo lz4-$(LZ4_VER).tar.gz checksum mismatch, expected=\"$(LZ4_SHA256)\" actual=\"$$LZ4_SHA256_ACTUAL\"; \
@@ -1770,9 +1697,8 @@ endif

 libzstd.a:
 	-rm -rf zstd-$(ZSTD_VER)
-ifeq (,$(wildcard ./zstd-$(ZSTD_VER).tar.gz))
-	curl --fail --output zstd-$(ZSTD_VER).tar.gz --location ${CURL_SSL_OPTS} ${ZSTD_DOWNLOAD_BASE}/v$(ZSTD_VER).tar.gz
-endif
+	curl -O -L ${CURL_SSL_OPTS} ${ZSTD_DOWNLOAD_BASE}/v$(ZSTD_VER).tar.gz
+	mv v$(ZSTD_VER).tar.gz zstd-$(ZSTD_VER).tar.gz
 	ZSTD_SHA256_ACTUAL=`$(SHA256_CMD) zstd-$(ZSTD_VER).tar.gz | cut -d ' ' -f 1`; \
 	if [ "$(ZSTD_SHA256)" != "$$ZSTD_SHA256_ACTUAL" ]; then \
 		echo zstd-$(ZSTD_VER).tar.gz checksum mismatch, expected=\"$(ZSTD_SHA256)\" actual=\"$$ZSTD_SHA256_ACTUAL\"; \
@@ -1834,26 +1760,20 @@ rocksdbjavastaticrelease: rocksdbjavastatic
 	cd java/target;jar -uf $(ROCKSDB_JAR_ALL) librocksdbjni-*.so librocksdbjni-*.jnilib
 	cd java/target/classes;jar -uf ../$(ROCKSDB_JAR_ALL) org/rocksdb/*.class org/rocksdb/util/*.class

-rocksdbjavastaticreleasedocker: rocksdbjavastatic rocksdbjavastaticdockerx86 rocksdbjavastaticdockerx86_64
-	cd java;jar -cf target/$(ROCKSDB_JAR_ALL) HISTORY*.md
-	cd java/target;jar -uf $(ROCKSDB_JAR_ALL) librocksdbjni-*.so librocksdbjni-*.jnilib
-	cd java/target/classes;jar -uf ../$(ROCKSDB_JAR_ALL) org/rocksdb/*.class org/rocksdb/util/*.class
-
-rocksdbjavastaticdockerx86:
-	mkdir -p java/target
-	DOCKER_LINUX_X86_CONTAINER=`docker ps -aqf name=rocksdb_linux_x86-be`; \
-	if [ -z "$$DOCKER_LINUX_X86_CONTAINER" ]; then \
-		docker container create --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host --name rocksdb_linux_x86-be evolvedbinary/rocksjava:centos6_x86-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh; \
-	fi
-	docker start -a rocksdb_linux_x86-be
-
-rocksdbjavastaticdockerx86_64:
-	mkdir -p java/target
+rocksdbjavastaticreleasedocker: rocksdbjavastatic
 	DOCKER_LINUX_X64_CONTAINER=`docker ps -aqf name=rocksdb_linux_x64-be`; \
 	if [ -z "$$DOCKER_LINUX_X64_CONTAINER" ]; then \
 		docker container create --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host --name rocksdb_linux_x64-be evolvedbinary/rocksjava:centos6_x64-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh; \
 	fi
 	docker start -a rocksdb_linux_x64-be
+	DOCKER_LINUX_X86_CONTAINER=`docker ps -aqf name=rocksdb_linux_x86-be`; \
+	if [ -z "$$DOCKER_LINUX_X86_CONTAINER" ]; then \
+		docker container create --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host --name rocksdb_linux_x86-be evolvedbinary/rocksjava:centos6_x86-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh; \
+	fi
+	docker start -a rocksdb_linux_x86-be
+	cd java;jar -cf target/$(ROCKSDB_JAR_ALL) HISTORY*.md
+	cd java/target;jar -uf $(ROCKSDB_JAR_ALL) librocksdbjni-*.so librocksdbjni-*.jnilib
+	cd java/target/classes;jar -uf ../$(ROCKSDB_JAR_ALL) org/rocksdb/*.class org/rocksdb/util/*.class

 rocksdbjavastaticdockerppc64le:
 	mkdir -p java/target
@@ -1863,14 +1783,6 @@ rocksdbjavastaticdockerppc64le:
 	fi
 	docker start -a rocksdb_linux_ppc64le-be

-rocksdbjavastaticdockerarm64v8:
-	mkdir -p java/target
-	DOCKER_LINUX_ARM64V8_CONTAINER=`docker ps -aqf name=rocksdb_linux_arm64v8-be`; \
-	if [ -z "$$DOCKER_LINUX_ARM64V8_CONTAINER" ]; then \
-		docker container create --attach stdin --attach stdout --attach stderr --volume `pwd`:/rocksdb-host --name rocksdb_linux_arm64v8-be evolvedbinary/rocksjava:centos7_arm64v8-be /rocksdb-host/java/crossbuild/docker-build-linux-centos.sh; \
-	fi
-	docker start -a rocksdb_linux_arm64v8-be
-
 rocksdbjavastaticpublish: rocksdbjavastaticrelease rocksdbjavastaticpublishcentral

 rocksdbjavastaticpublishdocker: rocksdbjavastaticreleasedocker rocksdbjavastaticpublishcentral
@@ -1986,7 +1898,7 @@ endif
 #  	Source files dependencies detection
 # ---------------------------------------------------------------------------

-all_sources = $(LIB_SOURCES) $(MAIN_SOURCES) $(MOCK_LIB_SOURCES) $(TOOL_LIB_SOURCES) $(BENCH_LIB_SOURCES) $(TEST_LIB_SOURCES) $(EXP_LIB_SOURCES) $(ANALYZER_LIB_SOURCES)
+all_sources = $(LIB_SOURCES) $(MAIN_SOURCES) $(MOCK_LIB_SOURCES) $(TOOL_LIB_SOURCES) $(BENCH_LIB_SOURCES) $(TEST_LIB_SOURCES) $(EXP_LIB_SOURCES)
 DEPFILES = $(all_sources:.cc=.cc.d)

 # Add proper dependency support so changing a .h file forces a .cc file to
@@ -1,5 +1,3 @@
-load("@fbcode_macros//build_defs:auto_headers.bzl", "AutoHeaders")
-
 REPO_PATH = package_name() + "/"

 BUCK_BINS = "buck-out/gen/" + REPO_PATH
@@ -67,14 +65,6 @@ is_opt_mode = build_mode.startswith("opt")
 if is_opt_mode:
    rocksdb_compiler_flags.append("-DNDEBUG")

-sanitizer = read_config("fbcode", "sanitizer")
-
-# Do not enable jemalloc if sanitizer presents. RocksDB will further detect
-# whether the binary is linked with jemalloc at runtime.
-if sanitizer == "":
-    rocksdb_compiler_flags.append("-DROCKSDB_JEMALLOC")
-    rocksdb_external_deps.append(("jemalloc", None, "headers"))
-
 cpp_library(
    name = "rocksdb_lib",
    srcs = [
@@ -89,7 +79,6 @@ cpp_library(
        "db/compaction_iterator.cc",
        "db/compaction_job.cc",
        "db/compaction_picker.cc",
-        "db/compaction_picker_fifo.cc",
        "db/compaction_picker_universal.cc",
        "db/convenience.cc",
        "db/db_filesnapshot.cc",
@@ -122,7 +111,6 @@ cpp_library(
        "db/merge_helper.cc",
        "db/merge_operator.cc",
        "db/range_del_aggregator.cc",
-        "db/range_tombstone_fragmenter.cc",
        "db/repair.cc",
        "db/snapshot_impl.cc",
        "db/table_cache.cc",
@@ -183,8 +171,6 @@ cpp_library(
        "table/cuckoo_table_builder.cc",
        "table/cuckoo_table_factory.cc",
        "table/cuckoo_table_reader.cc",
-        "table/data_block_footer.cc",
-        "table/data_block_hash_index.cc",
        "table/flush_block_policy.cc",
        "table/format.cc",
        "table/full_filter_block.cc",
@@ -200,7 +186,6 @@ cpp_library(
        "table/plain_table_index.cc",
        "table/plain_table_key_coding.cc",
        "table/plain_table_reader.cc",
-        "table/sst_file_reader.cc",
        "table/sst_file_writer.cc",
        "table/table_properties.cc",
        "table/two_level_iterator.cc",
@@ -226,7 +211,6 @@ cpp_library(
        "util/filename.cc",
        "util/filter_policy.cc",
        "util/hash.cc",
-        "util/jemalloc_nodump_allocator.cc",
        "util/log_buffer.cc",
        "util/murmurhash.cc",
        "util/random.cc",
@@ -234,19 +218,18 @@ cpp_library(
        "util/slice.cc",
        "util/sst_file_manager_impl.cc",
        "util/status.cc",
+        "util/status_message.cc",
        "util/string_util.cc",
        "util/sync_point.cc",
        "util/sync_point_impl.cc",
        "util/thread_local.cc",
        "util/threadpool_imp.cc",
-        "util/trace_replay.cc",
        "util/transaction_test_util.cc",
        "util/xxhash.cc",
        "utilities/backupable/backupable_db.cc",
        "utilities/blob_db/blob_compaction_filter.cc",
        "utilities/blob_db/blob_db.cc",
        "utilities/blob_db/blob_db_impl.cc",
-        "utilities/blob_db/blob_db_impl_filesnapshot.cc",
        "utilities/blob_db/blob_dump_tool.cc",
        "utilities/blob_db/blob_file.cc",
        "utilities/blob_db/blob_log_format.cc",
@@ -286,7 +269,6 @@ cpp_library(
        "utilities/simulator_cache/sim_cache.cc",
        "utilities/spatialdb/spatial_db.cc",
        "utilities/table_properties_collectors/compact_on_deletion_collector.cc",
-        "utilities/trace/file_trace_reader_writer.cc",
        "utilities/transactions/optimistic_transaction.cc",
        "utilities/transactions/optimistic_transaction_db_impl.cc",
        "utilities/transactions/pessimistic_transaction.cc",
@@ -304,7 +286,7 @@ cpp_library(
        "utilities/write_batch_with_index/write_batch_with_index.cc",
        "utilities/write_batch_with_index/write_batch_with_index_internal.cc",
    ],
-    auto_headers = AutoHeaders.RECURSIVE_GLOB,
+    headers = AutoHeaders.RECURSIVE_GLOB,
    arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
    compiler_flags = rocksdb_compiler_flags,
    preprocessor_flags = rocksdb_preprocessor_flags,
@@ -317,7 +299,6 @@ cpp_library(
    srcs = [
        "db/db_test_util.cc",
        "table/mock_table.cc",
-        "tools/trace_analyzer_tool.cc",
        "util/fault_injection_test_env.cc",
        "util/testharness.cc",
        "util/testutil.cc",
@@ -326,7 +307,7 @@ cpp_library(
        "utilities/col_buf_encoder.cc",
        "utilities/column_aware_encoding_util.cc",
    ],
-    auto_headers = AutoHeaders.RECURSIVE_GLOB,
+    headers = AutoHeaders.RECURSIVE_GLOB,
    arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
    compiler_flags = rocksdb_compiler_flags,
    preprocessor_flags = rocksdb_preprocessor_flags,
@@ -338,10 +319,9 @@ cpp_library(
    name = "rocksdb_tools_lib",
    srcs = [
        "tools/db_bench_tool.cc",
-        "tools/trace_analyzer_tool.cc",
        "util/testutil.cc",
    ],
-    auto_headers = AutoHeaders.RECURSIVE_GLOB,
+    headers = AutoHeaders.RECURSIVE_GLOB,
    arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
    compiler_flags = rocksdb_compiler_flags,
    preprocessor_flags = rocksdb_preprocessor_flags,
@@ -352,7 +332,7 @@ cpp_library(
 cpp_library(
    name = "env_basic_test_lib",
    srcs = ["env/env_basic_test.cc"],
-    auto_headers = AutoHeaders.RECURSIVE_GLOB,
+    headers = AutoHeaders.RECURSIVE_GLOB,
    arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
    compiler_flags = rocksdb_compiler_flags,
    preprocessor_flags = rocksdb_preprocessor_flags,
@@ -517,11 +497,6 @@ ROCKS_TESTS = [
        "table/cuckoo_table_reader_test.cc",
        "serial",
    ],
-    [
-        "data_block_hash_index_test",
-        "table/data_block_hash_index_test.cc",
-        "serial",
-    ],
    [
        "date_tiered_test",
        "utilities/date_tiered/date_tiered_test.cc",
@@ -932,11 +907,6 @@ ROCKS_TESTS = [
        "db/range_del_aggregator_test.cc",
        "serial",
    ],
-    [
-        "range_tombstone_fragmenter_test",
-        "db/range_tombstone_fragmenter_test.cc",
-        "serial",
-    ],
    [
        "rate_limiter_test",
        "util/rate_limiter_test.cc",
@@ -952,11 +922,6 @@ ROCKS_TESTS = [
        "db/repair_test.cc",
        "serial",
    ],
-    [
-        "repeatable_thread_test",
-        "util/repeatable_thread_test.cc",
-        "serial",
-    ],
    [
        "sim_cache_test",
        "utilities/simulator_cache/sim_cache_test.cc",
@@ -982,11 +947,6 @@ ROCKS_TESTS = [
        "tools/sst_dump_test.cc",
        "serial",
    ],
-    [
-        "sst_file_reader_test",
-        "table/sst_file_reader_test.cc",
-        "serial",
-    ],
    [
        "statistics_test",
        "monitoring/statistics_test.cc",
@@ -1022,11 +982,6 @@ ROCKS_TESTS = [
        "util/timer_queue_test.cc",
        "serial",
    ],
-    [
-        "trace_analyzer_test",
-        "tools/trace_analyzer_test.cc",
-        "serial",
-    ],
    [
        "transaction_test",
        "utilities/transactions/transaction_test.cc",
@@ -1109,19 +1064,20 @@ if not is_opt_mode:
        ttype = "gtest" if test_cfg[2] == "parallel" else "simple"
        test_bin = test_name + "_bin"

-        cpp_binary(
-            name = test_bin,
-            srcs = [test_cc],
-            arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
-            compiler_flags = rocksdb_compiler_flags,
-            preprocessor_flags = rocksdb_preprocessor_flags,
-            deps = [":rocksdb_test_lib"],
-            external_deps = rocksdb_external_deps,
+        cpp_binary (
+          name = test_bin,
+          srcs = [test_cc],
+          deps = [":rocksdb_test_lib"],
+          preprocessor_flags = rocksdb_preprocessor_flags,
+          arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
+          compiler_flags = rocksdb_compiler_flags,
+          external_deps = rocksdb_external_deps,
        )

        custom_unittest(
-            name = test_name,
-            command = [TEST_RUNNER, BUCK_BINS + test_bin],
-            type = ttype,
-            deps = [":" + test_bin],
+          name = test_name,
+          type = ttype,
+          deps = [":" + test_bin],
+          command = [TEST_RUNNER, BUCK_BINS + test_bin]
        )
+
@@ -12,7 +12,6 @@ At Facebook, we use RocksDB as storage engines in multiple data management servi
 6. LogDevice -- a distributed data store for logs [2]

 [1] https://research.facebook.com/publications/realtime-data-processing-at-facebook/
-
 [2] https://code.facebook.com/posts/357056558062811/logdevice-a-distributed-data-store-for-logs/

 ## LinkedIn
@@ -89,6 +88,3 @@ LzLabs is using RocksDB as a storage engine in their multi-database distributed

 ## ProfaneDB
 [ProfaneDB](https://profanedb.gitlab.io/) is a database for Protocol Buffers, and uses RocksDB for storage. It is accessible via gRPC, and the schema is defined using directly `.proto` files.
-
-## 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.
@@ -1,75 +1,15 @@
 version: 1.0.{build}
-
 image: Visual Studio 2017
-
-environment:
-  JAVA_HOME: C:\Program Files\Java\jdk1.8.0
-  THIRDPARTY_HOME: $(APPVEYOR_BUILD_FOLDER)\thirdparty
-  SNAPPY_HOME: $(THIRDPARTY_HOME)\snappy-1.1.7
-  SNAPPY_INCLUDE: $(SNAPPY_HOME);$(SNAPPY_HOME)\build
-  SNAPPY_LIB_DEBUG: $(SNAPPY_HOME)\build\Debug\snappy.lib
-  SNAPPY_LIB_RELEASE: $(SNAPPY_HOME)\build\Release\snappy.lib
-  LZ4_HOME: $(THIRDPARTY_HOME)\lz4-1.8.3
-  LZ4_INCLUDE: $(LZ4_HOME)\lib
-  LZ4_LIB_DEBUG: $(LZ4_HOME)\visual\VS2010\bin\x64_Debug\liblz4_static.lib
-  LZ4_LIB_RELEASE: $(LZ4_HOME)\visual\VS2010\bin\x64_Release\liblz4_static.lib
-  ZSTD_HOME: $(THIRDPARTY_HOME)\zstd-1.4.0
-  ZSTD_INCLUDE: $(ZSTD_HOME)\lib;$(ZSTD_HOME)\lib\dictBuilder
-  ZSTD_LIB_DEBUG: $(ZSTD_HOME)\build\VS2010\bin\x64_Debug\libzstd_static.lib
-  ZSTD_LIB_RELEASE: $(ZSTD_HOME)\build\VS2010\bin\x64_Release\libzstd_static.lib
-  matrix:
-    - APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2015
-      CMAKE_GENERATOR: Visual Studio 14 Win64
-      DEV_ENV: C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\devenv.com
-    - APPVEYOR_BUILD_WORKER_IMAGE: Visual Studio 2017
-      CMAKE_GENERATOR: Visual Studio 15 Win64
-      DEV_ENV: C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\Common7\IDE\devenv.com
-
-install:
-  - md %THIRDPARTY_HOME%
-  - echo "Building Snappy dependency..."
-  - cd %THIRDPARTY_HOME%
-  - curl --fail --silent --show-error --output snappy-1.1.7.zip --location https://github.com/google/snappy/archive/1.1.7.zip
-  - unzip snappy-1.1.7.zip
-  - cd snappy-1.1.7
-  - mkdir build
-  - cd build
-  - cmake -G "%CMAKE_GENERATOR%" ..
-  - msbuild Snappy.sln /p:Configuration=Debug /p:Platform=x64
-  - msbuild Snappy.sln /p:Configuration=Release /p:Platform=x64
-  - echo "Building LZ4 dependency..."
-  - cd %THIRDPARTY_HOME%
-  - curl --fail --silent --show-error --output lz4-1.8.3.zip --location https://github.com/lz4/lz4/archive/v1.8.3.zip
-  - unzip lz4-1.8.3.zip
-  - cd lz4-1.8.3\visual\VS2010
-  - ps: $CMD="$Env:DEV_ENV"; & $CMD lz4.sln /upgrade
-  - msbuild lz4.sln /p:Configuration=Debug /p:Platform=x64
-  - msbuild lz4.sln /p:Configuration=Release /p:Platform=x64
-  - echo "Building ZStd dependency..."
-  - cd %THIRDPARTY_HOME%
-  - curl --fail --silent --show-error --output zstd-1.4.0.zip --location https://github.com/facebook/zstd/archive/v1.4.0.zip
-  - unzip zstd-1.4.0.zip
-  - cd zstd-1.4.0\build\VS2010
-  - ps: $CMD="$Env:DEV_ENV"; & $CMD zstd.sln /upgrade
-  - msbuild zstd.sln /p:Configuration=Debug /p:Platform=x64
-  - msbuild zstd.sln /p:Configuration=Release /p:Platform=x64
-
 before_build:
-  - md %APPVEYOR_BUILD_FOLDER%\build
-  - cd %APPVEYOR_BUILD_FOLDER%\build
-  - cmake -G "%CMAKE_GENERATOR%" -DCMAKE_BUILD_TYPE=Debug -DOPTDBG=1 -DPORTABLE=1 -DSNAPPY=1 -DLZ4=1 -DZSTD=1 -DXPRESS=1 -DJNI=1 ..
-  - cd ..
-
+- md %APPVEYOR_BUILD_FOLDER%\build
+- cd %APPVEYOR_BUILD_FOLDER%\build
+- cmake -G "Visual Studio 15 Win64" -DOPTDBG=1 -DWITH_XPRESS=1 -DPORTABLE=1 ..
+- cd ..
 build:
  project: build\rocksdb.sln
  parallel: true
  verbosity: normal
-
 test:
-
 test_script:
-  - ps: build_tools\run_ci_db_test.ps1 -SuiteRun db_basic_test,db_test2,db_test,env_basic_test,env_test,db_merge_operand_test -Concurrency 8
-
-on_failure:
-  - cmd: 7z a build-failed.zip %APPVEYOR_BUILD_FOLDER%\build\ && appveyor PushArtifact build-failed.zip
+- ps: build_tools\run_ci_db_test.ps1 -SuiteRun db_basic_test,db_test2,db_test,env_basic_test,env_test -Concurrency 8

@@ -109,14 +109,12 @@ def generate_targets(repo_path):
        "rocksdb_test_lib",
        src_mk.get("MOCK_LIB_SOURCES", []) +
        src_mk.get("TEST_LIB_SOURCES", []) +
-        src_mk.get("EXP_LIB_SOURCES", []) +
-        src_mk.get("ANALYZER_LIB_SOURCES", []),
+        src_mk.get("EXP_LIB_SOURCES", []),
        [":rocksdb_lib"])
    # rocksdb_tools_lib
    TARGETS.add_library(
        "rocksdb_tools_lib",
        src_mk.get("BENCH_LIB_SOURCES", []) +
-        src_mk.get("ANALYZER_LIB_SOURCES", []) +
        ["util/testutil.cc"],
        [":rocksdb_lib"])

@@ -10,7 +10,7 @@ def pretty_list(lst, indent=8):

    if len(lst) == 1:
        return "\"%s\"" % lst[0]
-
+    
    separator = "\",\n%s\"" % (" " * indent)
    res = separator.join(sorted(lst))
    res = "\n" + (" " * indent) + "\"" + res + "\",\n" + (" " * (indent - 4))
@@ -31,16 +31,13 @@ class TARGETSBuilder:
        self.targets_file.close()

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

    def add_binary(self, name, srcs, deps=None):
@@ -2,9 +2,7 @@ from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
 from __future__ import unicode_literals
-rocksdb_target_header = """load("@fbcode_macros//build_defs:auto_headers.bzl", "AutoHeaders")
-
-REPO_PATH = package_name() + "/"
+rocksdb_target_header = """REPO_PATH = package_name() + "/"

 BUCK_BINS = "buck-out/gen/" + REPO_PATH

@@ -70,26 +68,18 @@ is_opt_mode = build_mode.startswith("opt")
 # doesn't harm and avoid forgetting to add it.
 if is_opt_mode:
    rocksdb_compiler_flags.append("-DNDEBUG")
-
-sanitizer = read_config("fbcode", "sanitizer")
-
-# Do not enable jemalloc if sanitizer presents. RocksDB will further detect
-# whether the binary is linked with jemalloc at runtime.
-if sanitizer == "":
-    rocksdb_compiler_flags.append("-DROCKSDB_JEMALLOC")
-    rocksdb_external_deps.append(("jemalloc", None, "headers"))
 """


 library_template = """
 cpp_library(
-    name = "{name}",
-    srcs = [{srcs}],
-    {headers_attr_prefix}headers = {headers},
+    name = "%s",
+    srcs = [%s],
+    headers = %s,
    arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
    compiler_flags = rocksdb_compiler_flags,
    preprocessor_flags = rocksdb_preprocessor_flags,
-    deps = [{deps}],
+    deps = [%s],
    external_deps = rocksdb_external_deps,
 )
 """
@@ -128,20 +118,21 @@ if not is_opt_mode:
        ttype = "gtest" if test_cfg[2] == "parallel" else "simple"
        test_bin = test_name + "_bin"

-        cpp_binary(
-            name = test_bin,
-            srcs = [test_cc],
-            arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
-            compiler_flags = rocksdb_compiler_flags,
-            preprocessor_flags = rocksdb_preprocessor_flags,
-            deps = [":rocksdb_test_lib"],
-            external_deps = rocksdb_external_deps,
+        cpp_binary (
+          name = test_bin,
+          srcs = [test_cc],
+          deps = [":rocksdb_test_lib"],
+          preprocessor_flags = rocksdb_preprocessor_flags,
+          arch_preprocessor_flags = rocksdb_arch_preprocessor_flags,
+          compiler_flags = rocksdb_compiler_flags,
+          external_deps = rocksdb_external_deps,
        )

        custom_unittest(
-            name = test_name,
-            command = [TEST_RUNNER, BUCK_BINS + test_bin],
-            type = ttype,
-            deps = [":" + test_bin],
+          name = test_name,
+          type = ttype,
+          deps = [":" + test_bin],
+          command = [TEST_RUNNER, BUCK_BINS + test_bin]
        )
+
 """
@@ -7,12 +7,12 @@
 // Name of the environment variables which need to be set by the entity which
 // triggers continuous runs so that code at the end of the file gets executed
 // and Sandcastle run starts.
-const ENV_POST_RECEIVE_HOOK = "POST_RECEIVE_HOOK";
-const ENV_HTTPS_APP_VALUE = "HTTPS_APP_VALUE";
-const ENV_HTTPS_TOKEN_VALUE = "HTTPS_TOKEN_VALUE";
+define("ENV_POST_RECEIVE_HOOK", "POST_RECEIVE_HOOK");
+define("ENV_HTTPS_APP_VALUE", "HTTPS_APP_VALUE");
+define("ENV_HTTPS_TOKEN_VALUE", "HTTPS_TOKEN_VALUE");

-const PRIMARY_TOKEN_FILE = '/home/krad/.sandcastle';
-const CONT_RUN_ALIAS = "leveldb";
+define("PRIMARY_TOKEN_FILE", '/home/krad/.sandcastle');
+define("CONT_RUN_ALIAS", "leveldb");

 //////////////////////////////////////////////////////////////////////
 /*  Run tests in sandcastle */
@@ -476,17 +476,6 @@ EOF
            COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_SCHED_GETCPU_PRESENT"
        fi
    fi
-
-    if ! test $ROCKSDB_DISABLE_ALIGNED_NEW; then
-        # Test whether c++17 aligned-new is supported
-        $CXX $PLATFORM_CXXFLAGS -faligned-new -x c++ - -o /dev/null 2>/dev/null <<EOF
-            struct alignas(1024) t {int a;};
-            int main() {}
-EOF
-        if [ "$?" = 0 ]; then
-            PLATFORM_CXXFLAGS="$PLATFORM_CXXFLAGS -faligned-new -DHAVE_ALIGNED_NEW"
-        fi
-    fi
 fi

 # TODO(tec): Fix -Wshorten-64-to-32 errors on FreeBSD and enable the warning.
@@ -527,8 +516,6 @@ if test -z "$PORTABLE"; then
  elif test -n "`echo $TARGET_ARCHITECTURE | grep ^arm`"; then
    # TODO: Handle this with approprite options.
    COMMON_FLAGS="$COMMON_FLAGS"
-  elif test -n "`echo $TARGET_ARCHITECTURE | grep ^aarch64`"; then
-    COMMON_FLAGS="$COMMON_FLAGS"
  elif [ "$TARGET_OS" == "IOS" ]; then
    COMMON_FLAGS="$COMMON_FLAGS"
  elif [ "$TARGET_OS" != "AIX" ] && [ "$TARGET_OS" != "SunOS" ]; then
@@ -1,18 +1,19 @@
-GCC_BASE=/mnt/gvfs/third-party2/gcc/112ec378fec7002ad3e09afde022e656049f7191/5.x/centos7-native/c447969
-CLANG_BASE=/mnt/gvfs/third-party2/llvm-fb/04999bdb3ce81a11073535dcb00b5e13dc1cbaf5/stable/centos7-native/c9f9104
-LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/92b0c8e5c8eecc71eb042594ce1ab3413799b385/5.x/gcc-5-glibc-2.23/339d858
-GLIBC_BASE=/mnt/gvfs/third-party2/glibc/3d8698d5973ba94f41620a80a67e4457fdf01e90/2.23/gcc-5-glibc-2.23/ca1d1c0
-SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/7f9bdaada18f59bc27ec2b0871eb8a6144343aef/1.1.3/gcc-5-glibc-2.23/9bc6787
-ZLIB_BASE=/mnt/gvfs/third-party2/zlib/22c2d65676fb7c23cfa797c4f6937f38b026f3cf/1.2.8/gcc-5-glibc-2.23/9bc6787
-BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/dc49a21c5fceec6456a7a28a94dcd16690af1337/1.0.6/gcc-5-glibc-2.23/9bc6787
-LZ4_BASE=/mnt/gvfs/third-party2/lz4/907b498203d297947f3bb70b9466f47e100f1873/r131/gcc-5-glibc-2.23/9bc6787
-ZSTD_BASE=/mnt/gvfs/third-party2/zstd/af6628a46758f1a15484a1760cd7294164bc5ba1/1.3.5/gcc-5-glibc-2.23/03859b5
-GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/0b9929d2588991c65a57168bf88aff2db87c5d48/2.2.0/gcc-5-glibc-2.23/9bc6787
-JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/b1a0e56c1e3e6929813a4331ade3a58ff083afbb/master/gcc-5-glibc-2.23/aa64d6b
-NUMA_BASE=/mnt/gvfs/third-party2/numa/9cbf2460284c669ed19c3ccb200a71f7dd7e53c7/2.0.11/gcc-5-glibc-2.23/9bc6787
-LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/bf3d7497fe4e6d007354f0adffa16ce3003f8338/1.3/gcc-5-glibc-2.23/b443de1
-TBB_BASE=/mnt/gvfs/third-party2/tbb/ff4e0b093534704d8abab678a4fd7f5ea7b094c7/2018_U5/gcc-5-glibc-2.23/9bc6787
-KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/b5c4a61a5c483ba24722005ae07895971a2ac707/4.0.9-36_fbk5_2933_gd092e3f/gcc-5-glibc-2.23/da39a3e
-BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/55031de95a2b46c82948743419a603b3d6aefe28/2.29.1/centos7-native/da39a3e
-VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/f3f697a28122e6bcd513273dd9c1ff23852fc59f/3.13.0/gcc-5-glibc-2.23/9bc6787
-LUA_BASE=/mnt/gvfs/third-party2/lua/f0cd714433206d5139df61659eb7b28b1dea6683/5.2.3/gcc-5-glibc-2.23/65372bd
+# shellcheck disable=SC2148
+GCC_BASE=/mnt/gvfs/third-party2/gcc/8219ec1bcedf8ad9da05e121e193364de2cc4f61/5.x/centos6-native/c447969
+CLANG_BASE=/mnt/gvfs/third-party2/llvm-fb/64d8d58e3d84f8bde7a029763d4f5baf39d0d5b9/stable/centos6-native/6aaf4de
+LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/ba9be983c81de7299b59fe71950c664a84dcb5f8/5.x/gcc-5-glibc-2.23/339d858
+GLIBC_BASE=/mnt/gvfs/third-party2/glibc/f20197cf3d4bd50339c9777aaa0b2ccadad9e2cb/2.23/gcc-5-glibc-2.23/ca1d1c0
+SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/6427ce8c7496e4ab06c2da81543b94c0de8be3d0/1.1.3/gcc-5-glibc-2.23/9bc6787
+ZLIB_BASE=/mnt/gvfs/third-party2/zlib/8f1e8b867d26efef93eac2fabbdb2e1d512665d7/1.2.8/gcc-5-glibc-2.23/9bc6787
+BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/70471c0571559fe0af7db6d7e8860b93a7eadfe1/1.0.6/gcc-5-glibc-2.23/9bc6787
+LZ4_BASE=/mnt/gvfs/third-party2/lz4/453c89d6f0e68cdf1c151c769197fabedad9cac8/r131/gcc-5-glibc-2.23/9bc6787
+ZSTD_BASE=/mnt/gvfs/third-party2/zstd/00a40fa5f8bd2cd0622f2e868552793aef37ccf4/1.3.0/gcc-5-glibc-2.23/03859b5
+GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/47eef08f9acb77de982fbda6047c26d330739538/2.2.0/gcc-5-glibc-2.23/9bc6787
+JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/4414ddc78df8008b35cc4adac23590ad29148584/master/gcc-5-glibc-2.23/d506c82
+NUMA_BASE=/mnt/gvfs/third-party2/numa/9d7ae2693d05d62f9a579cb21e6b717cf257a75d/2.0.11/gcc-5-glibc-2.23/9bc6787
+LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/2b2dd58e3a52ccf2c1d827def59e5f740de0ad15/1.2/gcc-5-glibc-2.23/b443de1
+TBB_BASE=/mnt/gvfs/third-party2/tbb/379addf7ab2468a2b4293b47456cfcd1c9cb318d/4.3/gcc-5-glibc-2.23/9bc6787
+KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/3f68f5fe65a85b7c2d3e66852268fbd1efdb3151/4.0.9-36_fbk5_2933_gd092e3f/gcc-5-glibc-2.23/da39a3e
+BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/b9fab0aec99d9c36408e810b2677e91c12807afd/2.28/centos6-native/da39a3e
+VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/423431d61786b20bcc3bde8972901130cb29e6b3/3.11.0/gcc-5-glibc-2.23/9bc6787
+LUA_BASE=/mnt/gvfs/third-party2/lua/3b0bb3bd9a0f690a069c479fcc0f7424fc7456d2/5.2.3/gcc-5-glibc-2.23/65372bd
@@ -64,12 +64,8 @@ class MatchErrorParser(ErrorParserBase):

 class CompilerErrorParser(MatchErrorParser):
    def __init__(self):
-        # format (compile error):
-        #   '<filename>:<line #>:<column #>: error: <error msg>'
-        # format (link error):
-        #   '<filename>:<line #>: error: <error msg>'
-        # The below regex catches both
-        super(CompilerErrorParser, self).__init__(r'\S+:\d+: error:')
+        # format: '<filename>:<line #>:<column #>: error: <error msg>'
+        super(CompilerErrorParser, self).__init__(r'\S+:\d+:\d+: error:')


 class ScanBuildErrorParser(MatchErrorParser):
@@ -43,15 +43,11 @@ if test -z $PIC_BUILD; then
  LZ4_INCLUDE=" -I $LZ4_BASE/include/"
  LZ4_LIBS=" $LZ4_BASE/lib/liblz4.a"
  CFLAGS+=" -DLZ4"
-fi

-ZSTD_INCLUDE=" -I $ZSTD_BASE/include/"
-if test -z $PIC_BUILD; then
+  ZSTD_INCLUDE=" -I $ZSTD_BASE/include/"
  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd.a"
-else
-  ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd_pic.a"
+  CFLAGS+=" -DZSTD"
 fi
-CFLAGS+=" -DZSTD"

 # location of gflags headers and libraries
 GFLAGS_INCLUDE=" -I $GFLAGS_BASE/include/"
@@ -5082,8 +5082,8 @@ sub openoutputfiles {
 	# Set reading FD if using --group (--ungroup does not need)
 	for my $fdno (1,2) {
 	    # Re-open the file for reading
-	    # so fdw can be closed separately
-	    # and fdr can be seeked separately (for --line-buffer)
+	    # so fdw can be closed seperately
+	    # and fdr can be seeked seperately (for --line-buffer)
 	    open(my $fdr,"<", $self->fh($fdno,'name')) ||
 		::die_bug("fdr: Cannot open ".$self->fh($fdno,'name'));
 	    $self->set_fh($fdno,'r',$fdr);
@@ -377,7 +377,7 @@ function send_to_ods {
    echo >&2 "ERROR: Key $key doesn't have a value."
    return
  fi
-  curl --silent "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build$git_br&key=$key&value=$value" \
+  curl -s "https://www.intern.facebook.com/intern/agent/ods_set.php?entity=rocksdb_build$git_br&key=$key&value=$value" \
    --connect-timeout 60
 }

@@ -89,8 +89,6 @@ CLANG="USE_CLANG=1"
 # using platform007 gives us gcc-8 or higher which has that bug fixed.
 TSAN="ROCKSDB_FBCODE_BUILD_WITH_PLATFORM007=1 COMPILE_WITH_TSAN=1"
 UBSAN="COMPILE_WITH_UBSAN=1"
-TSAN_CRASH='CRASH_TEST_EXT_ARGS="--compression_type=zstd --log2_keys_per_lock=22"'
-NON_TSAN_CRASH="CRASH_TEST_EXT_ARGS=--compression_type=zstd"
 DISABLE_JEMALLOC="DISABLE_JEMALLOC=1"
 HTTP_PROXY="https_proxy=http://fwdproxy.29.prn1:8080 http_proxy=http://fwdproxy.29.prn1:8080 ftp_proxy=http://fwdproxy.29.prn1:8080"
 SETUP_JAVA_ENV="export $HTTP_PROXY; export JAVA_HOME=/usr/local/jdk-8u60-64/; export PATH=\$JAVA_HOME/bin:\$PATH"
@@ -346,7 +344,7 @@ LITE_BUILD_COMMANDS="[
            $CLEANUP_ENV,
            {
                'name':'Build RocksDB debug version',
-                'shell':'make J=1 LITE=1 all check || $CONTRUN_NAME=lite $TASK_CREATION_TOOL',
+                'shell':'$LITE make J=1 all check || $CONTRUN_NAME=lite $TASK_CREATION_TOOL',
                'user':'root',
                $PARSER
            },
@@ -383,14 +381,14 @@ 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 $TASK_CREATION_TOOL',
+                'shell':'$SHM $DEBUG 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 $TASK_CREATION_TOOL',
+                'shell':'$SHM $DEBUG make J=1 crash_test || $CONTRUN_NAME=crash_test $TASK_CREATION_TOOL',
                'user':'root',
                $PARSER
            }
@@ -455,7 +453,7 @@ ASAN_CRASH_TEST_COMMANDS="[
            {
                'name':'Build and run RocksDB debug asan_crash_test',
                'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 asan_crash_test || $CONTRUN_NAME=asan_crash_test $TASK_CREATION_TOOL',
+                'shell':'$SHM $DEBUG make J=1 asan_crash_test || $CONTRUN_NAME=asan_crash_test $TASK_CREATION_TOOL',
                'user':'root',
                $PARSER
            },
@@ -497,7 +495,7 @@ UBSAN_CRASH_TEST_COMMANDS="[
            {
                'name':'Build and run RocksDB debug ubsan_crash_test',
                'timeout': 86400,
-                'shell':'$SHM $DEBUG $NON_TSAN_CRASH make J=1 ubsan_crash_test || $CONTRUN_NAME=ubsan_crash_test $TASK_CREATION_TOOL',
+                'shell':'$SHM $DEBUG make J=1 ubsan_crash_test || $CONTRUN_NAME=ubsan_crash_test $TASK_CREATION_TOOL',
                'user':'root',
                $PARSER
            },
@@ -563,7 +561,7 @@ 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':'set -o pipefail && $SHM $DEBUG $TSAN CRASH_TEST_KILL_ODD=1887 CRASH_TEST_EXT_ARGS=--log2_keys_per_lock=22  make J=1 crash_test || $CONTRUN_NAME=tsan_crash_test $TASK_CREATION_TOOL',
                'user':'root',
                $PARSER
            },
@@ -646,7 +644,7 @@ run_regression()

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

@@ -664,12 +662,12 @@ run_regression()

  # === lite build ===
  make clean
-  make LITE=1 -j$(nproc) static_lib
+  OPT=-DROCKSDB_LITE make -j$(nproc) static_lib
  send_size_to_ods static_lib_lite $(stat --printf="%s" librocksdb.a)
  strip librocksdb.a
  send_size_to_ods static_lib_lite_stripped $(stat --printf="%s" librocksdb.a)

-  make LITE=1 -j$(nproc) shared_lib
+  OPT=-DROCKSDB_LITE make -j$(nproc) shared_lib
  send_size_to_ods shared_lib_lite $(stat --printf="%s" `readlink -f librocksdb.so`)
  strip `readlink -f librocksdb.so`
  send_size_to_ods shared_lib_lite_stripped $(stat --printf="%s" `readlink -f librocksdb.so`)
@@ -104,8 +104,8 @@ touch "$OUTPUT"
 echo "Writing dependencies to $OUTPUT"

 # Compilers locations
-GCC_BASE=`readlink -f $TP2_LATEST/gcc/5.x/centos7-native/*/`
-CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/stable/centos7-native/*/`
+GCC_BASE=`readlink -f $TP2_LATEST/gcc/5.x/centos6-native/*/`
+CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/stable/centos6-native/*/`

 log_variable GCC_BASE
 log_variable CLANG_BASE
@@ -125,7 +125,7 @@ get_lib_base libunwind  LATEST  gcc-5-glibc-2.23
 get_lib_base tbb        LATEST  gcc-5-glibc-2.23

 get_lib_base kernel-headers 4.0.9-36_fbk5_2933_gd092e3f gcc-5-glibc-2.23
-get_lib_base binutils   LATEST centos7-native
+get_lib_base binutils   LATEST centos6-native 
 get_lib_base valgrind   LATEST gcc-5-glibc-2.23
 get_lib_base lua        5.2.3 gcc-5-glibc-2.23

@@ -64,8 +64,8 @@ class CacheTest : public testing::TestWithParam<std::string> {

  std::vector<int> deleted_keys_;
  std::vector<int> deleted_values_;
-  std::shared_ptr<Cache> cache_;
-  std::shared_ptr<Cache> cache2_;
+  shared_ptr<Cache> cache_;
+  shared_ptr<Cache> cache2_;

  CacheTest()
      : cache_(NewCache(kCacheSize, kNumShardBits, false)),
@@ -145,7 +145,7 @@ class CacheTest : public testing::TestWithParam<std::string> {
 CacheTest* CacheTest::current_;

 TEST_P(CacheTest, UsageTest) {
-  // cache is std::shared_ptr and will be automatically cleaned up.
+  // cache is shared_ptr and will be automatically cleaned up.
  const uint64_t kCapacity = 100000;
  auto cache = NewCache(kCapacity, 8, false);

@@ -173,7 +173,7 @@ TEST_P(CacheTest, UsageTest) {
 }

 TEST_P(CacheTest, PinnedUsageTest) {
-  // cache is std::shared_ptr and will be automatically cleaned up.
+  // cache is shared_ptr and will be automatically cleaned up.
  const uint64_t kCapacity = 100000;
  auto cache = NewCache(kCapacity, 8, false);

@@ -461,10 +461,8 @@ std::string LRUCacheShard::GetPrintableOptions() const {
 }

 LRUCache::LRUCache(size_t capacity, int num_shard_bits,
-                   bool strict_capacity_limit, double high_pri_pool_ratio,
-                   std::shared_ptr<MemoryAllocator> allocator)
-    : ShardedCache(capacity, num_shard_bits, strict_capacity_limit,
-                   std::move(allocator)) {
+                   bool strict_capacity_limit, double high_pri_pool_ratio)
+    : ShardedCache(capacity, num_shard_bits, strict_capacity_limit) {
  num_shards_ = 1 << num_shard_bits;
  shards_ = reinterpret_cast<LRUCacheShard*>(
      port::cacheline_aligned_alloc(sizeof(LRUCacheShard) * num_shards_));
@@ -539,14 +537,12 @@ double LRUCache::GetHighPriPoolRatio() {
 std::shared_ptr<Cache> NewLRUCache(const LRUCacheOptions& cache_opts) {
  return NewLRUCache(cache_opts.capacity, cache_opts.num_shard_bits,
                     cache_opts.strict_capacity_limit,
-                     cache_opts.high_pri_pool_ratio,
-                     cache_opts.memory_allocator);
+                     cache_opts.high_pri_pool_ratio);
 }

-std::shared_ptr<Cache> NewLRUCache(
-    size_t capacity, int num_shard_bits, bool strict_capacity_limit,
-    double high_pri_pool_ratio,
-    std::shared_ptr<MemoryAllocator> memory_allocator) {
+std::shared_ptr<Cache> NewLRUCache(size_t capacity, int num_shard_bits,
+                                   bool strict_capacity_limit,
+                                   double high_pri_pool_ratio) {
  if (num_shard_bits >= 20) {
    return nullptr;  // the cache cannot be sharded into too many fine pieces
  }
@@ -558,8 +554,7 @@ std::shared_ptr<Cache> NewLRUCache(
    num_shard_bits = GetDefaultCacheShardBits(capacity);
  }
  return std::make_shared<LRUCache>(capacity, num_shard_bits,
-                                    strict_capacity_limit, high_pri_pool_ratio,
-                                    std::move(memory_allocator));
+                                    strict_capacity_limit, high_pri_pool_ratio);
 }

 }  // namespace rocksdb
@@ -279,8 +279,7 @@ class ALIGN_AS(CACHE_LINE_SIZE) LRUCacheShard : public CacheShard {
 class LRUCache : public ShardedCache {
 public:
  LRUCache(size_t capacity, int num_shard_bits, bool strict_capacity_limit,
-           double high_pri_pool_ratio,
-           std::shared_ptr<MemoryAllocator> memory_allocator = nullptr);
+           double high_pri_pool_ratio);
  virtual ~LRUCache();
  virtual const char* Name() const override { return "LRUCache"; }
  virtual CacheShard* GetShard(int shard) override;
@@ -20,10 +20,8 @@
 namespace rocksdb {

 ShardedCache::ShardedCache(size_t capacity, int num_shard_bits,
-                           bool strict_capacity_limit,
-                           std::shared_ptr<MemoryAllocator> allocator)
-    : Cache(std::move(allocator)),
-      num_shard_bits_(num_shard_bits),
+                           bool strict_capacity_limit)
+    : num_shard_bits_(num_shard_bits),
      capacity_(capacity),
      strict_capacity_limit_(strict_capacity_limit),
      last_id_(1) {}
@@ -144,9 +142,6 @@ std::string ShardedCache::GetPrintableOptions() const {
             strict_capacity_limit_);
    ret.append(buffer);
  }
-  snprintf(buffer, kBufferSize, "    memory_allocator : %s\n",
-           memory_allocator() ? memory_allocator()->Name() : "None");
-  ret.append(buffer);
  ret.append(GetShard(0)->GetPrintableOptions());
  return ret;
 }
@@ -47,8 +47,7 @@ class CacheShard {
 // Keys are sharded by the highest num_shard_bits bits of hash value.
 class ShardedCache : public Cache {
 public:
-  ShardedCache(size_t capacity, int num_shard_bits, bool strict_capacity_limit,
-               std::shared_ptr<MemoryAllocator> memory_allocator = nullptr);
+  ShardedCache(size_t capacity, int num_shard_bits, bool strict_capacity_limit);
  virtual ~ShardedCache() = default;
  virtual const char* Name() const override = 0;
  virtual CacheShard* GetShard(int shard) = 0;
@@ -18,7 +18,6 @@
 #include "db/event_helpers.h"
 #include "db/internal_stats.h"
 #include "db/merge_helper.h"
-#include "db/range_del_aggregator.h"
 #include "db/table_cache.h"
 #include "db/version_edit.h"
 #include "monitoring/iostats_context_imp.h"
@@ -66,9 +65,8 @@ Status BuildTable(
    const std::string& dbname, Env* env, const ImmutableCFOptions& ioptions,
    const MutableCFOptions& mutable_cf_options, const EnvOptions& env_options,
    TableCache* table_cache, InternalIterator* iter,
-    std::vector<std::unique_ptr<FragmentedRangeTombstoneIterator>>
-        range_del_iters,
-    FileMetaData* meta, const InternalKeyComparator& internal_comparator,
+    std::unique_ptr<InternalIterator> range_del_iter, FileMetaData* meta,
+    const InternalKeyComparator& internal_comparator,
    const std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
        int_tbl_prop_collector_factories,
    uint32_t column_family_id, const std::string& column_family_name,
@@ -88,10 +86,12 @@ Status BuildTable(
  Status s;
  meta->fd.file_size = 0;
  iter->SeekToFirst();
-  std::unique_ptr<CompactionRangeDelAggregator> range_del_agg(
-      new CompactionRangeDelAggregator(&internal_comparator, snapshots));
-  for (auto& range_del_iter : range_del_iters) {
-    range_del_agg->AddTombstones(std::move(range_del_iter));
+  std::unique_ptr<RangeDelAggregator> range_del_agg(
+      new RangeDelAggregator(internal_comparator, snapshots));
+  s = range_del_agg->AddTombstones(std::move(range_del_iter));
+  if (!s.ok()) {
+    // may be non-ok if a range tombstone key is unparsable
+    return s;
  }

  std::string fname = TableFileName(ioptions.cf_paths, meta->fd.GetNumber(),
@@ -104,9 +104,9 @@ Status BuildTable(

  if (iter->Valid() || !range_del_agg->IsEmpty()) {
    TableBuilder* builder;
-    std::unique_ptr<WritableFileWriter> file_writer;
+    unique_ptr<WritableFileWriter> file_writer;
    {
-      std::unique_ptr<WritableFile> file;
+      unique_ptr<WritableFile> file;
 #ifndef NDEBUG
      bool use_direct_writes = env_options.use_direct_writes;
      TEST_SYNC_POINT_CALLBACK("BuildTable:create_file", &use_direct_writes);
@@ -121,9 +121,8 @@ Status BuildTable(
      file->SetIOPriority(io_priority);
      file->SetWriteLifeTimeHint(write_hint);

-      file_writer.reset(new WritableFileWriter(std::move(file), fname,
-                                               env_options, ioptions.statistics,
-                                               ioptions.listeners));
+      file_writer.reset(new WritableFileWriter(std::move(file), env_options,
+                                               ioptions.statistics));
      builder = NewTableBuilder(
          ioptions, mutable_cf_options, internal_comparator,
          int_tbl_prop_collector_factories, column_family_id,
@@ -158,10 +157,8 @@ Status BuildTable(
      }
    }

-    auto range_del_it = range_del_agg->NewIterator();
-    for (range_del_it->SeekToFirst(); range_del_it->Valid();
-         range_del_it->Next()) {
-      auto tombstone = range_del_it->Tombstone();
+    for (auto it = range_del_agg->NewIterator(); it->Valid(); it->Next()) {
+      auto tombstone = it->Tombstone();
      auto kv = tombstone.Serialize();
      builder->Add(kv.first.Encode(), kv.second);
      meta->UpdateBoundariesForRange(kv.first, tombstone.SerializeEndKey(),
@@ -9,7 +9,6 @@
 #include <string>
 #include <utility>
 #include <vector>
-#include "db/range_tombstone_fragmenter.h"
 #include "db/table_properties_collector.h"
 #include "options/cf_options.h"
 #include "rocksdb/comparator.h"
@@ -36,6 +35,7 @@ class VersionEdit;
 class TableBuilder;
 class WritableFileWriter;
 class InternalStats;
+class InternalIterator;

 // @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
@@ -66,9 +66,8 @@ extern Status BuildTable(
    const std::string& dbname, Env* env, const ImmutableCFOptions& options,
    const MutableCFOptions& mutable_cf_options, const EnvOptions& env_options,
    TableCache* table_cache, InternalIterator* iter,
-    std::vector<std::unique_ptr<FragmentedRangeTombstoneIterator>>
-        range_del_iters,
-    FileMetaData* meta, const InternalKeyComparator& internal_comparator,
+    std::unique_ptr<InternalIterator> range_del_iter, FileMetaData* meta,
+    const InternalKeyComparator& internal_comparator,
    const std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
        int_tbl_prop_collector_factories,
    uint32_t column_family_id, const std::string& column_family_name,
@@ -33,7 +33,6 @@
 #include "rocksdb/utilities/backupable_db.h"
 #include "rocksdb/utilities/checkpoint.h"
 #include "rocksdb/utilities/db_ttl.h"
-#include "rocksdb/utilities/memory_util.h"
 #include "rocksdb/utilities/optimistic_transaction_db.h"
 #include "rocksdb/utilities/transaction.h"
 #include "rocksdb/utilities/transaction_db.h"
@@ -42,10 +41,6 @@
 #include "rocksdb/perf_context.h"
 #include "utilities/merge_operators.h"

-#include <vector>
-#include <unordered_set>
-#include <map>
-
 using rocksdb::BytewiseComparator;
 using rocksdb::Cache;
 using rocksdb::ColumnFamilyDescriptor;
@@ -113,12 +108,8 @@ using rocksdb::TransactionLogIterator;
 using rocksdb::BatchResult;
 using rocksdb::PerfLevel;
 using rocksdb::PerfContext;
-using rocksdb::MemoryUtil;

 using std::shared_ptr;
-using std::vector;
-using std::unordered_set;
-using std::map;

 extern "C" {

@@ -151,20 +142,14 @@ struct rocksdb_writablefile_t    { WritableFile*     rep; };
 struct rocksdb_wal_iterator_t { TransactionLogIterator* rep; };
 struct rocksdb_wal_readoptions_t { TransactionLogIterator::ReadOptions rep; };
 struct rocksdb_filelock_t        { FileLock*         rep; };
-struct rocksdb_logger_t {
-  std::shared_ptr<Logger> rep;
-};
-struct rocksdb_cache_t {
-  std::shared_ptr<Cache> rep;
-};
+struct rocksdb_logger_t          { shared_ptr<Logger>  rep; };
+struct rocksdb_cache_t           { shared_ptr<Cache>   rep; };
 struct rocksdb_livefiles_t       { std::vector<LiveFileMetaData> rep; };
 struct rocksdb_column_family_handle_t  { ColumnFamilyHandle* rep; };
 struct rocksdb_envoptions_t      { EnvOptions        rep; };
 struct rocksdb_ingestexternalfileoptions_t  { IngestExternalFileOptions rep; };
 struct rocksdb_sstfilewriter_t   { SstFileWriter*    rep; };
-struct rocksdb_ratelimiter_t {
-  std::shared_ptr<RateLimiter> rep;
-};
+struct rocksdb_ratelimiter_t     { shared_ptr<RateLimiter>      rep; };
 struct rocksdb_perfcontext_t     { PerfContext*      rep; };
 struct rocksdb_pinnableslice_t {
  PinnableSlice rep;
@@ -2417,7 +2402,7 @@ void rocksdb_options_set_bytes_per_sync(

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

 void rocksdb_options_set_allow_concurrent_memtable_write(rocksdb_options_t* opt,
@@ -2448,20 +2433,6 @@ void rocksdb_options_set_max_write_buffer_number_to_maintain(
  opt->rep.max_write_buffer_number_to_maintain = n;
 }

-void rocksdb_options_set_enable_pipelined_write(rocksdb_options_t* opt,
-                                                unsigned char v) {
-  opt->rep.enable_pipelined_write = v;
-}
-
-void rocksdb_options_set_max_subcompactions(rocksdb_options_t* opt,
-                                            uint32_t n) {
-  opt->rep.max_subcompactions = n;
-}
-
-void rocksdb_options_set_max_background_jobs(rocksdb_options_t* opt, int n) {
-  opt->rep.max_background_jobs = n;
-}
-
 void rocksdb_options_set_max_background_compactions(rocksdb_options_t* opt, int n) {
  opt->rep.max_background_compactions = n;
 }
@@ -3550,18 +3521,6 @@ const char* rocksdb_livefiles_largestkey(
  return lf->rep[index].largestkey.data();
 }

-uint64_t rocksdb_livefiles_entries(
-    const rocksdb_livefiles_t* lf,
-    int index) {
-  return lf->rep[index].num_entries;
-}
-
-uint64_t rocksdb_livefiles_deletions(
-    const rocksdb_livefiles_t* lf,
-    int index) {
-  return lf->rep[index].num_deletions;
-}
-
 extern void rocksdb_livefiles_destroy(
  const rocksdb_livefiles_t* lf) {
  delete lf;
@@ -4128,98 +4087,6 @@ const char* rocksdb_pinnableslice_value(const rocksdb_pinnableslice_t* v,
  *vlen = v->rep.size();
  return v->rep.data();
 }
-
-// container to keep databases and caches in order to use rocksdb::MemoryUtil
-struct rocksdb_memory_consumers_t {
-  std::vector<rocksdb_t*> dbs;
-  std::unordered_set<rocksdb_cache_t*> caches;
-};
-
-// initializes new container of memory consumers
-rocksdb_memory_consumers_t* rocksdb_memory_consumers_create() {
-  return new rocksdb_memory_consumers_t;
-}
-
-// adds datatabase to the container of memory consumers
-void rocksdb_memory_consumers_add_db(rocksdb_memory_consumers_t* consumers,
-                                     rocksdb_t* db) {
-  consumers->dbs.push_back(db);
-}
-
-// adds cache to the container of memory consumers
-void rocksdb_memory_consumers_add_cache(rocksdb_memory_consumers_t* consumers,
-                                        rocksdb_cache_t* cache) {
-  consumers->caches.insert(cache);
-}
-
-// deletes container with memory consumers
-void rocksdb_memory_consumers_destroy(rocksdb_memory_consumers_t* consumers) {
-  delete consumers;
-}
-
-// contains memory usage statistics provided by rocksdb::MemoryUtil
-struct rocksdb_memory_usage_t {
-  uint64_t mem_table_total;
-  uint64_t mem_table_unflushed;
-  uint64_t mem_table_readers_total;
-  uint64_t cache_total;
-};
-
-// estimates amount of memory occupied by consumers (dbs and caches)
-rocksdb_memory_usage_t* rocksdb_approximate_memory_usage_create(
-    rocksdb_memory_consumers_t* consumers, char** errptr) {
-
-  vector<DB*> dbs;
-  for (auto db : consumers->dbs) {
-    dbs.push_back(db->rep);
-  }
-
-  unordered_set<const Cache*> cache_set;
-  for (auto cache : consumers->caches) {
-    cache_set.insert(const_cast<const Cache*>(cache->rep.get()));
-  }
-
-  std::map<rocksdb::MemoryUtil::UsageType, uint64_t> usage_by_type;
-
-  auto status = MemoryUtil::GetApproximateMemoryUsageByType(dbs, cache_set,
-                                                            &usage_by_type);
-  if (SaveError(errptr, status)) {
-    return nullptr;
-  }
-
-  auto result = new rocksdb_memory_usage_t;
-  result->mem_table_total = usage_by_type[MemoryUtil::kMemTableTotal];
-  result->mem_table_unflushed = usage_by_type[MemoryUtil::kMemTableUnFlushed];
-  result->mem_table_readers_total = usage_by_type[MemoryUtil::kTableReadersTotal];
-  result->cache_total = usage_by_type[MemoryUtil::kCacheTotal];
-  return result;
-}
-
-uint64_t rocksdb_approximate_memory_usage_get_mem_table_total(
-    rocksdb_memory_usage_t* memory_usage) {
-  return memory_usage->mem_table_total;
-}
-
-uint64_t rocksdb_approximate_memory_usage_get_mem_table_unflushed(
-    rocksdb_memory_usage_t* memory_usage) {
-  return memory_usage->mem_table_unflushed;
-}
-
-uint64_t rocksdb_approximate_memory_usage_get_mem_table_readers_total(
-    rocksdb_memory_usage_t* memory_usage) {
-  return memory_usage->mem_table_readers_total;
-}
-
-uint64_t rocksdb_approximate_memory_usage_get_cache_total(
-    rocksdb_memory_usage_t* memory_usage) {
-  return memory_usage->cache_total;
-}
-
-// deletes container with memory usage estimates
-void rocksdb_approximate_memory_usage_destroy(rocksdb_memory_usage_t* usage) {
-  delete usage;
-}
-
 }  // end extern "C"

 #endif  // !ROCKSDB_LITE
@@ -19,8 +19,11 @@

 // Can not use port/port.h macros as this is a c file
 #ifdef OS_WIN
+
 #include <windows.h>

+#define snprintf _snprintf
+
 // Ok for uniqueness
 int geteuid() {
  int result = 0;
@@ -31,11 +34,6 @@ int geteuid() {
  return result;
 }

-// VS < 2015
-#if defined(_MSC_VER) && (_MSC_VER < 1900)
-#define snprintf _snprintf
-#endif
-
 #endif

 const char* phase = "";
@@ -49,19 +47,12 @@ static void StartPhase(const char* name) {
  fprintf(stderr, "=== Test %s\n", name);
  phase = name;
 }
-#ifdef _MSC_VER
-#pragma warning(push)
-#pragma warning (disable: 4996) // getenv security warning
-#endif
 static const char* GetTempDir(void) {
    const char* ret = getenv("TEST_TMPDIR");
    if (ret == NULL || ret[0] == '\0')
        ret = "/tmp";
    return ret;
 }
-#ifdef _MSC_VER
-#pragma warning(pop)
-#endif

 #define CheckNoError(err)                                               \
  if ((err) != NULL) {                                                  \
@@ -652,7 +643,7 @@ int main(int argc, char** argv) {
    rocksdb_sstfilewriter_t* writer =
        rocksdb_sstfilewriter_create(env_opt, io_options);

-    remove(sstfilename);
+    unlink(sstfilename);
    rocksdb_sstfilewriter_open(writer, sstfilename, &err);
    CheckNoError(err);
    rocksdb_sstfilewriter_put(writer, "sstk1", 5, "v1", 2, &err);
@@ -673,7 +664,7 @@ int main(int argc, char** argv) {
    CheckGet(db, roptions, "sstk2", "v2");
    CheckGet(db, roptions, "sstk3", "v3");

-    remove(sstfilename);
+    unlink(sstfilename);
    rocksdb_sstfilewriter_open(writer, sstfilename, &err);
    CheckNoError(err);
    rocksdb_sstfilewriter_put(writer, "sstk2", 5, "v4", 2, &err);
@@ -1343,47 +1334,6 @@ int main(int argc, char** argv) {
    rocksdb_destroy_db(options, dbname, &err);
  }

-  // Check memory usage stats
-  StartPhase("approximate_memory_usage");
-  {
-    // Create database
-    db = rocksdb_open(options, dbname, &err);
-    CheckNoError(err);
-
-    rocksdb_memory_consumers_t* consumers;
-    consumers = rocksdb_memory_consumers_create();
-    rocksdb_memory_consumers_add_db(consumers, db);
-    rocksdb_memory_consumers_add_cache(consumers, cache);
-
-    // take memory usage report before write-read operation
-    rocksdb_memory_usage_t* mu1;
-    mu1 = rocksdb_approximate_memory_usage_create(consumers, &err);
-    CheckNoError(err);
-
-    // Put data (this should affect memtables)
-    rocksdb_put(db, woptions, "memory", 6, "test", 4, &err);
-    CheckNoError(err);
-    CheckGet(db, roptions, "memory", "test");
-
-    // take memory usage report after write-read operation
-    rocksdb_memory_usage_t* mu2;
-    mu2 = rocksdb_approximate_memory_usage_create(consumers, &err);
-    CheckNoError(err);
-
-    // amount of memory used within memtables should grow
-    CheckCondition(rocksdb_approximate_memory_usage_get_mem_table_total(mu2) >=
-                   rocksdb_approximate_memory_usage_get_mem_table_total(mu1));
-    CheckCondition(rocksdb_approximate_memory_usage_get_mem_table_unflushed(mu2) >=
-                   rocksdb_approximate_memory_usage_get_mem_table_unflushed(mu1));
-
-    rocksdb_memory_consumers_destroy(consumers);
-    rocksdb_approximate_memory_usage_destroy(mu1);
-    rocksdb_approximate_memory_usage_destroy(mu2);
-    rocksdb_close(db);
-    rocksdb_destroy_db(options, dbname, &err);
-    CheckNoError(err);
-  }
-
  StartPhase("cuckoo_options");
  {
    rocksdb_cuckoo_table_options_t* cuckoo_options;
@@ -1725,7 +1675,7 @@ int main(int argc, char** argv) {
    db = rocksdb_open(options, dbname, &err);
    CheckNoError(err);
  }
-
+  
  StartPhase("cleanup");
  rocksdb_close(db);
  rocksdb_options_destroy(options);
@@ -20,12 +20,10 @@
 #include <limits>

 #include "db/compaction_picker.h"
-#include "db/compaction_picker_fifo.h"
 #include "db/compaction_picker_universal.h"
 #include "db/db_impl.h"
 #include "db/internal_stats.h"
 #include "db/job_context.h"
-#include "db/range_del_aggregator.h"
 #include "db/table_properties_collector.h"
 #include "db/version_set.h"
 #include "db/write_controller.h"
@@ -107,6 +105,9 @@ void GetIntTblPropCollectorFactory(
    int_tbl_prop_collector_factories->emplace_back(
        new UserKeyTablePropertiesCollectorFactory(collector_factories[i]));
  }
+  // Add collector to collect internal key statistics
+  int_tbl_prop_collector_factories->emplace_back(
+      new InternalKeyPropertiesCollectorFactory);
 }

 Status CheckCompressionSupported(const ColumnFamilyOptions& cf_options) {
@@ -944,16 +945,26 @@ Status ColumnFamilyData::RangesOverlapWithMemtables(
  super_version->imm->AddIterators(read_opts, &merge_iter_builder);
  ScopedArenaIterator memtable_iter(merge_iter_builder.Finish());

-  auto read_seq = super_version->current->version_set()->LastSequence();
-  ReadRangeDelAggregator range_del_agg(&internal_comparator_, read_seq);
+  std::vector<InternalIterator*> memtable_range_del_iters;
  auto* active_range_del_iter =
-      super_version->mem->NewRangeTombstoneIterator(read_opts, read_seq);
-  range_del_agg.AddTombstones(
-      std::unique_ptr<FragmentedRangeTombstoneIterator>(active_range_del_iter));
-  super_version->imm->AddRangeTombstoneIterators(read_opts, nullptr /* arena */,
-                                                 &range_del_agg);
-
+      super_version->mem->NewRangeTombstoneIterator(read_opts);
+  if (active_range_del_iter != nullptr) {
+    memtable_range_del_iters.push_back(active_range_del_iter);
+  }
+  super_version->imm->AddRangeTombstoneIterators(read_opts,
+                                                 &memtable_range_del_iters);
+  RangeDelAggregator range_del_agg(internal_comparator_, {} /* snapshots */,
+                                   false /* collapse_deletions */);
  Status status;
+  {
+    std::unique_ptr<InternalIterator> memtable_range_del_iter(
+        NewMergingIterator(&internal_comparator_,
+                           memtable_range_del_iters.empty()
+                               ? nullptr
+                               : &memtable_range_del_iters[0],
+                           static_cast<int>(memtable_range_del_iters.size())));
+    status = range_del_agg.AddTombstones(std::move(memtable_range_del_iter));
+  }
  for (size_t i = 0; i < ranges.size() && status.ok() && !*overlap; ++i) {
    auto* vstorage = super_version->current->storage_info();
    auto* ucmp = vstorage->InternalComparator()->user_comparator();
@@ -47,7 +47,7 @@ class EnvCounter : public EnvWrapper {
  int GetNumberOfNewWritableFileCalls() {
    return num_new_writable_file_;
  }
-  Status NewWritableFile(const std::string& f, std::unique_ptr<WritableFile>* r,
+  Status NewWritableFile(const std::string& f, unique_ptr<WritableFile>* r,
                         const EnvOptions& soptions) override {
    ++num_new_writable_file_;
    return EnvWrapper::NewWritableFile(f, r, soptions);
@@ -486,9 +486,9 @@ class ColumnFamilyTestBase : public testing::Test {
  void CopyFile(const std::string& source, const std::string& destination,
                uint64_t size = 0) {
    const EnvOptions soptions;
-    std::unique_ptr<SequentialFile> srcfile;
+    unique_ptr<SequentialFile> srcfile;
    ASSERT_OK(env_->NewSequentialFile(source, &srcfile, soptions));
-    std::unique_ptr<WritableFile> destfile;
+    unique_ptr<WritableFile> destfile;
    ASSERT_OK(env_->NewWritableFile(destination, &destfile, soptions));

    if (size == 0) {
@@ -513,16 +513,6 @@ class ColumnFamilyTestBase : public testing::Test {
    return static_cast<int>(files.size());
  }

-  void RecalculateWriteStallConditions(ColumnFamilyData* cfd,
-      const MutableCFOptions& mutable_cf_options)  {
-    // add lock to avoid race condition between
-    // `RecalculateWriteStallConditions` which writes to CFStats and
-    // background `DBImpl::DumpStats()` threads which read CFStats
-    dbfull()->TEST_LockMutex();
-    cfd->RecalculateWriteStallConditions(mutable_cf_options);
-    dbfull()-> TEST_UnlockMutex();
-  }
-
  std::vector<ColumnFamilyHandle*> handles_;
  std::vector<std::string> names_;
  std::vector<std::set<std::string>> keys_;
@@ -2510,139 +2500,139 @@ TEST_P(ColumnFamilyTest, WriteStallSingleColumnFamily) {
  mutable_cf_options.disable_auto_compactions = false;

  vstorage->TEST_set_estimated_compaction_needed_bytes(50);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage->TEST_set_estimated_compaction_needed_bytes(201);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(400);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(500);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25 / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(450);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(205);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(202);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(201);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(198);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage->TEST_set_estimated_compaction_needed_bytes(399);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(599);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(2001);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(3001);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage->TEST_set_estimated_compaction_needed_bytes(390);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(100);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage->set_l0_delay_trigger_count(100);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(101);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());

  vstorage->set_l0_delay_trigger_count(0);
  vstorage->TEST_set_estimated_compaction_needed_bytes(300);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25 / 1.25, GetDbDelayedWriteRate());

  vstorage->set_l0_delay_trigger_count(101);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25 / 1.25 / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(200);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25 / 1.25, GetDbDelayedWriteRate());

  vstorage->set_l0_delay_trigger_count(0);
  vstorage->TEST_set_estimated_compaction_needed_bytes(0);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  mutable_cf_options.disable_auto_compactions = true;
  dbfull()->TEST_write_controler().set_delayed_write_rate(kBaseRate);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage->set_l0_delay_trigger_count(50);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(0, GetDbDelayedWriteRate());
@@ -2650,7 +2640,7 @@ TEST_P(ColumnFamilyTest, WriteStallSingleColumnFamily) {

  vstorage->set_l0_delay_trigger_count(60);
  vstorage->TEST_set_estimated_compaction_needed_bytes(300);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(0, GetDbDelayedWriteRate());
@@ -2659,14 +2649,14 @@ TEST_P(ColumnFamilyTest, WriteStallSingleColumnFamily) {
  mutable_cf_options.disable_auto_compactions = false;
  vstorage->set_l0_delay_trigger_count(70);
  vstorage->TEST_set_estimated_compaction_needed_bytes(500);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage->set_l0_delay_trigger_count(71);
  vstorage->TEST_set_estimated_compaction_needed_bytes(501);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());
@@ -2691,31 +2681,31 @@ TEST_P(ColumnFamilyTest, CompactionSpeedupSingleColumnFamily) {
  mutable_cf_options.hard_pending_compaction_bytes_limit = 2000;

  vstorage->TEST_set_estimated_compaction_needed_bytes(40);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(50);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(300);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(45);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(7);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(9);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(6);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  // Speed up threshold = min(4 * 2, 4 + (12 - 4)/4) = 6
@@ -2724,15 +2714,15 @@ TEST_P(ColumnFamilyTest, CompactionSpeedupSingleColumnFamily) {
  mutable_cf_options.level0_stop_writes_trigger = 30;

  vstorage->set_l0_delay_trigger_count(5);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(7);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(3);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());
 }

@@ -2759,53 +2749,53 @@ TEST_P(ColumnFamilyTest, WriteStallTwoColumnFamilies) {
  mutable_cf_options1.soft_pending_compaction_bytes_limit = 500;

  vstorage->TEST_set_estimated_compaction_needed_bytes(50);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(201);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(!dbfull()->TEST_write_controler().NeedsDelay());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(600);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(70);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate, GetDbDelayedWriteRate());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(800);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(300);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25 / 1.25, GetDbDelayedWriteRate());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(700);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());

  vstorage->TEST_set_estimated_compaction_needed_bytes(500);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25 / 1.25, GetDbDelayedWriteRate());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(600);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_TRUE(!IsDbWriteStopped());
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_EQ(kBaseRate / 1.25, GetDbDelayedWriteRate());
@@ -2838,41 +2828,41 @@ TEST_P(ColumnFamilyTest, CompactionSpeedupTwoColumnFamilies) {
  mutable_cf_options1.level0_slowdown_writes_trigger = 16;

  vstorage->TEST_set_estimated_compaction_needed_bytes(40);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(60);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(30);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(70);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->TEST_set_estimated_compaction_needed_bytes(20);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage1->TEST_set_estimated_compaction_needed_bytes(3);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(9);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage1->set_l0_delay_trigger_count(2);
-  RecalculateWriteStallConditions(cfd1, mutable_cf_options);
+  cfd1->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(6, dbfull()->TEST_BGCompactionsAllowed());

  vstorage->set_l0_delay_trigger_count(0);
-  RecalculateWriteStallConditions(cfd, mutable_cf_options);
+  cfd->RecalculateWriteStallConditions(mutable_cf_options);
  ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());
 }

@@ -308,55 +308,6 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
  delete db;
 }

-TEST_F(CompactFilesTest, SentinelCompressionType) {
-  if (!Zlib_Supported()) {
-    fprintf(stderr, "zlib compression not supported, skip this test\n");
-    return;
-  }
-  if (!Snappy_Supported()) {
-    fprintf(stderr, "snappy compression not supported, skip this test\n");
-    return;
-  }
-  // Check that passing `CompressionType::kDisableCompressionOption` to
-  // `CompactFiles` causes it to use the column family compression options.
-  for (auto compaction_style :
-       {CompactionStyle::kCompactionStyleLevel,
-        CompactionStyle::kCompactionStyleUniversal,
-        CompactionStyle::kCompactionStyleNone}) {
-    DestroyDB(db_name_, Options());
-    Options options;
-    options.compaction_style = compaction_style;
-    // L0: Snappy, L1: ZSTD, L2: Snappy
-    options.compression_per_level = {CompressionType::kSnappyCompression,
-                                     CompressionType::kZlibCompression,
-                                     CompressionType::kSnappyCompression};
-    options.create_if_missing = true;
-    FlushedFileCollector* collector = new FlushedFileCollector();
-    options.listeners.emplace_back(collector);
-    DB* db = nullptr;
-    ASSERT_OK(DB::Open(options, db_name_, &db));
-
-    db->Put(WriteOptions(), "key", "val");
-    db->Flush(FlushOptions());
-
-    auto l0_files = collector->GetFlushedFiles();
-    ASSERT_EQ(1, l0_files.size());
-
-    // L0->L1 compaction, so output should be ZSTD-compressed
-    CompactionOptions compaction_opts;
-    compaction_opts.compression = CompressionType::kDisableCompressionOption;
-    ASSERT_OK(db->CompactFiles(compaction_opts, l0_files, 1));
-
-    rocksdb::TablePropertiesCollection all_tables_props;
-    ASSERT_OK(db->GetPropertiesOfAllTables(&all_tables_props));
-    for (const auto& name_and_table_props : all_tables_props) {
-      ASSERT_EQ(CompressionTypeToString(CompressionType::kZlibCompression),
-                name_and_table_props.second->compression_name);
-    }
-    delete db;
-  }
-}
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -25,12 +25,22 @@ CompactedDBImpl::~CompactedDBImpl() {
 }

 size_t CompactedDBImpl::FindFile(const Slice& key) {
+  size_t left = 0;
  size_t right = files_.num_files - 1;
-  auto cmp = [&](const FdWithKeyRange& f, const Slice& k) -> bool {
-    return user_comparator_->Compare(ExtractUserKey(f.largest_key), k) < 0;
-  };
-  return static_cast<size_t>(std::lower_bound(files_.files,
-                            files_.files + right, key, cmp) - files_.files);
+  while (left < right) {
+    size_t mid = (left + right) >> 1;
+    const FdWithKeyRange& f = files_.files[mid];
+    if (user_comparator_->Compare(ExtractUserKey(f.largest_key), key) < 0) {
+      // Key at "mid.largest" is < "target".  Therefore all
+      // files at or before "mid" are uninteresting.
+      left = mid + 1;
+    } else {
+      // Key at "mid.largest" is >= "target".  Therefore all files
+      // after "mid" are uninteresting.
+      right = mid;
+    }
+  }
+  return right;
 }

 Status CompactedDBImpl::Get(const ReadOptions& options, ColumnFamilyHandle*,
@@ -147,7 +157,6 @@ Status CompactedDBImpl::Open(const Options& options,
  std::unique_ptr<CompactedDBImpl> db(new CompactedDBImpl(db_options, dbname));
  Status s = db->Init(options);
  if (s.ok()) {
-    db->StartTimedTasks();
    ROCKS_LOG_INFO(db->immutable_db_options_.info_log,
                   "Opened the db as fully compacted mode");
    LogFlush(db->immutable_db_options_.info_log);
@@ -67,11 +67,10 @@ class CompactedDBImpl : public DBImpl {
  virtual Status EnableFileDeletions(bool /*force*/) override {
    return Status::NotSupported("Not supported in compacted db mode.");
  }
-  virtual Status GetLiveFiles(std::vector<std::string>& ret,
-                              uint64_t* manifest_file_size,
-                              bool /*flush_memtable*/) override {
-    return DBImpl::GetLiveFiles(ret, manifest_file_size,
-                                false /* flush_memtable */);
+  virtual Status GetLiveFiles(std::vector<std::string>&,
+                              uint64_t* /*manifest_file_size*/,
+                              bool /*flush_memtable*/ = true) override {
+    return Status::NotSupported("Not supported in compacted db mode.");
  }
  using DBImpl::Flush;
  virtual Status Flush(const FlushOptions& /*options*/,
@@ -23,43 +23,6 @@

 namespace rocksdb {

-const uint64_t kRangeTombstoneSentinel =
-    PackSequenceAndType(kMaxSequenceNumber, kTypeRangeDeletion);
-
-int sstableKeyCompare(const Comparator* user_cmp, const InternalKey& a,
-                      const InternalKey& b) {
-  auto c = user_cmp->Compare(a.user_key(), b.user_key());
-  if (c != 0) {
-    return c;
-  }
-  auto a_footer = ExtractInternalKeyFooter(a.Encode());
-  auto b_footer = ExtractInternalKeyFooter(b.Encode());
-  if (a_footer == kRangeTombstoneSentinel) {
-    if (b_footer != kRangeTombstoneSentinel) {
-      return -1;
-    }
-  } else if (b_footer == kRangeTombstoneSentinel) {
-    return 1;
-  }
-  return 0;
-}
-
-int sstableKeyCompare(const Comparator* user_cmp, const InternalKey* a,
-                      const InternalKey& b) {
-  if (a == nullptr) {
-    return -1;
-  }
-  return sstableKeyCompare(user_cmp, *a, b);
-}
-
-int sstableKeyCompare(const Comparator* user_cmp, const InternalKey& a,
-                      const InternalKey* b) {
-  if (b == nullptr) {
-    return -1;
-  }
-  return sstableKeyCompare(user_cmp, a, *b);
-}
-
 uint64_t TotalFileSize(const std::vector<FileMetaData*>& files) {
  uint64_t sum = 0;
  for (size_t i = 0; i < files.size() && files[i]; i++) {
@@ -118,49 +81,6 @@ void Compaction::GetBoundaryKeys(
  }
 }

-std::vector<CompactionInputFiles> Compaction::PopulateWithAtomicBoundaries(
-    VersionStorageInfo* vstorage, std::vector<CompactionInputFiles> inputs) {
-  const Comparator* ucmp = vstorage->InternalComparator()->user_comparator();
-  for (size_t i = 0; i < inputs.size(); i++) {
-    if (inputs[i].level == 0 || inputs[i].files.empty()) {
-      continue;
-    }
-    inputs[i].atomic_compaction_unit_boundaries.reserve(inputs[i].files.size());
-    AtomicCompactionUnitBoundary cur_boundary;
-    size_t first_atomic_idx = 0;
-    auto add_unit_boundary = [&](size_t to) {
-      if (first_atomic_idx == to) return;
-      for (size_t k = first_atomic_idx; k < to; k++) {
-        inputs[i].atomic_compaction_unit_boundaries.push_back(cur_boundary);
-      }
-      first_atomic_idx = to;
-    };
-    for (size_t j = 0; j < inputs[i].files.size(); j++) {
-      const auto* f = inputs[i].files[j];
-      if (j == 0) {
-        // First file in a level.
-        cur_boundary.smallest = &f->smallest;
-        cur_boundary.largest = &f->largest;
-      } else if (sstableKeyCompare(ucmp, *cur_boundary.largest, f->smallest) ==
-                 0) {
-        // SSTs overlap but the end key of the previous file was not
-        // artificially extended by a range tombstone. Extend the current
-        // boundary.
-        cur_boundary.largest = &f->largest;
-      } else {
-        // Atomic compaction unit has ended.
-        add_unit_boundary(j);
-        cur_boundary.smallest = &f->smallest;
-        cur_boundary.largest = &f->largest;
-      }
-    }
-    add_unit_boundary(inputs[i].files.size());
-    assert(inputs[i].files.size() ==
-           inputs[i].atomic_compaction_unit_boundaries.size());
-  }
-  return inputs;
-}
-
 // helper function to determine if compaction is creating files at the
 // bottommost level
 bool Compaction::IsBottommostLevel(
@@ -235,7 +155,7 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
      output_compression_(_compression),
      output_compression_opts_(_compression_opts),
      deletion_compaction_(_deletion_compaction),
-      inputs_(PopulateWithAtomicBoundaries(vstorage, std::move(_inputs))),
+      inputs_(std::move(_inputs)),
      grandparents_(std::move(_grandparents)),
      score_(_score),
      bottommost_level_(IsBottommostLevel(output_level_, vstorage, inputs_)),
@@ -411,14 +331,12 @@ const char* Compaction::InputLevelSummary(
    if (!is_first) {
      len +=
          snprintf(scratch->buffer + len, sizeof(scratch->buffer) - len, " + ");
-      len = std::min(len, static_cast<int>(sizeof(scratch->buffer)));
    } else {
      is_first = false;
    }
    len += snprintf(scratch->buffer + len, sizeof(scratch->buffer) - len,
                    "%" ROCKSDB_PRIszt "@%d", input_level.size(),
                    input_level.level);
-    len = std::min(len, static_cast<int>(sizeof(scratch->buffer)));
  }
  snprintf(scratch->buffer + len, sizeof(scratch->buffer) - len,
           " files to L%d", output_level());
@@ -15,43 +15,11 @@

 namespace rocksdb {

-// Utility for comparing sstable boundary keys. Returns -1 if either a or b is
-// null which provides the property that a==null indicates a key that is less
-// than any key and b==null indicates a key that is greater than any key. Note
-// that the comparison is performed primarily on the user-key portion of the
-// key. If the user-keys compare equal, an additional test is made to sort
-// range tombstone sentinel keys before other keys with the same user-key. The
-// result is that 2 user-keys will compare equal if they differ purely on
-// their sequence number and value, but the range tombstone sentinel for that
-// user-key will compare not equal. This is necessary because the range
-// tombstone sentinel key is set as the largest key for an sstable even though
-// that key never appears in the database. We don't want adjacent sstables to
-// be considered overlapping if they are separated by the range tombstone
-// sentinel.
-int sstableKeyCompare(const Comparator* user_cmp, const InternalKey& a,
-                      const InternalKey& b);
-int sstableKeyCompare(const Comparator* user_cmp, const InternalKey* a,
-                      const InternalKey& b);
-int sstableKeyCompare(const Comparator* user_cmp, const InternalKey& a,
-                      const InternalKey* b);
-
-// An AtomicCompactionUnitBoundary represents a range of keys [smallest,
-// largest] that exactly spans one ore more neighbouring SSTs on the same
-// level. Every pair of  SSTs in this range "overlap" (i.e., the largest
-// user key of one file is the smallest user key of the next file). These
-// boundaries are propagated down to RangeDelAggregator during compaction
-// to provide safe truncation boundaries for range tombstones.
-struct AtomicCompactionUnitBoundary {
-  const InternalKey* smallest = nullptr;
-  const InternalKey* largest = nullptr;
-};
-
 // The structure that manages compaction input files associated
 // with the same physical level.
 struct CompactionInputFiles {
  int level;
  std::vector<FileMetaData*> files;
-  std::vector<AtomicCompactionUnitBoundary> atomic_compaction_unit_boundaries;
  inline bool empty() const { return files.empty(); }
  inline size_t size() const { return files.size(); }
  inline void clear() { files.clear(); }
@@ -128,12 +96,6 @@ class Compaction {
    return inputs_[compaction_input_level][i];
  }

-  const std::vector<AtomicCompactionUnitBoundary>* boundaries(
-      size_t compaction_input_level) const {
-    assert(compaction_input_level < inputs_.size());
-    return &inputs_[compaction_input_level].atomic_compaction_unit_boundaries;
-  }
-
  // Returns the list of file meta data of the specified compaction
  // input level.
  // REQUIREMENT: "compaction_input_level" must be >= 0 and
@@ -300,13 +262,6 @@ class Compaction {
                              const std::vector<CompactionInputFiles>& inputs,
                              Slice* smallest_key, Slice* largest_key);

-  // Get the atomic file boundaries for all files in the compaction. Necessary
-  // in order to avoid the scenario described in
-  // https://github.com/facebook/rocksdb/pull/4432#discussion_r221072219 and plumb
-  // down appropriate key boundaries to RangeDelAggregator during compaction.
-  static std::vector<CompactionInputFiles> PopulateWithAtomicBoundaries(
-      VersionStorageInfo* vstorage, std::vector<CompactionInputFiles> inputs);
-
  // helper function to determine if compaction with inputs and storage is
  // bottommost
  static bool IsBottommostLevel(
@@ -18,8 +18,8 @@ CompactionIterator::CompactionIterator(
    SequenceNumber earliest_write_conflict_snapshot,
    const SnapshotChecker* snapshot_checker, Env* env,
    bool report_detailed_time, bool expect_valid_internal_key,
-    CompactionRangeDelAggregator* range_del_agg, const Compaction* compaction,
-    const CompactionFilter* compaction_filter,
+    RangeDelAggregator* range_del_agg,
+    const Compaction* compaction, const CompactionFilter* compaction_filter,
    const std::atomic<bool>* shutting_down,
    const SequenceNumber preserve_deletes_seqnum)
    : CompactionIterator(
@@ -36,7 +36,7 @@ CompactionIterator::CompactionIterator(
    SequenceNumber earliest_write_conflict_snapshot,
    const SnapshotChecker* snapshot_checker, Env* env,
    bool report_detailed_time, bool expect_valid_internal_key,
-    CompactionRangeDelAggregator* range_del_agg,
+    RangeDelAggregator* range_del_agg,
    std::unique_ptr<CompactionProxy> compaction,
    const CompactionFilter* compaction_filter,
    const std::atomic<bool>* shutting_down,
@@ -77,12 +77,6 @@ CompactionIterator::CompactionIterator(
    earliest_snapshot_ = snapshots_->at(0);
    latest_snapshot_ = snapshots_->back();
  }
-#ifndef NDEBUG
-  // findEarliestVisibleSnapshot assumes this ordering.
-  for (size_t i = 1; i < snapshots_->size(); ++i) {
-    assert(snapshots_->at(i - 1) <= snapshots_->at(i));
-  }
-#endif
  if (compaction_filter_ != nullptr) {
    if (compaction_filter_->IgnoreSnapshots()) {
      ignore_snapshots_ = true;
@@ -511,31 +505,6 @@ void CompactionIterator::NextFromInput() {
        ++iter_stats_.num_optimized_del_drop_obsolete;
      }
      input_->Next();
-    } else if ((ikey_.type == kTypeDeletion) && bottommost_level_ &&
-               ikeyNotNeededForIncrementalSnapshot()) {
-      // Handle the case where we have a delete key at the bottom most level
-      // We can skip outputting the key iff there are no subsequent puts for this
-      // key
-      ParsedInternalKey next_ikey;
-      input_->Next();
-      // Skip over all versions of this key that happen to occur in the same snapshot
-      // range as the delete
-      while (input_->Valid() &&
-             ParseInternalKey(input_->key(), &next_ikey) &&
-             cmp_->Equal(ikey_.user_key, next_ikey.user_key) &&
-             (prev_snapshot == 0 || next_ikey.sequence > prev_snapshot ||
-              (snapshot_checker_ != nullptr &&
-               UNLIKELY(!snapshot_checker_->IsInSnapshot(next_ikey.sequence,
-                                                         prev_snapshot))))) {
-        input_->Next();
-      }
-      // If you find you still need to output a row with this key, we need to output the
-      // delete too
-      if (input_->Valid() && ParseInternalKey(input_->key(), &next_ikey) &&
-          cmp_->Equal(ikey_.user_key, next_ikey.user_key)) {
-        valid_ = true;
-        at_next_ = true;
-      }
    } else if (ikey_.type == kTypeMerge) {
      if (!merge_helper_->HasOperator()) {
        status_ = Status::InvalidArgument(
@@ -634,23 +603,18 @@ void CompactionIterator::PrepareOutput() {
 inline SequenceNumber CompactionIterator::findEarliestVisibleSnapshot(
    SequenceNumber in, SequenceNumber* prev_snapshot) {
  assert(snapshots_->size());
-  auto snapshots_iter = std::lower_bound(
-      snapshots_->begin(), snapshots_->end(), in);
-  if (snapshots_iter == snapshots_->begin()) {
-    *prev_snapshot = 0;
-  } else {
-    *prev_snapshot = *std::prev(snapshots_iter);
-    assert(*prev_snapshot < in);
-  }
-  for (; snapshots_iter != snapshots_->end(); ++snapshots_iter) {
-    auto cur = *snapshots_iter;
-    assert(in <= cur);
-    if (snapshot_checker_ == nullptr ||
-        snapshot_checker_->IsInSnapshot(in, cur)) {
+  SequenceNumber prev = kMaxSequenceNumber;
+  for (const auto cur : *snapshots_) {
+    assert(prev == kMaxSequenceNumber || prev <= cur);
+    if (cur >= in && (snapshot_checker_ == nullptr ||
+                      snapshot_checker_->IsInSnapshot(in, cur))) {
+      *prev_snapshot = prev == kMaxSequenceNumber ? 0 : prev;
      return cur;
    }
-    *prev_snapshot = cur;
+    prev = cur;
+    assert(prev < kMaxSequenceNumber);
  }
+  *prev_snapshot = prev;
  return kMaxSequenceNumber;
 }

@@ -64,7 +64,7 @@ class CompactionIterator {
                     SequenceNumber earliest_write_conflict_snapshot,
                     const SnapshotChecker* snapshot_checker, Env* env,
                     bool report_detailed_time, bool expect_valid_internal_key,
-                     CompactionRangeDelAggregator* range_del_agg,
+                     RangeDelAggregator* range_del_agg,
                     const Compaction* compaction = nullptr,
                     const CompactionFilter* compaction_filter = nullptr,
                     const std::atomic<bool>* shutting_down = nullptr,
@@ -77,7 +77,7 @@ class CompactionIterator {
                     SequenceNumber earliest_write_conflict_snapshot,
                     const SnapshotChecker* snapshot_checker, Env* env,
                     bool report_detailed_time, bool expect_valid_internal_key,
-                     CompactionRangeDelAggregator* range_del_agg,
+                     RangeDelAggregator* range_del_agg,
                     std::unique_ptr<CompactionProxy> compaction,
                     const CompactionFilter* compaction_filter = nullptr,
                     const std::atomic<bool>* shutting_down = nullptr,
@@ -141,7 +141,7 @@ class CompactionIterator {
  Env* env_;
  bool report_detailed_time_;
  bool expect_valid_internal_key_;
-  CompactionRangeDelAggregator* range_del_agg_;
+  RangeDelAggregator* range_del_agg_;
  std::unique_ptr<CompactionProxy> compaction_;
  const CompactionFilter* compaction_filter_;
  const std::atomic<bool>* shutting_down_;
@@ -221,15 +221,10 @@ class CompactionIteratorTest : public testing::TestWithParam<bool> {
      MergeOperator* merge_op = nullptr, CompactionFilter* filter = nullptr,
      bool bottommost_level = false,
      SequenceNumber earliest_write_conflict_snapshot = kMaxSequenceNumber) {
-    std::unique_ptr<InternalIterator> unfragmented_range_del_iter(
+    std::unique_ptr<InternalIterator> range_del_iter(
        new test::VectorIterator(range_del_ks, range_del_vs));
-    auto tombstone_list = std::make_shared<FragmentedRangeTombstoneList>(
-        std::move(unfragmented_range_del_iter), icmp_);
-    std::unique_ptr<FragmentedRangeTombstoneIterator> range_del_iter(
-        new FragmentedRangeTombstoneIterator(tombstone_list, icmp_,
-                                             kMaxSequenceNumber));
-    range_del_agg_.reset(new CompactionRangeDelAggregator(&icmp_, snapshots_));
-    range_del_agg_->AddTombstones(std::move(range_del_iter));
+    range_del_agg_.reset(new RangeDelAggregator(icmp_, snapshots_));
+    ASSERT_OK(range_del_agg_->AddTombstones(std::move(range_del_iter)));

    std::unique_ptr<CompactionIterator::CompactionProxy> compaction;
    if (filter || bottommost_level) {
@@ -252,8 +247,9 @@ class CompactionIteratorTest : public testing::TestWithParam<bool> {
    c_iter_.reset(new CompactionIterator(
        iter_.get(), cmp_, merge_helper_.get(), last_sequence, &snapshots_,
        earliest_write_conflict_snapshot, snapshot_checker_.get(),
-        Env::Default(), false /* report_detailed_time */, false,
-        range_del_agg_.get(), std::move(compaction), filter, &shutting_down_));
+        Env::Default(), false /* report_detailed_time */,
+        false, range_del_agg_.get(), std::move(compaction), filter,
+        &shutting_down_));
  }

  void AddSnapshot(SequenceNumber snapshot,
@@ -297,7 +293,7 @@ class CompactionIteratorTest : public testing::TestWithParam<bool> {
  std::unique_ptr<MergeHelper> merge_helper_;
  std::unique_ptr<LoggingForwardVectorIterator> iter_;
  std::unique_ptr<CompactionIterator> c_iter_;
-  std::unique_ptr<CompactionRangeDelAggregator> range_del_agg_;
+  std::unique_ptr<RangeDelAggregator> range_del_agg_;
  std::unique_ptr<SnapshotChecker> snapshot_checker_;
  std::atomic<bool> shutting_down_{false};
  FakeCompaction* compaction_proxy_;
@@ -676,12 +672,8 @@ TEST_P(CompactionIteratorTest, ZeroOutSequenceAtBottomLevel) {
 TEST_P(CompactionIteratorTest, RemoveDeletionAtBottomLevel) {
  AddSnapshot(1);
  RunTest({test::KeyStr("a", 1, kTypeDeletion),
-           test::KeyStr("b", 3, kTypeDeletion),
-           test::KeyStr("b", 1, kTypeValue)},
-          {"", "", ""},
-          {test::KeyStr("b", 3, kTypeDeletion),
-           test::KeyStr("b", 0, kTypeValue)},
-          {"", ""},
+           test::KeyStr("b", 2, kTypeDeletion)},
+          {"", ""}, {test::KeyStr("b", 2, kTypeDeletion)}, {""},
          kMaxSequenceNumber /*last_commited_seq*/, nullptr /*merge_operator*/,
          nullptr /*compaction_filter*/, true /*bottommost_level*/);
 }
@@ -850,26 +842,13 @@ TEST_F(CompactionIteratorWithSnapshotCheckerTest,
      {test::KeyStr("a", 1, kTypeDeletion), test::KeyStr("b", 2, kTypeDeletion),
       test::KeyStr("c", 3, kTypeDeletion)},
      {"", "", ""},
-      {},
+      {test::KeyStr("b", 2, kTypeDeletion),
+       test::KeyStr("c", 3, kTypeDeletion)},
      {"", ""}, kMaxSequenceNumber /*last_commited_seq*/,
      nullptr /*merge_operator*/, nullptr /*compaction_filter*/,
      true /*bottommost_level*/);
 }

-TEST_F(CompactionIteratorWithSnapshotCheckerTest,
-       NotRemoveDeletionIfValuePresentToEarlierSnapshot) {
-  AddSnapshot(2,1);
-  RunTest(
-      {test::KeyStr("a", 4, kTypeDeletion), test::KeyStr("a", 1, kTypeValue),
-          test::KeyStr("b", 3, kTypeValue)},
-      {"", "", ""},
-      {test::KeyStr("a", 4, kTypeDeletion), test::KeyStr("a", 0, kTypeValue),
-            test::KeyStr("b", 3, kTypeValue)},
-      {"", "", ""}, kMaxSequenceNumber /*last_commited_seq*/,
-      nullptr /*merge_operator*/, nullptr /*compaction_filter*/,
-      true /*bottommost_level*/);
-}
-
 TEST_F(CompactionIteratorWithSnapshotCheckerTest,
       NotRemoveSingleDeletionIfNotVisibleToEarliestSnapshot) {
  AddSnapshot(2, 1);
@@ -36,7 +36,6 @@
 #include "db/memtable_list.h"
 #include "db/merge_context.h"
 #include "db/merge_helper.h"
-#include "db/range_del_aggregator.h"
 #include "db/version_set.h"
 #include "monitoring/iostats_context_imp.h"
 #include "monitoring/perf_context_imp.h"
@@ -201,7 +200,7 @@ struct CompactionJob::SubcompactionState {
    return *this;
  }

-  // Because member std::unique_ptrs do not have these.
+  // Because member unique_ptrs do not have these.
  SubcompactionState(const SubcompactionState&) = delete;

  SubcompactionState& operator=(const SubcompactionState&) = delete;
@@ -512,10 +511,7 @@ void CompactionJob::GenSubcompactionBoundaries() {
  // size of data covered by keys in that range
  uint64_t sum = 0;
  std::vector<RangeWithSize> ranges;
-  // Get input version from CompactionState since it's already referenced
-  // earlier in SetInputVersioCompaction::SetInputVersion and will not change
-  // when db_mutex_ is released below
-  auto* v = compact_->compaction->input_version();
+  auto* v = cfd->current();
  for (auto it = bounds.begin();;) {
    const Slice a = *it;
    it++;
@@ -525,13 +521,7 @@ void CompactionJob::GenSubcompactionBoundaries() {
    }

    const Slice b = *it;
-
-    // ApproximateSize could potentially create table reader iterator to seek
-    // to the index block and may incur I/O cost in the process. Unlock db
-    // mutex to reduce contention
-    db_mutex_->Unlock();
    uint64_t size = versions_->ApproximateSize(v, a, b, start_lvl, out_lvl + 1);
-    db_mutex_->Lock();
    ranges.emplace_back(a, b, size);
    sum += size;
  }
@@ -603,11 +593,13 @@ Status CompactionJob::Run() {
    thread.join();
  }

+  if (output_directory_) {
+    output_directory_->Fsync();
+  }
+
  compaction_stats_.micros = env_->NowMicros() - start_micros;
  MeasureTime(stats_, COMPACTION_TIME, compaction_stats_.micros);

-  TEST_SYNC_POINT("CompactionJob::Run:BeforeVerify");
-
  // Check if any thread encountered an error during execution
  Status status;
  for (const auto& state : compact_->sub_compact_states) {
@@ -617,10 +609,6 @@ Status CompactionJob::Run() {
    }
  }

-  if (status.ok() && output_directory_) {
-    status = output_directory_->Fsync();
-  }
-
  if (status.ok()) {
    thread_pool.clear();
    std::vector<const FileMetaData*> files_meta;
@@ -805,13 +793,10 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
 void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  assert(sub_compact != nullptr);
  ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
-  CompactionRangeDelAggregator range_del_agg(&cfd->internal_comparator(),
-                                             existing_snapshots_);
-
-  // Although the v2 aggregator is what the level iterator(s) know about,
-  // the AddTombstones calls will be propagated down to the v1 aggregator.
+  std::unique_ptr<RangeDelAggregator> range_del_agg(
+      new RangeDelAggregator(cfd->internal_comparator(), existing_snapshots_));
  std::unique_ptr<InternalIterator> input(versions_->MakeInputIterator(
-      sub_compact->compaction, &range_del_agg, env_optiosn_for_read_));
+      sub_compact->compaction, range_del_agg.get(), env_optiosn_for_read_));

  AutoThreadOperationStageUpdater stage_updater(
      ThreadStatus::STAGE_COMPACTION_PROCESS_KV);
@@ -900,7 +885,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
      input.get(), cfd->user_comparator(), &merge, versions_->LastSequence(),
      &existing_snapshots_, earliest_write_conflict_snapshot_,
      snapshot_checker_, env_, ShouldReportDetailedTime(env_, stats_), false,
-      &range_del_agg, sub_compact->compaction, compaction_filter,
+      range_del_agg.get(), sub_compact->compaction, compaction_filter,
      shutting_down_, preserve_deletes_seqnum_));
  auto c_iter = sub_compact->c_iter.get();
  c_iter->SeekToFirst();
@@ -1038,9 +1023,9 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
        next_key = &c_iter->key();
      }
      CompactionIterationStats range_del_out_stats;
-      status =
-          FinishCompactionOutputFile(input_status, sub_compact, &range_del_agg,
-                                     &range_del_out_stats, next_key);
+      status = FinishCompactionOutputFile(input_status, sub_compact,
+                                          range_del_agg.get(),
+                                          &range_del_out_stats, next_key);
      RecordDroppedKeys(range_del_out_stats,
                        &sub_compact->compaction_job_stats);
      if (sub_compact->outputs.size() == 1) {
@@ -1090,7 +1075,8 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  }

  if (status.ok() && sub_compact->builder == nullptr &&
-      sub_compact->outputs.size() == 0 && !range_del_agg.IsEmpty()) {
+      sub_compact->outputs.size() == 0 &&
+      !range_del_agg->IsEmpty()) {
    // handle subcompaction containing only range deletions
    status = OpenCompactionOutputFile(sub_compact);
  }
@@ -1099,8 +1085,8 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
  // close the output file.
  if (sub_compact->builder != nullptr) {
    CompactionIterationStats range_del_out_stats;
-    Status s = FinishCompactionOutputFile(status, sub_compact, &range_del_agg,
-                                          &range_del_out_stats);
+    Status s = FinishCompactionOutputFile(
+        status, sub_compact, range_del_agg.get(), &range_del_out_stats);
    if (status.ok()) {
      status = s;
    }
@@ -1165,7 +1151,7 @@ void CompactionJob::RecordDroppedKeys(

 Status CompactionJob::FinishCompactionOutputFile(
    const Status& input_status, SubcompactionState* sub_compact,
-    CompactionRangeDelAggregator* range_del_agg,
+    RangeDelAggregator* range_del_agg,
    CompactionIterationStats* range_del_out_stats,
    const Slice* next_table_min_key /* = nullptr */) {
  AutoThreadOperationStageUpdater stage_updater(
@@ -1204,19 +1190,10 @@ Status CompactionJob::FinishCompactionOutputFile(
      lower_bound = nullptr;
    }
    if (next_table_min_key != nullptr) {
-      // This may be the last file in the subcompaction in some cases, so we
-      // need to compare the end key of subcompaction with the next file start
-      // key. When the end key is chosen by the subcompaction, we know that
-      // it must be the biggest key in output file. Therefore, it is safe to
-      // use the smaller key as the upper bound of the output file, to ensure
-      // that there is no overlapping between different output files.
+      // This isn't the last file in the subcompaction, so extend until the next
+      // file starts.
      upper_bound_guard = ExtractUserKey(*next_table_min_key);
-      if (sub_compact->end != nullptr &&
-          ucmp->Compare(upper_bound_guard, *sub_compact->end) >= 0) {
-        upper_bound = sub_compact->end;
-      } else {
-        upper_bound = &upper_bound_guard;
-      }
+      upper_bound = &upper_bound_guard;
    } else {
      // This is the last file in the subcompaction, so extend until the
      // subcompaction ends.
@@ -1226,45 +1203,18 @@ Status CompactionJob::FinishCompactionOutputFile(
    if (existing_snapshots_.size() > 0) {
      earliest_snapshot = existing_snapshots_[0];
    }
-    bool has_overlapping_endpoints;
-    if (upper_bound != nullptr && meta->largest.size() > 0) {
-      has_overlapping_endpoints =
-          ucmp->Compare(meta->largest.user_key(), *upper_bound) == 0;
-    } else {
-      has_overlapping_endpoints = false;
-    }
-
-    // The end key of the subcompaction must be bigger or equal to the upper
-    // bound. If the end of subcompaction is null or the upper bound is null,
-    // it means that this file is the last file in the compaction. So there
-    // will be no overlapping between this file and others.
-    assert(sub_compact->end == nullptr ||
-           upper_bound == nullptr ||
-           ucmp->Compare(*upper_bound , *sub_compact->end) <= 0);
-    auto it = range_del_agg->NewIterator(lower_bound, upper_bound,
-                                         has_overlapping_endpoints);
-    // Position the range tombstone output iterator. There may be tombstone
-    // fragments that are entirely out of range, so make sure that we do not
-    // include those.
+    auto it = range_del_agg->NewIterator();
    if (lower_bound != nullptr) {
      it->Seek(*lower_bound);
-    } else {
-      it->SeekToFirst();
    }
    for (; it->Valid(); it->Next()) {
      auto tombstone = it->Tombstone();
-      if (upper_bound != nullptr) {
-        int cmp = ucmp->Compare(*upper_bound, tombstone.start_key_);
-        if ((has_overlapping_endpoints && cmp < 0) ||
-            (!has_overlapping_endpoints && cmp <= 0)) {
-          // Tombstones starting after upper_bound only need to be included in
-          // the next table. If the current SST ends before upper_bound, i.e.,
-          // `has_overlapping_endpoints == false`, we can also skip over range
-          // tombstones that start exactly at upper_bound. Such range tombstones
-          // will be included in the next file and are not relevant to the point
-          // keys or endpoints of the current file.
-          break;
-        }
+      if (upper_bound != nullptr &&
+          ucmp->Compare(*upper_bound, tombstone.start_key_) <= 0) {
+        // Tombstones starting at upper_bound or later only need to be included
+        // in the next table. Break because subsequent tombstones will start
+        // even later.
+        break;
      }

      if (bottommost_level_ && tombstone.seq_ <= earliest_snapshot) {
@@ -1276,8 +1226,6 @@ Status CompactionJob::FinishCompactionOutputFile(
      }

      auto kv = tombstone.Serialize();
-      assert(lower_bound == nullptr ||
-             ucmp->Compare(*lower_bound, kv.second) < 0);
      sub_compact->builder->Add(kv.first.Encode(), kv.second);
      InternalKey smallest_candidate = std::move(kv.first);
      if (lower_bound != nullptr &&
@@ -1359,7 +1307,9 @@ Status CompactionJob::FinishCompactionOutputFile(
    // VersionEdit.
    assert(!sub_compact->outputs.empty());
    sub_compact->outputs.pop_back();
-    meta = nullptr;
+    sub_compact->builder.reset();
+    sub_compact->current_output_file_size = 0;
+    return s;
  }

  if (s.ok() && (current_entries > 0 || tp.num_range_deletions > 0)) {
@@ -1481,7 +1431,7 @@ Status CompactionJob::OpenCompactionOutputFile(
      TableFileCreationReason::kCompaction);
 #endif  // !ROCKSDB_LITE
  // Make the output file
-  std::unique_ptr<WritableFile> writable_file;
+  unique_ptr<WritableFile> writable_file;
 #ifndef NDEBUG
  bool syncpoint_arg = env_options_.use_direct_writes;
  TEST_SYNC_POINT_CALLBACK("CompactionJob::OpenCompactionOutputFile",
@@ -1513,11 +1463,8 @@ Status CompactionJob::OpenCompactionOutputFile(
  writable_file->SetWriteLifeTimeHint(write_hint_);
  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, env_options_,
-                             db_options_.statistics.get(), listeners));
+  sub_compact->outfile.reset(new WritableFileWriter(
+      std::move(writable_file), env_options_, db_options_.statistics.get()));

  // If the Column family flag is to only optimize filters for hits,
  // we can skip creating filters if this is the bottommost_level where
@@ -29,8 +29,8 @@
 #include "db/version_edit.h"
 #include "db/write_controller.h"
 #include "db/write_thread.h"
-#include "options/cf_options.h"
 #include "options/db_options.h"
+#include "options/cf_options.h"
 #include "port/port.h"
 #include "rocksdb/compaction_filter.h"
 #include "rocksdb/compaction_job_stats.h"
@@ -104,7 +104,7 @@ class CompactionJob {

  Status FinishCompactionOutputFile(
      const Status& input_status, SubcompactionState* sub_compact,
-      CompactionRangeDelAggregator* range_del_agg,
+      RangeDelAggregator* range_del_agg,
      CompactionIterationStats* range_del_out_stats,
      const Slice* next_table_min_key = nullptr);
  Status InstallCompactionResults(const MutableCFOptions& mutable_cf_options);
@@ -79,7 +79,7 @@ class CompactionJobTest : public testing::Test {
        shutting_down_(false),
        preserve_deletes_seqnum_(0),
        mock_table_factory_(new mock::MockTableFactory()),
-        error_handler_(nullptr, db_options_, &mutex_) {
+        error_handler_(db_options_, &mutex_) {
    EXPECT_OK(env_->CreateDirIfMissing(dbname_));
    db_options_.db_paths.emplace_back(dbname_,
                                      std::numeric_limits<uint64_t>::max());
@@ -200,12 +200,12 @@ class CompactionJobTest : public testing::Test {
    new_db.SetLastSequence(0);

    const std::string manifest = DescriptorFileName(dbname_, 1);
-    std::unique_ptr<WritableFile> file;
+    unique_ptr<WritableFile> file;
    Status s = env_->NewWritableFile(
        manifest, &file, env_->OptimizeForManifestWrite(env_options_));
    ASSERT_OK(s);
-    std::unique_ptr<WritableFileWriter> file_writer(
-        new WritableFileWriter(std::move(file), manifest, env_options_));
+    unique_ptr<WritableFileWriter> file_writer(
+        new WritableFileWriter(std::move(file), env_options_));
    {
      log::Writer log(std::move(file_writer), 0, false);
      std::string record;
@@ -18,7 +18,6 @@
 #include <queue>
 #include <string>
 #include <utility>
-#include <vector>
 #include "db/column_family.h"
 #include "monitoring/statistics.h"
 #include "util/filename.h"
@@ -37,7 +36,6 @@ uint64_t TotalCompensatedFileSize(const std::vector<FileMetaData*>& files) {
  }
  return sum;
 }
-}  // anonymous namespace

 bool FindIntraL0Compaction(const std::vector<FileMetaData*>& level_files,
                           size_t min_files_to_compact,
@@ -51,7 +49,7 @@ bool FindIntraL0Compaction(const std::vector<FileMetaData*>& level_files,
  // increasing.
  size_t new_compact_bytes_per_del_file = 0;
  for (span_len = 1; span_len < level_files.size(); ++span_len) {
-    compact_bytes += static_cast<size_t>(level_files[span_len]->fd.file_size);
+    compact_bytes += level_files[span_len]->fd.file_size;
    new_compact_bytes_per_del_file = compact_bytes / span_len;
    if (level_files[span_len]->being_compacted ||
        new_compact_bytes_per_del_file > compact_bytes_per_del_file) {
@@ -71,6 +69,7 @@ bool FindIntraL0Compaction(const std::vector<FileMetaData*>& level_files,
  }
  return false;
 }
+}  // anonymous namespace

 // Determine compression type, based on user options, level of the output
 // file and whether compression is disabled.
@@ -220,8 +219,7 @@ void CompactionPicker::GetRange(const std::vector<CompactionInputFiles>& inputs,

 bool CompactionPicker::ExpandInputsToCleanCut(const std::string& /*cf_name*/,
                                              VersionStorageInfo* vstorage,
-                                              CompactionInputFiles* inputs,
-                                              InternalKey** next_smallest) {
+                                              CompactionInputFiles* inputs) {
  // This isn't good compaction
  assert(!inputs->empty());

@@ -244,8 +242,7 @@ bool CompactionPicker::ExpandInputsToCleanCut(const std::string& /*cf_name*/,
    GetRange(*inputs, &smallest, &largest);
    inputs->clear();
    vstorage->GetOverlappingInputs(level, &smallest, &largest, &inputs->files,
-                                   hint_index, &hint_index, true,
-                                   next_smallest);
+                                   hint_index, &hint_index);
  } while (inputs->size() > old_size);

  // we started off with inputs non-empty and the previous loop only grew
@@ -318,29 +315,13 @@ Compaction* CompactionPicker::CompactFiles(
  // shouldn't have been released since.
  assert(!FilesRangeOverlapWithCompaction(input_files, output_level));

-  CompressionType compression_type;
-  if (compact_options.compression == kDisableCompressionOption) {
-    int base_level;
-    if (ioptions_.compaction_style == kCompactionStyleLevel) {
-      base_level = vstorage->base_level();
-    } else {
-      base_level = 1;
-    }
-    compression_type =
-        GetCompressionType(ioptions_, vstorage, mutable_cf_options,
-                           output_level, base_level);
-  } else {
-    // TODO(ajkr): `CompactionOptions` offers configurable `CompressionType`
-    // without configurable `CompressionOptions`, which is inconsistent.
-    compression_type = compact_options.compression;
-  }
-  auto c = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, input_files, output_level,
-      compact_options.output_file_size_limit,
-      mutable_cf_options.max_compaction_bytes, output_path_id, compression_type,
-      GetCompressionOptions(ioptions_, vstorage, output_level),
-      compact_options.max_subcompactions,
-      /* grandparents */ {}, true);
+  auto c =
+      new Compaction(vstorage, ioptions_, mutable_cf_options, input_files,
+                     output_level, compact_options.output_file_size_limit,
+                     mutable_cf_options.max_compaction_bytes, output_path_id,
+                     compact_options.compression, ioptions_.compression_opts,
+                     compact_options.max_subcompactions,
+                     /* grandparents */ {}, true);
  RegisterCompaction(c);
  return c;
 }
@@ -652,6 +633,7 @@ Compaction* CompactionPicker::CompactRange(
      uint64_t s = inputs[i]->compensated_file_size;
      total += s;
      if (total >= limit) {
+        **compaction_end = inputs[i + 1]->smallest;
        covering_the_whole_range = false;
        inputs.files.resize(i + 1);
        break;
@@ -660,10 +642,7 @@ Compaction* CompactionPicker::CompactRange(
  }
  assert(output_path_id < static_cast<uint32_t>(ioptions_.cf_paths.size()));

-  InternalKey key_storage;
-  InternalKey* next_smallest = &key_storage;
-  if (ExpandInputsToCleanCut(cf_name, vstorage, &inputs, &next_smallest) ==
-      false) {
+  if (ExpandInputsToCleanCut(cf_name, vstorage, &inputs) == false) {
    // manual compaction is now multi-threaded, so it can
    // happen that ExpandWhileOverlapping fails
    // we handle it higher in RunManualCompaction
@@ -671,10 +650,8 @@ Compaction* CompactionPicker::CompactRange(
    return nullptr;
  }

-  if (covering_the_whole_range || !next_smallest) {
+  if (covering_the_whole_range) {
    *compaction_end = nullptr;
-  } else {
-    **compaction_end = *next_smallest;
  }

  CompactionInputFiles output_level_inputs;
@@ -944,8 +921,8 @@ Status CompactionPicker::SanitizeCompactionInputFiles(
  // any currently-existing files.
  for (auto file_num : *input_files) {
    bool found = false;
-    for (const auto& level_meta : cf_meta.levels) {
-      for (const auto& file_meta : level_meta.files) {
+    for (auto level_meta : cf_meta.levels) {
+      for (auto file_meta : level_meta.files) {
        if (file_num == TableFileNameToNumber(file_meta.name)) {
          if (file_meta.being_compacted) {
            return Status::Aborted("Specified compaction input file " +
@@ -1548,4 +1525,217 @@ Compaction* LevelCompactionPicker::PickCompaction(
  return builder.PickCompaction();
 }

+#ifndef ROCKSDB_LITE
+bool FIFOCompactionPicker::NeedsCompaction(
+    const VersionStorageInfo* vstorage) const {
+  const int kLevel0 = 0;
+  return vstorage->CompactionScore(kLevel0) >= 1;
+}
+
+namespace {
+uint64_t GetTotalFilesSize(
+    const std::vector<FileMetaData*>& files) {
+  uint64_t total_size = 0;
+  for (const auto& f : files) {
+    total_size += f->fd.file_size;
+  }
+  return total_size;
+}
+}  // anonymous namespace
+
+Compaction* FIFOCompactionPicker::PickTTLCompaction(
+    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
+  assert(mutable_cf_options.compaction_options_fifo.ttl > 0);
+
+  const int kLevel0 = 0;
+  const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
+  uint64_t total_size = GetTotalFilesSize(level_files);
+
+  int64_t _current_time;
+  auto status = ioptions_.env->GetCurrentTime(&_current_time);
+  if (!status.ok()) {
+    ROCKS_LOG_BUFFER(log_buffer,
+                     "[%s] FIFO compaction: Couldn't get current time: %s. "
+                     "Not doing compactions based on TTL. ",
+                     cf_name.c_str(), status.ToString().c_str());
+    return nullptr;
+  }
+  const uint64_t current_time = static_cast<uint64_t>(_current_time);
+
+  std::vector<CompactionInputFiles> inputs;
+  inputs.emplace_back();
+  inputs[0].level = 0;
+
+  // avoid underflow
+  if (current_time > mutable_cf_options.compaction_options_fifo.ttl) {
+    for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
+      auto f = *ritr;
+      if (f->fd.table_reader != nullptr &&
+          f->fd.table_reader->GetTableProperties() != nullptr) {
+        auto creation_time =
+            f->fd.table_reader->GetTableProperties()->creation_time;
+        if (creation_time == 0 ||
+            creation_time >= (current_time -
+                              mutable_cf_options.compaction_options_fifo.ttl)) {
+          break;
+        }
+        total_size -= f->compensated_file_size;
+        inputs[0].files.push_back(f);
+      }
+    }
+  }
+
+  // Return a nullptr and proceed to size-based FIFO compaction if:
+  // 1. there are no files older than ttl OR
+  // 2. there are a few files older than ttl, but deleting them will not bring
+  //    the total size to be less than max_table_files_size threshold.
+  if (inputs[0].files.empty() ||
+      total_size >
+          mutable_cf_options.compaction_options_fifo.max_table_files_size) {
+    return nullptr;
+  }
+
+  for (const auto& f : inputs[0].files) {
+    ROCKS_LOG_BUFFER(log_buffer,
+                     "[%s] FIFO compaction: picking file %" PRIu64
+                     " with creation time %" PRIu64 " for deletion",
+                     cf_name.c_str(), f->fd.GetNumber(),
+                     f->fd.table_reader->GetTableProperties()->creation_time);
+  }
+
+  Compaction* c = new Compaction(
+      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
+      kNoCompression, ioptions_.compression_opts, /* max_subcompactions */ 0,
+      {}, /* is manual */ false, vstorage->CompactionScore(0),
+      /* is deletion compaction */ true, CompactionReason::kFIFOTtl);
+  return c;
+}
+
+Compaction* FIFOCompactionPicker::PickSizeCompaction(
+    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
+  const int kLevel0 = 0;
+  const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
+  uint64_t total_size = GetTotalFilesSize(level_files);
+
+  if (total_size <=
+          mutable_cf_options.compaction_options_fifo.max_table_files_size ||
+      level_files.size() == 0) {
+    // total size not exceeded
+    if (mutable_cf_options.compaction_options_fifo.allow_compaction &&
+        level_files.size() > 0) {
+      CompactionInputFiles comp_inputs;
+      // try to prevent same files from being compacted multiple times, which
+      // could produce large files that may never TTL-expire. Achieve this by
+      // disallowing compactions with files larger than memtable (inflate its
+      // size by 10% to account for uncompressed L0 files that may have size
+      // slightly greater than memtable size limit).
+      size_t max_compact_bytes_per_del_file =
+          static_cast<size_t>(MultiplyCheckOverflow(
+              static_cast<uint64_t>(mutable_cf_options.write_buffer_size),
+              1.1));
+      if (FindIntraL0Compaction(
+              level_files,
+              mutable_cf_options
+                  .level0_file_num_compaction_trigger /* min_files_to_compact */
+              ,
+              max_compact_bytes_per_del_file, &comp_inputs)) {
+        Compaction* c = new Compaction(
+            vstorage, ioptions_, mutable_cf_options, {comp_inputs}, 0,
+            16 * 1024 * 1024 /* output file size limit */,
+            0 /* max compaction bytes, not applicable */,
+            0 /* output path ID */, mutable_cf_options.compression,
+            ioptions_.compression_opts, 0 /* max_subcompactions */, {},
+            /* is manual */ false, vstorage->CompactionScore(0),
+            /* is deletion compaction */ false,
+            CompactionReason::kFIFOReduceNumFiles);
+        return c;
+      }
+    }
+
+    ROCKS_LOG_BUFFER(
+        log_buffer,
+        "[%s] FIFO compaction: nothing to do. Total size %" PRIu64
+        ", max size %" PRIu64 "\n",
+        cf_name.c_str(), total_size,
+        mutable_cf_options.compaction_options_fifo.max_table_files_size);
+    return nullptr;
+  }
+
+  if (!level0_compactions_in_progress_.empty()) {
+    ROCKS_LOG_BUFFER(
+        log_buffer,
+        "[%s] FIFO compaction: Already executing compaction. No need "
+        "to run parallel compactions since compactions are very fast",
+        cf_name.c_str());
+    return nullptr;
+  }
+
+  std::vector<CompactionInputFiles> inputs;
+  inputs.emplace_back();
+  inputs[0].level = 0;
+
+  for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
+    auto f = *ritr;
+    total_size -= f->compensated_file_size;
+    inputs[0].files.push_back(f);
+    char tmp_fsize[16];
+    AppendHumanBytes(f->fd.GetFileSize(), tmp_fsize, sizeof(tmp_fsize));
+    ROCKS_LOG_BUFFER(log_buffer,
+                     "[%s] FIFO compaction: picking file %" PRIu64
+                     " with size %s for deletion",
+                     cf_name.c_str(), f->fd.GetNumber(), tmp_fsize);
+    if (total_size <=
+        mutable_cf_options.compaction_options_fifo.max_table_files_size) {
+      break;
+    }
+  }
+
+  Compaction* c = new Compaction(
+      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
+      kNoCompression, ioptions_.compression_opts, /* max_subcompactions */ 0,
+      {}, /* is manual */ false, vstorage->CompactionScore(0),
+      /* is deletion compaction */ true, CompactionReason::kFIFOMaxSize);
+  return c;
+}
+
+Compaction* FIFOCompactionPicker::PickCompaction(
+    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
+  assert(vstorage->num_levels() == 1);
+
+  Compaction* c = nullptr;
+  if (mutable_cf_options.compaction_options_fifo.ttl > 0) {
+    c = PickTTLCompaction(cf_name, mutable_cf_options, vstorage, log_buffer);
+  }
+  if (c == nullptr) {
+    c = PickSizeCompaction(cf_name, mutable_cf_options, vstorage, log_buffer);
+  }
+  RegisterCompaction(c);
+  return c;
+}
+
+Compaction* FIFOCompactionPicker::CompactRange(
+    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+    VersionStorageInfo* vstorage, int input_level, int output_level,
+    uint32_t /*output_path_id*/, uint32_t /*max_subcompactions*/,
+    const InternalKey* /*begin*/, const InternalKey* /*end*/,
+    InternalKey** compaction_end, bool* /*manual_conflict*/) {
+#ifdef NDEBUG
+  (void)input_level;
+  (void)output_level;
+#endif
+  assert(input_level == 0);
+  assert(output_level == 0);
+  *compaction_end = nullptr;
+  LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, ioptions_.info_log);
+  Compaction* c =
+      PickCompaction(cf_name, mutable_cf_options, vstorage, &log_buffer);
+  log_buffer.FlushBufferToLog();
+  return c;
+}
+
+#endif  // !ROCKSDB_LITE
+
 }  // namespace rocksdb
@@ -151,8 +151,7 @@ class CompactionPicker {
  // Will return false if it is impossible to apply this compaction.
  bool ExpandInputsToCleanCut(const std::string& cf_name,
                              VersionStorageInfo* vstorage,
-                              CompactionInputFiles* inputs,
-                              InternalKey** next_smallest = nullptr);
+                              CompactionInputFiles* inputs);

  // Returns true if any one of the parent files are being compacted
  bool IsRangeInCompaction(VersionStorageInfo* vstorage,
@@ -235,6 +234,42 @@ class LevelCompactionPicker : public CompactionPicker {
 };

 #ifndef ROCKSDB_LITE
+class FIFOCompactionPicker : public CompactionPicker {
+ public:
+  FIFOCompactionPicker(const ImmutableCFOptions& ioptions,
+                       const InternalKeyComparator* icmp)
+      : CompactionPicker(ioptions, icmp) {}
+
+  virtual Compaction* PickCompaction(const std::string& cf_name,
+                                     const MutableCFOptions& mutable_cf_options,
+                                     VersionStorageInfo* version,
+                                     LogBuffer* log_buffer) override;
+
+  virtual Compaction* CompactRange(
+      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
+      VersionStorageInfo* vstorage, int input_level, int output_level,
+      uint32_t output_path_id, uint32_t max_subcompactions,
+      const InternalKey* begin, const InternalKey* end,
+      InternalKey** compaction_end, bool* manual_conflict) override;
+
+  // The maximum allowed output level.  Always returns 0.
+  virtual int MaxOutputLevel() const override { return 0; }
+
+  virtual bool NeedsCompaction(
+      const VersionStorageInfo* vstorage) const override;
+
+ private:
+  Compaction* PickTTLCompaction(const std::string& cf_name,
+                                const MutableCFOptions& mutable_cf_options,
+                                VersionStorageInfo* version,
+                                LogBuffer* log_buffer);
+
+  Compaction* PickSizeCompaction(const std::string& cf_name,
+                                 const MutableCFOptions& mutable_cf_options,
+                                 VersionStorageInfo* version,
+                                 LogBuffer* log_buffer);
+};
+
 class NullCompactionPicker : public CompactionPicker {
 public:
  NullCompactionPicker(const ImmutableCFOptions& ioptions,
@@ -272,11 +307,6 @@ class NullCompactionPicker : public CompactionPicker {
 };
 #endif  // !ROCKSDB_LITE

-bool FindIntraL0Compaction(const std::vector<FileMetaData*>& level_files,
-                           size_t min_files_to_compact,
-                           uint64_t max_compact_bytes_per_del_file,
-                           CompactionInputFiles* comp_inputs);
-
 CompressionType GetCompressionType(const ImmutableCFOptions& ioptions,
                                   const VersionStorageInfo* vstorage,
                                   const MutableCFOptions& mutable_cf_options,
@@ -1,235 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under both the GPLv2 (found in the
-//  COPYING file in the root directory) and Apache 2.0 License
-//  (found in the LICENSE.Apache file in the root directory).
-//
-// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
-// Use of this source code is governed by a BSD-style license that can be
-// found in the LICENSE file. See the AUTHORS file for names of contributors.
-
-#include "db/compaction_picker_fifo.h"
-#ifndef ROCKSDB_LITE
-
-#ifndef __STDC_FORMAT_MACROS
-#define __STDC_FORMAT_MACROS
-#endif
-
-#include <inttypes.h>
-#include <string>
-#include <vector>
-#include "db/column_family.h"
-#include "util/log_buffer.h"
-#include "util/string_util.h"
-
-namespace rocksdb {
-namespace {
-uint64_t GetTotalFilesSize(const std::vector<FileMetaData*>& files) {
-  uint64_t total_size = 0;
-  for (const auto& f : files) {
-    total_size += f->fd.file_size;
-  }
-  return total_size;
-}
-}  // anonymous namespace
-
-bool FIFOCompactionPicker::NeedsCompaction(
-    const VersionStorageInfo* vstorage) const {
-  const int kLevel0 = 0;
-  return vstorage->CompactionScore(kLevel0) >= 1;
-}
-
-Compaction* FIFOCompactionPicker::PickTTLCompaction(
-    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
-  assert(mutable_cf_options.compaction_options_fifo.ttl > 0);
-
-  const int kLevel0 = 0;
-  const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
-  uint64_t total_size = GetTotalFilesSize(level_files);
-
-  int64_t _current_time;
-  auto status = ioptions_.env->GetCurrentTime(&_current_time);
-  if (!status.ok()) {
-    ROCKS_LOG_BUFFER(log_buffer,
-                     "[%s] FIFO compaction: Couldn't get current time: %s. "
-                     "Not doing compactions based on TTL. ",
-                     cf_name.c_str(), status.ToString().c_str());
-    return nullptr;
-  }
-  const uint64_t current_time = static_cast<uint64_t>(_current_time);
-
-  std::vector<CompactionInputFiles> inputs;
-  inputs.emplace_back();
-  inputs[0].level = 0;
-
-  // avoid underflow
-  if (current_time > mutable_cf_options.compaction_options_fifo.ttl) {
-    for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
-      auto f = *ritr;
-      if (f->fd.table_reader != nullptr &&
-          f->fd.table_reader->GetTableProperties() != nullptr) {
-        auto creation_time =
-            f->fd.table_reader->GetTableProperties()->creation_time;
-        if (creation_time == 0 ||
-            creation_time >= (current_time -
-                              mutable_cf_options.compaction_options_fifo.ttl)) {
-          break;
-        }
-        total_size -= f->compensated_file_size;
-        inputs[0].files.push_back(f);
-      }
-    }
-  }
-
-  // Return a nullptr and proceed to size-based FIFO compaction if:
-  // 1. there are no files older than ttl OR
-  // 2. there are a few files older than ttl, but deleting them will not bring
-  //    the total size to be less than max_table_files_size threshold.
-  if (inputs[0].files.empty() ||
-      total_size >
-          mutable_cf_options.compaction_options_fifo.max_table_files_size) {
-    return nullptr;
-  }
-
-  for (const auto& f : inputs[0].files) {
-    ROCKS_LOG_BUFFER(log_buffer,
-                     "[%s] FIFO compaction: picking file %" PRIu64
-                     " with creation time %" PRIu64 " for deletion",
-                     cf_name.c_str(), f->fd.GetNumber(),
-                     f->fd.table_reader->GetTableProperties()->creation_time);
-  }
-
-  Compaction* c = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
-      kNoCompression, ioptions_.compression_opts, /* max_subcompactions */ 0,
-      {}, /* is manual */ false, vstorage->CompactionScore(0),
-      /* is deletion compaction */ true, CompactionReason::kFIFOTtl);
-  return c;
-}
-
-Compaction* FIFOCompactionPicker::PickSizeCompaction(
-    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
-  const int kLevel0 = 0;
-  const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
-  uint64_t total_size = GetTotalFilesSize(level_files);
-
-  if (total_size <=
-          mutable_cf_options.compaction_options_fifo.max_table_files_size ||
-      level_files.size() == 0) {
-    // total size not exceeded
-    if (mutable_cf_options.compaction_options_fifo.allow_compaction &&
-        level_files.size() > 0) {
-      CompactionInputFiles comp_inputs;
-      // try to prevent same files from being compacted multiple times, which
-      // could produce large files that may never TTL-expire. Achieve this by
-      // disallowing compactions with files larger than memtable (inflate its
-      // size by 10% to account for uncompressed L0 files that may have size
-      // slightly greater than memtable size limit).
-      size_t max_compact_bytes_per_del_file =
-          static_cast<size_t>(MultiplyCheckOverflow(
-              static_cast<uint64_t>(mutable_cf_options.write_buffer_size),
-              1.1));
-      if (FindIntraL0Compaction(
-              level_files,
-              mutable_cf_options
-                  .level0_file_num_compaction_trigger /* min_files_to_compact */
-              ,
-              max_compact_bytes_per_del_file, &comp_inputs)) {
-        Compaction* c = new Compaction(
-            vstorage, ioptions_, mutable_cf_options, {comp_inputs}, 0,
-            16 * 1024 * 1024 /* output file size limit */,
-            0 /* max compaction bytes, not applicable */,
-            0 /* output path ID */, mutable_cf_options.compression,
-            ioptions_.compression_opts, 0 /* max_subcompactions */, {},
-            /* is manual */ false, vstorage->CompactionScore(0),
-            /* is deletion compaction */ false,
-            CompactionReason::kFIFOReduceNumFiles);
-        return c;
-      }
-    }
-
-    ROCKS_LOG_BUFFER(
-        log_buffer,
-        "[%s] FIFO compaction: nothing to do. Total size %" PRIu64
-        ", max size %" PRIu64 "\n",
-        cf_name.c_str(), total_size,
-        mutable_cf_options.compaction_options_fifo.max_table_files_size);
-    return nullptr;
-  }
-
-  if (!level0_compactions_in_progress_.empty()) {
-    ROCKS_LOG_BUFFER(
-        log_buffer,
-        "[%s] FIFO compaction: Already executing compaction. No need "
-        "to run parallel compactions since compactions are very fast",
-        cf_name.c_str());
-    return nullptr;
-  }
-
-  std::vector<CompactionInputFiles> inputs;
-  inputs.emplace_back();
-  inputs[0].level = 0;
-
-  for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
-    auto f = *ritr;
-    total_size -= f->compensated_file_size;
-    inputs[0].files.push_back(f);
-    char tmp_fsize[16];
-    AppendHumanBytes(f->fd.GetFileSize(), tmp_fsize, sizeof(tmp_fsize));
-    ROCKS_LOG_BUFFER(log_buffer,
-                     "[%s] FIFO compaction: picking file %" PRIu64
-                     " with size %s for deletion",
-                     cf_name.c_str(), f->fd.GetNumber(), tmp_fsize);
-    if (total_size <=
-        mutable_cf_options.compaction_options_fifo.max_table_files_size) {
-      break;
-    }
-  }
-
-  Compaction* c = new Compaction(
-      vstorage, ioptions_, mutable_cf_options, std::move(inputs), 0, 0, 0, 0,
-      kNoCompression, ioptions_.compression_opts, /* max_subcompactions */ 0,
-      {}, /* is manual */ false, vstorage->CompactionScore(0),
-      /* is deletion compaction */ true, CompactionReason::kFIFOMaxSize);
-  return c;
-}
-
-Compaction* FIFOCompactionPicker::PickCompaction(
-    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, LogBuffer* log_buffer) {
-  assert(vstorage->num_levels() == 1);
-
-  Compaction* c = nullptr;
-  if (mutable_cf_options.compaction_options_fifo.ttl > 0) {
-    c = PickTTLCompaction(cf_name, mutable_cf_options, vstorage, log_buffer);
-  }
-  if (c == nullptr) {
-    c = PickSizeCompaction(cf_name, mutable_cf_options, vstorage, log_buffer);
-  }
-  RegisterCompaction(c);
-  return c;
-}
-
-Compaction* FIFOCompactionPicker::CompactRange(
-    const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-    VersionStorageInfo* vstorage, int input_level, int output_level,
-    uint32_t /*output_path_id*/, uint32_t /*max_subcompactions*/,
-    const InternalKey* /*begin*/, const InternalKey* /*end*/,
-    InternalKey** compaction_end, bool* /*manual_conflict*/) {
-#ifdef NDEBUG
-  (void)input_level;
-  (void)output_level;
-#endif
-  assert(input_level == 0);
-  assert(output_level == 0);
-  *compaction_end = nullptr;
-  LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, ioptions_.info_log);
-  Compaction* c =
-      PickCompaction(cf_name, mutable_cf_options, vstorage, &log_buffer);
-  log_buffer.FlushBufferToLog();
-  return c;
-}
-
-}  // namespace rocksdb
-#endif  // !ROCKSDB_LITE
@@ -1,52 +0,0 @@
-//  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
-#ifndef ROCKSDB_LITE
-
-#include "db/compaction_picker.h"
-
-namespace rocksdb {
-class FIFOCompactionPicker : public CompactionPicker {
- public:
-  FIFOCompactionPicker(const ImmutableCFOptions& ioptions,
-                       const InternalKeyComparator* icmp)
-      : CompactionPicker(ioptions, icmp) {}
-
-  virtual Compaction* PickCompaction(const std::string& cf_name,
-                                     const MutableCFOptions& mutable_cf_options,
-                                     VersionStorageInfo* version,
-                                     LogBuffer* log_buffer) override;
-
-  virtual Compaction* CompactRange(
-      const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
-      VersionStorageInfo* vstorage, int input_level, int output_level,
-      uint32_t output_path_id, uint32_t max_subcompactions,
-      const InternalKey* begin, const InternalKey* end,
-      InternalKey** compaction_end, bool* manual_conflict) override;
-
-  // The maximum allowed output level.  Always returns 0.
-  virtual int MaxOutputLevel() const override { return 0; }
-
-  virtual bool NeedsCompaction(
-      const VersionStorageInfo* vstorage) const override;
-
- private:
-  Compaction* PickTTLCompaction(const std::string& cf_name,
-                                const MutableCFOptions& mutable_cf_options,
-                                VersionStorageInfo* version,
-                                LogBuffer* log_buffer);
-
-  Compaction* PickSizeCompaction(const std::string& cf_name,
-                                 const MutableCFOptions& mutable_cf_options,
-                                 VersionStorageInfo* version,
-                                 LogBuffer* log_buffer);
-};
-}  // namespace rocksdb
-#endif  // !ROCKSDB_LITE
@@ -4,12 +4,10 @@
 //  (found in the LICENSE.Apache file in the root directory).

 #include "db/compaction_picker.h"
-
 #include <limits>
 #include <string>
 #include <utility>
 #include "db/compaction.h"
-#include "db/compaction_picker_fifo.h"
 #include "db/compaction_picker_universal.h"

 #include "util/logging.h"
@@ -92,8 +90,8 @@ class CompactionPickerTest : public testing::Test {
    f->fd = FileDescriptor(file_number, path_id, file_size);
    f->smallest = InternalKey(smallest, smallest_seq, kTypeValue);
    f->largest = InternalKey(largest, largest_seq, kTypeValue);
-    f->fd.smallest_seqno = smallest_seq;
-    f->fd.largest_seqno = largest_seq;
+    f->smallest_seqno = smallest_seq;
+    f->largest_seqno = largest_seq;
    f->compensated_file_size = file_size;
    f->refs = 0;
    vstorage_->AddFile(level, f);
@@ -97,17 +97,17 @@ void GetSmallestLargestSeqno(const std::vector<FileMetaData*>& files,
                             SequenceNumber* largest_seqno) {
  bool is_first = true;
  for (FileMetaData* f : files) {
-    assert(f->fd.smallest_seqno <= f->fd.largest_seqno);
+    assert(f->smallest_seqno <= f->largest_seqno);
    if (is_first) {
      is_first = false;
-      *smallest_seqno = f->fd.smallest_seqno;
-      *largest_seqno = f->fd.largest_seqno;
+      *smallest_seqno = f->smallest_seqno;
+      *largest_seqno = f->largest_seqno;
    } else {
-      if (f->fd.smallest_seqno < *smallest_seqno) {
-        *smallest_seqno = f->fd.smallest_seqno;
+      if (f->smallest_seqno < *smallest_seqno) {
+        *smallest_seqno = f->smallest_seqno;
      }
-      if (f->fd.largest_seqno > *largest_seqno) {
-        *largest_seqno = f->fd.largest_seqno;
+      if (f->largest_seqno > *largest_seqno) {
+        *largest_seqno = f->largest_seqno;
      }
    }
  }
@@ -365,11 +365,11 @@ Compaction* UniversalCompactionPicker::PickCompaction(
  size_t level_index = 0U;
  if (c->start_level() == 0) {
    for (auto f : *c->inputs(0)) {
-      assert(f->fd.smallest_seqno <= f->fd.largest_seqno);
+      assert(f->smallest_seqno <= f->largest_seqno);
      if (is_first) {
        is_first = false;
      }
-      prev_smallest_seqno = f->fd.smallest_seqno;
+      prev_smallest_seqno = f->smallest_seqno;
    }
    level_index = 1U;
  }
@@ -35,7 +35,7 @@ Status DeleteFilesInRanges(DB* db, ColumnFamilyHandle* column_family,
 Status VerifySstFileChecksum(const Options& options,
                             const EnvOptions& env_options,
                             const std::string& file_path) {
-  std::unique_ptr<RandomAccessFile> file;
+  unique_ptr<RandomAccessFile> file;
  uint64_t file_size;
  InternalKeyComparator internal_comparator(options.comparator);
  ImmutableCFOptions ioptions(options);
@@ -46,7 +46,7 @@ Status VerifySstFileChecksum(const Options& options,
  } else {
    return s;
  }
-  std::unique_ptr<TableReader> table_reader;
+  unique_ptr<TableReader> table_reader;
  std::unique_ptr<RandomAccessFileReader> file_reader(
      new RandomAccessFileReader(std::move(file), file_path));
  const bool kImmortal = true;
@@ -37,7 +37,7 @@ class CorruptionTest : public testing::Test {
 public:
  test::ErrorEnv env_;
  std::string dbname_;
-  std::shared_ptr<Cache> tiny_cache_;
+  shared_ptr<Cache> tiny_cache_;
  Options options_;
  DB* db_;

@@ -485,7 +485,7 @@ TEST_F(CorruptionTest, FileSystemStateCorrupted) {
    db_ = nullptr;

    if (iter == 0) {  // corrupt file size
-      std::unique_ptr<WritableFile> file;
+      unique_ptr<WritableFile> file;
      env_.NewWritableFile(filename, &file, EnvOptions());
      file->Append(Slice("corrupted sst"));
      file.reset();
@@ -794,7 +794,7 @@ TEST_F(DBBasicTest, ChecksumTest) {
  BlockBasedTableOptions table_options;
  Options options = CurrentOptions();
  // change when new checksum type added
-  int max_checksum = static_cast<int>(kxxHash64);
+  int max_checksum = static_cast<int>(kxxHash);
  const int kNumPerFile = 2;

  // generate one table with each type of checksum
@@ -809,7 +809,7 @@ TEST_F(DBBasicTest, ChecksumTest) {
  }

  // verify data with each type of checksum
-  for (int i = 0; i <= kxxHash64; ++i) {
+  for (int i = 0; i <= kxxHash; ++i) {
    table_options.checksum = static_cast<ChecksumType>(i);
    options.table_factory.reset(NewBlockBasedTableFactory(table_options));
    Reopen(options);
@@ -881,7 +881,7 @@ class TestEnv : public EnvWrapper {
    int GetCloseCount() { return close_count; }

    virtual Status NewLogger(const std::string& /*fname*/,
-                             std::shared_ptr<Logger>* result) {
+                             shared_ptr<Logger>* result) {
      result->reset(new TestLogger(this));
      return Status::OK();
    }
@@ -349,7 +349,7 @@ TEST_F(DBBlockCacheTest, IndexAndFilterBlocksStats) {
  // 200 bytes are enough to hold the first two blocks
  std::shared_ptr<Cache> cache = NewLRUCache(200, 0, false);
  table_options.block_cache = cache;
-  table_options.filter_policy.reset(NewBloomFilterPolicy(20, true));
+  table_options.filter_policy.reset(NewBloomFilterPolicy(20));
  options.table_factory.reset(new BlockBasedTableFactory(table_options));
  CreateAndReopenWithCF({"pikachu"}, options);

@@ -22,12 +22,11 @@ class DBBloomFilterTest : public DBTestBase {

 class DBBloomFilterTestWithParam
    : public DBTestBase,
-      public testing::WithParamInterface<std::tuple<bool, bool, uint32_t>> {
+      public testing::WithParamInterface<std::tuple<bool, bool>> {
  //                             public testing::WithParamInterface<bool> {
 protected:
  bool use_block_based_filter_;
  bool partition_filters_;
-  uint32_t format_version_;

 public:
  DBBloomFilterTestWithParam() : DBTestBase("/db_bloom_filter_tests") {}
@@ -37,12 +36,9 @@ class DBBloomFilterTestWithParam
  void SetUp() override {
    use_block_based_filter_ = std::get<0>(GetParam());
    partition_filters_ = std::get<1>(GetParam());
-    format_version_ = std::get<2>(GetParam());
  }
 };

-class DBBloomFilterTestDefFormatVersion : public DBBloomFilterTestWithParam {};
-
 class SliceTransformLimitedDomainGeneric : public SliceTransform {
  const char* Name() const override {
    return "SliceTransformLimitedDomainGeneric";
@@ -66,7 +62,7 @@ class SliceTransformLimitedDomainGeneric : public SliceTransform {
 // KeyMayExist can lead to a few false positives, but not false negatives.
 // To make test deterministic, use a much larger number of bits per key-20 than
 // bits in the key, so that false positives are eliminated
-TEST_P(DBBloomFilterTestDefFormatVersion, KeyMayExist) {
+TEST_P(DBBloomFilterTestWithParam, KeyMayExist) {
  do {
    ReadOptions ropts;
    std::string value;
@@ -147,7 +143,6 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloomCustomPrefixExtractor) {
    options.prefix_extractor =
        std::make_shared<SliceTransformLimitedDomainGeneric>();
    options.statistics = rocksdb::CreateDBStatistics();
-    get_perf_context()->EnablePerLevelPerfContext();
    BlockBasedTableOptions bbto;
    bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
    if (partition_filters) {
@@ -172,39 +167,20 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloomCustomPrefixExtractor) {

    ASSERT_EQ("foo", Get("barbarbar"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
-    ASSERT_EQ(
-        0,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);
    ASSERT_EQ("foo2", Get("barbarbar2"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
-    ASSERT_EQ(
-        0,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);
    ASSERT_EQ("NOT_FOUND", Get("barbarbar3"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
-    ASSERT_EQ(
-        0,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);

    ASSERT_EQ("NOT_FOUND", Get("barfoofoo"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 1);
-    ASSERT_EQ(
-        1,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);

    ASSERT_EQ("NOT_FOUND", Get("foobarbar"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 2);
-    ASSERT_EQ(
-        2,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);

    ro.total_order_seek = true;
    ASSERT_TRUE(db_->Get(ro, "foobarbar", &value).IsNotFound());
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 2);
-    ASSERT_EQ(
-        2,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);
-    get_perf_context()->Reset();
  }
 }

@@ -213,7 +189,6 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloom) {
    Options options = last_options_;
    options.prefix_extractor.reset(NewFixedPrefixTransform(8));
    options.statistics = rocksdb::CreateDBStatistics();
-    get_perf_context()->EnablePerLevelPerfContext();
    BlockBasedTableOptions bbto;
    bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
    if (partition_filters) {
@@ -252,10 +227,6 @@ TEST_F(DBBloomFilterTest, GetFilterByPrefixBloom) {
    ro.total_order_seek = true;
    ASSERT_TRUE(db_->Get(ro, "foobarbar", &value).IsNotFound());
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 2);
-    ASSERT_EQ(
-        2,
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful);
-    get_perf_context()->Reset();
  }
 }

@@ -264,7 +235,6 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
    Options options = last_options_;
    options.prefix_extractor.reset(NewFixedPrefixTransform(3));
    options.statistics = rocksdb::CreateDBStatistics();
-    get_perf_context()->EnablePerLevelPerfContext();

    BlockBasedTableOptions bbto;
    bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
@@ -412,14 +382,6 @@ TEST_F(DBBloomFilterTest, WholeKeyFilterProp) {
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 12);
    ASSERT_EQ("bar", Get("barfoo"));
    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 12);
-    uint64_t bloom_filter_useful_all_levels = 0;
-    for (auto& kv : (*(get_perf_context()->level_to_perf_context))) {
-      if (kv.second.bloom_filter_useful > 0) {
-        bloom_filter_useful_all_levels += kv.second.bloom_filter_useful;
-      }
-    }
-    ASSERT_EQ(12, bloom_filter_useful_all_levels);
-    get_perf_context()->Reset();
  }
 }

@@ -439,11 +401,6 @@ TEST_P(DBBloomFilterTestWithParam, BloomFilter) {
      table_options.index_type =
          BlockBasedTableOptions::IndexType::kTwoLevelIndexSearch;
    }
-    table_options.format_version = format_version_;
-    if (format_version_ >= 4) {
-      // value delta encoding challenged more with index interval > 1
-      table_options.index_block_restart_interval = 8;
-    }
    table_options.metadata_block_size = 32;
    options.table_factory.reset(NewBlockBasedTableFactory(table_options));

@@ -499,34 +456,15 @@ TEST_P(DBBloomFilterTestWithParam, BloomFilter) {
  } while (ChangeCompactOptions());
 }

-#ifndef ROCKSDB_VALGRIND_RUN
-INSTANTIATE_TEST_CASE_P(
-    FormatDef, DBBloomFilterTestDefFormatVersion,
-    ::testing::Values(std::make_tuple(true, false, test::kDefaultFormatVersion),
-                      std::make_tuple(false, true, test::kDefaultFormatVersion),
-                      std::make_tuple(false, false,
-                                      test::kDefaultFormatVersion)));
-
-INSTANTIATE_TEST_CASE_P(
-    FormatDef, DBBloomFilterTestWithParam,
-    ::testing::Values(std::make_tuple(true, false, test::kDefaultFormatVersion),
-                      std::make_tuple(false, true, test::kDefaultFormatVersion),
-                      std::make_tuple(false, false,
-                                      test::kDefaultFormatVersion)));
-
-INSTANTIATE_TEST_CASE_P(
-    FormatLatest, DBBloomFilterTestWithParam,
-    ::testing::Values(std::make_tuple(true, false, test::kLatestFormatVersion),
-                      std::make_tuple(false, true, test::kLatestFormatVersion),
-                      std::make_tuple(false, false,
-                                      test::kLatestFormatVersion)));
-#endif  // ROCKSDB_VALGRIND_RUN
+INSTANTIATE_TEST_CASE_P(DBBloomFilterTestWithParam, DBBloomFilterTestWithParam,
+                        ::testing::Values(std::make_tuple(true, false),
+                                          std::make_tuple(false, true),
+                                          std::make_tuple(false, false)));

 TEST_F(DBBloomFilterTest, BloomFilterRate) {
  while (ChangeFilterOptions()) {
    Options options = CurrentOptions();
    options.statistics = rocksdb::CreateDBStatistics();
-    get_perf_context()->EnablePerLevelPerfContext();
    CreateAndReopenWithCF({"pikachu"}, options);

    const int maxKey = 10000;
@@ -548,10 +486,6 @@ TEST_F(DBBloomFilterTest, BloomFilterRate) {
      ASSERT_EQ("NOT_FOUND", Get(1, Key(i + 33333)));
    }
    ASSERT_GE(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), maxKey * 0.98);
-    ASSERT_GE(
-        (*(get_perf_context()->level_to_perf_context))[0].bloom_filter_useful,
-        maxKey * 0.98);
-    get_perf_context()->Reset();
  }
 }

@@ -897,7 +831,7 @@ TEST_P(BloomStatsTestWithParam, BloomStatsTestWithIter) {
  ASSERT_OK(Put(key1, value1, WriteOptions()));
  ASSERT_OK(Put(key3, value3, WriteOptions()));

-  std::unique_ptr<Iterator> iter(dbfull()->NewIterator(ReadOptions()));
+  unique_ptr<Iterator> iter(dbfull()->NewIterator(ReadOptions()));

  // check memtable bloom stats
  iter->Seek(key1);
@@ -1073,7 +1007,6 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
  options.table_factory.reset(NewBlockBasedTableFactory(bbto));
  options.optimize_filters_for_hits = true;
  options.statistics = rocksdb::CreateDBStatistics();
-  get_perf_context()->EnablePerLevelPerfContext();
  CreateAndReopenWithCF({"mypikachu"}, options);

  int numkeys = 200000;
@@ -1120,14 +1053,6 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
  // no bloom filter. Most keys be checked bloom filters twice.
  ASSERT_GT(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 65000 * 2);
  ASSERT_LT(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 120000 * 2);
-  uint64_t bloom_filter_useful_all_levels = 0;
-  for (auto& kv : (*(get_perf_context()->level_to_perf_context))) {
-    if (kv.second.bloom_filter_useful > 0) {
-      bloom_filter_useful_all_levels += kv.second.bloom_filter_useful;
-    }
-  }
-  ASSERT_GT(bloom_filter_useful_all_levels, 65000 * 2);
-  ASSERT_LT(bloom_filter_useful_all_levels, 120000 * 2);

  for (int i = 0; i < numkeys; i += 2) {
    ASSERT_EQ(Get(1, Key(i)), "val");
@@ -1237,7 +1162,6 @@ TEST_F(DBBloomFilterTest, OptimizeFiltersForHits) {
  ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_FILTER_HIT));
  ASSERT_EQ(2 /* index and data block */,
            TestGetTickerCount(options, BLOCK_CACHE_ADD));
-  get_perf_context()->Reset();
 }

 int CountIter(std::unique_ptr<Iterator>& iter, const Slice& key) {
@@ -340,10 +340,9 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
  Arena arena;
  {
    InternalKeyComparator icmp(options.comparator);
-    ReadRangeDelAggregator range_del_agg(&icmp,
-                                         kMaxSequenceNumber /* upper_bound */);
-    ScopedArenaIterator iter(dbfull()->NewInternalIterator(
-        &arena, &range_del_agg, kMaxSequenceNumber, handles_[1]));
+    RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
+    ScopedArenaIterator iter(
+        dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[1]));
    iter->SeekToFirst();
    ASSERT_OK(iter->status());
    while (iter->Valid()) {
@@ -430,10 +429,9 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
  count = 0;
  {
    InternalKeyComparator icmp(options.comparator);
-    ReadRangeDelAggregator range_del_agg(&icmp,
-                                         kMaxSequenceNumber /* upper_bound */);
-    ScopedArenaIterator iter(dbfull()->NewInternalIterator(
-        &arena, &range_del_agg, kMaxSequenceNumber, handles_[1]));
+    RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
+    ScopedArenaIterator iter(
+        dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[1]));
    iter->SeekToFirst();
    ASSERT_OK(iter->status());
    while (iter->Valid()) {
@@ -648,10 +646,9 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
    int total = 0;
    Arena arena;
    InternalKeyComparator icmp(options.comparator);
-    ReadRangeDelAggregator range_del_agg(&icmp,
-                                         kMaxSequenceNumber /* snapshots */);
-    ScopedArenaIterator iter(dbfull()->NewInternalIterator(
-        &arena, &range_del_agg, kMaxSequenceNumber));
+    RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
+    ScopedArenaIterator iter(
+        dbfull()->NewInternalIterator(&arena, &range_del_agg));
    iter->SeekToFirst();
    ASSERT_OK(iter->status());
    while (iter->Valid()) {
@@ -851,7 +848,7 @@ TEST_F(DBTestCompactionFilter, SkipUntilWithBloomFilter) {
  DestroyAndReopen(options);

  Put("0000000010", "v10");
-  Put("0000000020", "v20");  // skipped
+  Put("0000000020", "v20"); // skipped
  Put("0000000050", "v50");
  Flush();

@@ -8,13 +8,11 @@
 // found in the LICENSE file. See the AUTHORS file for names of contributors.

 #include "db/db_test_util.h"
-#include "port/port.h"
 #include "port/stack_trace.h"
+#include "port/port.h"
 #include "rocksdb/experimental.h"
 #include "rocksdb/utilities/convenience.h"
-#include "util/fault_injection_test_env.h"
 #include "util/sync_point.h"
-
 namespace rocksdb {

 // SYNC_POINT is not supported in released Windows mode.
@@ -122,12 +120,13 @@ class SstStatsCollector : public EventListener {
 public:
  SstStatsCollector() : num_ssts_creation_started_(0) {}

-  void OnTableFileCreationStarted(
-      const TableFileCreationBriefInfo& /* info */) override {
+  void OnTableFileCreationStarted(const TableFileCreationBriefInfo& /* info */) override {
    ++num_ssts_creation_started_;
  }

-  int num_ssts_creation_started() { return num_ssts_creation_started_; }
+  int num_ssts_creation_started() {
+    return num_ssts_creation_started_;
+  }

 private:
  std::atomic<int> num_ssts_creation_started_;
@@ -2479,7 +2478,6 @@ TEST_P(DBCompactionTestWithParam, ManualLevelCompactionOutputPathId) {

    // Compaction range overlaps files
    Compact(1, "p1", "p9", 1);
-    ASSERT_OK(dbfull()->TEST_WaitForCompact());
    ASSERT_EQ("0,1", FilesPerLevel(1));
    ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
    ASSERT_EQ(0, GetSstFileCount(options.db_paths[0].path));
@@ -2495,7 +2493,6 @@ TEST_P(DBCompactionTestWithParam, ManualLevelCompactionOutputPathId) {

    // Compact just the new range
    Compact(1, "b", "f", 1);
-    ASSERT_OK(dbfull()->TEST_WaitForCompact());
    ASSERT_EQ("0,2", FilesPerLevel(1));
    ASSERT_EQ(2, GetSstFileCount(options.db_paths[1].path));
    ASSERT_EQ(0, GetSstFileCount(options.db_paths[0].path));
@@ -2512,7 +2509,6 @@ TEST_P(DBCompactionTestWithParam, ManualLevelCompactionOutputPathId) {
    compact_options.target_path_id = 1;
    compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
    db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);
-    ASSERT_OK(dbfull()->TEST_WaitForCompact());

    ASSERT_EQ("0,1", FilesPerLevel(1));
    ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
@@ -2955,12 +2951,6 @@ TEST_F(DBCompactionTest, SuggestCompactRangeNoTwoLevel0Compactions) {
  dbfull()->TEST_WaitForCompact();
 }

-static std::string ShortKey(int i) {
-  assert(i < 10000);
-  char buf[100];
-  snprintf(buf, sizeof(buf), "key%04d", i);
-  return std::string(buf);
-}

 TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
  int32_t trivial_move = 0;
@@ -2973,28 +2963,10 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
      [&](void* /*arg*/) { non_trivial_move++; });
  rocksdb::SyncPoint::GetInstance()->EnableProcessing();

-  // The key size is guaranteed to be <= 8
-  class ShortKeyComparator : public Comparator {
-    int Compare(const rocksdb::Slice& a,
-                const rocksdb::Slice& b) const override {
-      assert(a.size() <= 8);
-      assert(b.size() <= 8);
-      return BytewiseComparator()->Compare(a, b);
-    }
-    const char* Name() const override { return "ShortKeyComparator"; }
-    void FindShortestSeparator(std::string* start,
-                               const rocksdb::Slice& limit) const override {
-      return BytewiseComparator()->FindShortestSeparator(start, limit);
-    }
-    void FindShortSuccessor(std::string* key) const override {
-      return BytewiseComparator()->FindShortSuccessor(key);
-    }
-  } short_key_cmp;
  Options options = CurrentOptions();
  options.target_file_size_base = 100000000;
  options.write_buffer_size = 100000000;
  options.max_subcompactions = max_subcompactions_;
-  options.comparator = &short_key_cmp;
  DestroyAndReopen(options);

  int32_t value_size = 10 * 1024;  // 10 KB
@@ -3004,7 +2976,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
  // File with keys [ 0 => 99 ]
  for (int i = 0; i < 100; i++) {
    values.push_back(RandomString(&rnd, value_size));
-    ASSERT_OK(Put(ShortKey(i), values[i]));
+    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

@@ -3021,7 +2993,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
  // File with keys [ 100 => 199 ]
  for (int i = 100; i < 200; i++) {
    values.push_back(RandomString(&rnd, value_size));
-    ASSERT_OK(Put(ShortKey(i), values[i]));
+    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

@@ -3039,7 +3011,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
  // File with keys [ 200 => 299 ]
  for (int i = 200; i < 300; i++) {
    values.push_back(RandomString(&rnd, value_size));
-    ASSERT_OK(Put(ShortKey(i), values[i]));
+    ASSERT_OK(Put(Key(i), values[i]));
  }
  ASSERT_OK(Flush());

@@ -3057,7 +3029,7 @@ TEST_P(DBCompactionTestWithParam, ForceBottommostLevelCompaction) {
  ASSERT_EQ(non_trivial_move, 0);

  for (int i = 0; i < 300; i++) {
-    ASSERT_EQ(Get(ShortKey(i)), values[i]);
+    ASSERT_EQ(Get(Key(i)), values[i]);
  }

  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
@@ -3529,13 +3501,12 @@ TEST_F(DBCompactionTest, CompactRangeDelayedByL0FileCount) {
      // ensure the auto compaction doesn't finish until manual compaction has
      // had a chance to be delayed.
      rocksdb::SyncPoint::GetInstance()->LoadDependency(
-          {{"DBImpl::WaitUntilFlushWouldNotStallWrites:StallWait",
-            "CompactionJob::Run():End"}});
+          {{"DBImpl::CompactRange:StallWait", "CompactionJob::Run():End"}});
    } else {
      // ensure the auto-compaction doesn't finish until manual compaction has
      // continued without delay.
      rocksdb::SyncPoint::GetInstance()->LoadDependency(
-          {{"DBImpl::FlushMemTable:StallWaitDone", "CompactionJob::Run():End"}});
+          {{"DBImpl::CompactRange:StallWaitDone", "CompactionJob::Run():End"}});
    }
    rocksdb::SyncPoint::GetInstance()->EnableProcessing();

@@ -3583,13 +3554,12 @@ TEST_F(DBCompactionTest, CompactRangeDelayedByImmMemTableCount) {
      // ensure the flush doesn't finish until manual compaction has had a
      // chance to be delayed.
      rocksdb::SyncPoint::GetInstance()->LoadDependency(
-          {{"DBImpl::WaitUntilFlushWouldNotStallWrites:StallWait",
-            "FlushJob::WriteLevel0Table"}});
+          {{"DBImpl::CompactRange:StallWait", "FlushJob::WriteLevel0Table"}});
    } else {
      // ensure the flush doesn't finish until manual compaction has continued
      // without delay.
      rocksdb::SyncPoint::GetInstance()->LoadDependency(
-          {{"DBImpl::FlushMemTable:StallWaitDone",
+          {{"DBImpl::CompactRange:StallWaitDone",
            "FlushJob::WriteLevel0Table"}});
    }
    rocksdb::SyncPoint::GetInstance()->EnableProcessing();
@@ -3599,7 +3569,6 @@ TEST_F(DBCompactionTest, CompactRangeDelayedByImmMemTableCount) {
      ASSERT_OK(Put(Key(0), RandomString(&rnd, 1024)));
      FlushOptions flush_opts;
      flush_opts.wait = false;
-      flush_opts.allow_write_stall = true;
      dbfull()->Flush(flush_opts);
    }

@@ -3635,7 +3604,7 @@ TEST_F(DBCompactionTest, CompactRangeShutdownWhileDelayed) {
    // The auto-compaction waits until the manual compaction finishes to ensure
    // the signal comes from closing CF/DB, not from compaction making progress.
    rocksdb::SyncPoint::GetInstance()->LoadDependency(
-        {{"DBImpl::WaitUntilFlushWouldNotStallWrites:StallWait",
+        {{"DBImpl::CompactRange:StallWait",
          "DBCompactionTest::CompactRangeShutdownWhileDelayed:PreShutdown"},
         {"DBCompactionTest::CompactRangeShutdownWhileDelayed:PostManual",
          "CompactionJob::Run():End"}});
@@ -3686,21 +3655,18 @@ TEST_F(DBCompactionTest, CompactRangeSkipFlushAfterDelay) {
  // began. So it unblocks CompactRange and precludes its flush. Throughout the
  // test, stall conditions are upheld via high L0 file count.
  rocksdb::SyncPoint::GetInstance()->LoadDependency(
-      {{"DBImpl::WaitUntilFlushWouldNotStallWrites:StallWait",
+      {{"DBImpl::CompactRange:StallWait",
        "DBCompactionTest::CompactRangeSkipFlushAfterDelay:PreFlush"},
       {"DBCompactionTest::CompactRangeSkipFlushAfterDelay:PostFlush",
-        "DBImpl::FlushMemTable:StallWaitDone"},
-       {"DBImpl::FlushMemTable:StallWaitDone", "CompactionJob::Run():End"}});
+        "DBImpl::CompactRange:StallWaitDone"},
+       {"DBImpl::CompactRange:StallWaitDone", "CompactionJob::Run():End"}});
  rocksdb::SyncPoint::GetInstance()->EnableProcessing();

-  //used for the delayable flushes
-  FlushOptions flush_opts;
-  flush_opts.allow_write_stall = true;
  for (int i = 0; i < kNumL0FilesLimit - 1; ++i) {
    for (int j = 0; j < 2; ++j) {
      ASSERT_OK(Put(Key(j), RandomString(&rnd, 1024)));
    }
-    dbfull()->Flush(flush_opts);
+    Flush();
  }
  auto manual_compaction_thread = port::Thread([this]() {
    CompactRangeOptions cro;
@@ -3710,7 +3676,7 @@ TEST_F(DBCompactionTest, CompactRangeSkipFlushAfterDelay) {

  TEST_SYNC_POINT("DBCompactionTest::CompactRangeSkipFlushAfterDelay:PreFlush");
  Put(ToString(0), RandomString(&rnd, 1024));
-  dbfull()->Flush(flush_opts);
+  Flush();
  Put(ToString(0), RandomString(&rnd, 1024));
  TEST_SYNC_POINT("DBCompactionTest::CompactRangeSkipFlushAfterDelay:PostFlush");
  manual_compaction_thread.join();
@@ -3987,167 +3953,6 @@ INSTANTIATE_TEST_CASE_P(
                      CompactionPri::kOldestSmallestSeqFirst,
                      CompactionPri::kMinOverlappingRatio));

-class NoopMergeOperator : public MergeOperator {
- public:
-  NoopMergeOperator() {}
-
-  virtual bool FullMergeV2(const MergeOperationInput& /*merge_in*/,
-                           MergeOperationOutput* merge_out) const override {
-    std::string val("bar");
-    merge_out->new_value = val;
-    return true;
-  }
-
-  virtual const char* Name() const override { return "Noop"; }
-};
-
-TEST_F(DBCompactionTest, PartialManualCompaction) {
-  Options opts = CurrentOptions();
-  opts.num_levels = 3;
-  opts.level0_file_num_compaction_trigger = 10;
-  opts.compression = kNoCompression;
-  opts.merge_operator.reset(new NoopMergeOperator());
-  opts.target_file_size_base = 10240;
-  DestroyAndReopen(opts);
-
-  Random rnd(301);
-  for (auto i = 0; i < 8; ++i) {
-    for (auto j = 0; j < 10; ++j) {
-      Merge("foo", RandomString(&rnd, 1024));
-    }
-    Flush();
-  }
-
-  MoveFilesToLevel(2);
-
-  std::string prop;
-  EXPECT_TRUE(dbfull()->GetProperty(DB::Properties::kLiveSstFilesSize, &prop));
-  uint64_t max_compaction_bytes = atoi(prop.c_str()) / 2;
-  ASSERT_OK(dbfull()->SetOptions(
-      {{"max_compaction_bytes", std::to_string(max_compaction_bytes)}}));
-
-  CompactRangeOptions cro;
-  cro.bottommost_level_compaction = BottommostLevelCompaction::kForce;
-  dbfull()->CompactRange(cro, nullptr, nullptr);
-}
-
-TEST_F(DBCompactionTest, ManualCompactionFailsInReadOnlyMode) {
-  // Regression test for bug where manual compaction hangs forever when the DB
-  // is in read-only mode. Verify it now at least returns, despite failing.
-  const int kNumL0Files = 4;
-  std::unique_ptr<FaultInjectionTestEnv> mock_env(
-      new FaultInjectionTestEnv(Env::Default()));
-  Options opts = CurrentOptions();
-  opts.disable_auto_compactions = true;
-  opts.env = mock_env.get();
-  DestroyAndReopen(opts);
-
-  Random rnd(301);
-  for (int i = 0; i < kNumL0Files; ++i) {
-    // Make sure files are overlapping in key-range to prevent trivial move.
-    Put("key1", RandomString(&rnd, 1024));
-    Put("key2", RandomString(&rnd, 1024));
-    Flush();
-  }
-  ASSERT_EQ(kNumL0Files, NumTableFilesAtLevel(0));
-
-  // Enter read-only mode by failing a write.
-  mock_env->SetFilesystemActive(false);
-  // Make sure this is outside `CompactRange`'s range so that it doesn't fail
-  // early trying to flush memtable.
-  ASSERT_NOK(Put("key3", RandomString(&rnd, 1024)));
-
-  // In the bug scenario, the first manual compaction would fail and forget to
-  // unregister itself, causing the second one to hang forever due to conflict
-  // with a non-running compaction.
-  CompactRangeOptions cro;
-  cro.exclusive_manual_compaction = false;
-  Slice begin_key("key1");
-  Slice end_key("key2");
-  ASSERT_NOK(dbfull()->CompactRange(cro, &begin_key, &end_key));
-  ASSERT_NOK(dbfull()->CompactRange(cro, &begin_key, &end_key));
-
-  // Close before mock_env destruct.
-  Close();
-}
-
-// FixFileIngestionCompactionDeadlock tests and verifies that compaction and
-// file ingestion do not cause deadlock in the event of write stall triggered
-// by number of L0 files reaching level0_stop_writes_trigger.
-TEST_P(DBCompactionTestWithParam, FixFileIngestionCompactionDeadlock) {
-  const int kNumKeysPerFile = 100;
-  // Generate SST files.
-  Options options = CurrentOptions();
-
-  // Generate an external SST file containing a single key, i.e. 99
-  std::string sst_files_dir = dbname_ + "/sst_files/";
-  test::DestroyDir(env_, sst_files_dir);
-  ASSERT_OK(env_->CreateDir(sst_files_dir));
-  SstFileWriter sst_writer(EnvOptions(), options);
-  const std::string sst_file_path = sst_files_dir + "test.sst";
-  ASSERT_OK(sst_writer.Open(sst_file_path));
-  ASSERT_OK(sst_writer.Put(Key(kNumKeysPerFile - 1), "value"));
-  ASSERT_OK(sst_writer.Finish());
-
-  SyncPoint::GetInstance()->DisableProcessing();
-  SyncPoint::GetInstance()->ClearAllCallBacks();
-  SyncPoint::GetInstance()->LoadDependency({
-      {"DBImpl::IngestExternalFile:AfterIncIngestFileCounter",
-       "BackgroundCallCompaction:0"},
-  });
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  options.write_buffer_size = 110 << 10;  // 110KB
-  options.level0_file_num_compaction_trigger =
-      options.level0_stop_writes_trigger;
-  options.max_subcompactions = max_subcompactions_;
-  options.memtable_factory.reset(new SpecialSkipListFactory(kNumKeysPerFile));
-  DestroyAndReopen(options);
-  Random rnd(301);
-
-  // Generate level0_stop_writes_trigger L0 files to trigger write stop
-  for (int i = 0; i != options.level0_file_num_compaction_trigger; ++i) {
-    for (int j = 0; j != kNumKeysPerFile; ++j) {
-      ASSERT_OK(Put(Key(j), RandomString(&rnd, 990)));
-    }
-    if (0 == i) {
-      // When we reach here, the memtables have kNumKeysPerFile keys. Note that
-      // flush is not yet triggered. We need to write an extra key so that the
-      // write path will call PreprocessWrite and flush the previous key-value
-      // pairs to e flushed. After that, there will be the newest key in the
-      // memtable, and a bunch of L0 files. Since there is already one key in
-      // the memtable, then for i = 1, 2, ..., we do not have to write this
-      // extra key to trigger flush.
-      ASSERT_OK(Put("", ""));
-    }
-    dbfull()->TEST_WaitForFlushMemTable();
-    ASSERT_EQ(NumTableFilesAtLevel(0 /*level*/, 0 /*cf*/), i + 1);
-  }
-  // When we reach this point, there will be level0_stop_writes_trigger L0
-  // files and one extra key (99) in memory, which overlaps with the external
-  // SST file. Write stall triggers, and can be cleared only after compaction
-  // reduces the number of L0 files.
-
-  // Compaction will also be triggered since we have reached the threshold for
-  // auto compaction. Note that compaction may begin after the following file
-  // ingestion thread and waits for ingestion to finish.
-
-  // Thread to ingest file with overlapping key range with the current
-  // memtable. Consequently ingestion will trigger a flush. The flush MUST
-  // proceed without waiting for the write stall condition to clear, otherwise
-  // deadlock can happen.
-  port::Thread ingestion_thr([&]() {
-    IngestExternalFileOptions ifo;
-    Status s = db_->IngestExternalFile({sst_file_path}, ifo);
-    ASSERT_OK(s);
-  });
-
-  // More write to trigger write stop
-  ingestion_thr.join();
-  ASSERT_OK(dbfull()->TEST_WaitForCompact());
-  Close();
-}
-
 #endif // !defined(ROCKSDB_LITE)
 }  // namespace rocksdb

@@ -27,7 +27,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase) {
    return;
  }
  // Use InMemoryEnv, or it would be too slow.
-  std::unique_ptr<Env> env(new MockEnv(env_));
+  unique_ptr<Env> env(new MockEnv(env_));

  const int kNKeys = 1000;
  int keys[kNKeys];
@@ -125,7 +125,6 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  int kMaxKey = 1000000;

  Options options = CurrentOptions();
-  options.compression = kNoCompression;
  options.create_if_missing = true;
  options.write_buffer_size = 20480;
  options.max_write_buffer_number = 2;
@@ -168,8 +167,8 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
  ASSERT_EQ(4U, int_prop);

-  // Insert extra about 28K to L0. After they are compacted to L4, the base
-  // level should be changed to L3.
+  // Insert extra about 28K to L0. After they are compacted to L4, base level
+  // should be changed to L3.
  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "true"},
  }));
@@ -190,7 +189,13 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
  ASSERT_TRUE(db_->GetProperty("rocksdb.num-files-at-level2", &str_prop));
  ASSERT_EQ("0", str_prop);

-  // Write even more data while leaving the base level at L3.
+  // Trigger parallel compaction, and the first one would change the base
+  // level.
+  // Hold compaction jobs to make sure
+  rocksdb::SyncPoint::GetInstance()->SetCallBack(
+      "CompactionJob::Run():Start",
+      [&](void* /*arg*/) { env_->SleepForMicroseconds(100000); });
+  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "true"},
  }));
@@ -203,12 +208,18 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
      {"disable_auto_compactions", "false"},
  }));
  Flush();
+  // Wait for 200 milliseconds before proceeding compactions to make sure two
+  // parallel ones are executed.
+  env_->SleepForMicroseconds(200000);
  dbfull()->TEST_WaitForCompact();
  ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
  ASSERT_EQ(3U, int_prop);
+  rocksdb::SyncPoint::GetInstance()->DisableProcessing();

-  // Fill up L0, and then run an (auto) L0->Lmax compaction to raise the base
-  // level to 2.
+  // Trigger a condition that the compaction changes base level and L0->Lbase
+  // happens at the same time.
+  // We try to make last levels' targets to be 40K, 160K, 640K, add triggers
+  // another compaction from 40K->160K.
  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "true"},
  }));
@@ -218,31 +229,23 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
                  RandomString(&rnd, 380)));
  }
-
-  // Make sure that the compaction starts before the last bit of data is
-  // flushed, so that the base level isn't raised to L1.
-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
-      {"CompactionJob::Run():Start", "DynamicLevelMaxBytesBase2:0"},
-  });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
  ASSERT_OK(dbfull()->SetOptions({
      {"disable_auto_compactions", "false"},
  }));
-
-  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:0");
  Flush();
  dbfull()->TEST_WaitForCompact();
  ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
  ASSERT_EQ(2U, int_prop);
+
+  // A manual compaction will trigger the base level to become L2
+  // Keep Writing data until base level changed 2->1. There will be L0->L2
+  // compaction going on at the same time.
  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();

-  // Write more data until the base level changes to L1. There will be
-  // a manual compaction going on at the same time.
  rocksdb::SyncPoint::GetInstance()->LoadDependency({
-      {"CompactionJob::Run():Start", "DynamicLevelMaxBytesBase2:1"},
-      {"DynamicLevelMaxBytesBase2:2", "CompactionJob::Run():End"},
+      {"CompactionJob::Run():Start", "DynamicLevelMaxBytesBase2:0"},
+      {"DynamicLevelMaxBytesBase2:1", "CompactionJob::Run():End"},
      {"DynamicLevelMaxBytesBase2:compact_range_finish",
       "FlushJob::WriteLevel0Table"},
  });
@@ -254,12 +257,12 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
    TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:compact_range_finish");
  });

-  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:1");
+  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:0");
  for (int i = 0; i < 2; i++) {
    ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
                  RandomString(&rnd, 380)));
  }
-  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:2");
+  TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:1");

  Flush();

@@ -40,7 +40,7 @@ TEST_F(DBEncryptionTest, CheckEncrypted) {
      continue;
    }
    auto filePath = dbname_ + "/" + *it;
-    std::unique_ptr<SequentialFile> seqFile;
+    unique_ptr<SequentialFile> seqFile;
    auto envOptions = EnvOptions(CurrentOptions());
    status = defaultEnv->NewSequentialFile(filePath, &seqFile, envOptions);
    ASSERT_OK(status);
@@ -44,7 +44,7 @@ Status DBImpl::EnableFileDeletions(bool force) {
  // Job id == 0 means that this is not our background process, but rather
  // user thread
  JobContext job_context(0);
-  bool file_deletion_enabled = false;
+  bool should_purge_files = false;
  {
    InstrumentedMutexLock l(&mutex_);
    if (force) {
@@ -54,18 +54,19 @@ Status DBImpl::EnableFileDeletions(bool force) {
      --disable_delete_obsolete_files_;
    }
    if (disable_delete_obsolete_files_ == 0)  {
-      file_deletion_enabled = true;
+      ROCKS_LOG_INFO(immutable_db_options_.info_log, "File Deletions Enabled");
+      should_purge_files = true;
      FindObsoleteFiles(&job_context, true);
      bg_cv_.SignalAll();
+    } else {
+      ROCKS_LOG_WARN(
+          immutable_db_options_.info_log,
+          "File Deletions Enable, but not really enabled. Counter: %d",
+          disable_delete_obsolete_files_);
    }
  }
-  if (file_deletion_enabled) {
-    ROCKS_LOG_INFO(immutable_db_options_.info_log, "File Deletions Enabled");
+  if (should_purge_files)  {
    PurgeObsoleteFiles(job_context);
-  } else {
-    ROCKS_LOG_WARN(immutable_db_options_.info_log,
-                   "File Deletions Enable, but not really enabled. Counter: %d",
-                   disable_delete_obsolete_files_);
  }
  job_context.Clean();
  LogFlush(immutable_db_options_.info_log);
@@ -86,28 +87,19 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
  if (flush_memtable) {
    // flush all dirty data to disk.
    Status status;
-    if (immutable_db_options_.atomic_flush) {
-      autovector<ColumnFamilyData*> cfds;
-      SelectColumnFamiliesForAtomicFlush(&cfds);
+    for (auto cfd : *versions_->GetColumnFamilySet()) {
+      if (cfd->IsDropped()) {
+        continue;
+      }
+      cfd->Ref();
      mutex_.Unlock();
-      status = AtomicFlushMemTables(cfds, FlushOptions(),
-                                    FlushReason::kGetLiveFiles);
+      status = FlushMemTable(cfd, FlushOptions(), FlushReason::kGetLiveFiles);
+      TEST_SYNC_POINT("DBImpl::GetLiveFiles:1");
+      TEST_SYNC_POINT("DBImpl::GetLiveFiles:2");
      mutex_.Lock();
-    } else {
-      for (auto cfd : *versions_->GetColumnFamilySet()) {
-        if (cfd->IsDropped()) {
-          continue;
-        }
-        cfd->Ref();
-        mutex_.Unlock();
-        status = FlushMemTable(cfd, FlushOptions(), FlushReason::kGetLiveFiles);
-        TEST_SYNC_POINT("DBImpl::GetLiveFiles:1");
-        TEST_SYNC_POINT("DBImpl::GetLiveFiles:2");
-        mutex_.Lock();
-        cfd->Unref();
-        if (!status.ok()) {
-          break;
-        }
+      cfd->Unref();
+      if (!status.ok()) {
+        break;
      }
    }
    versions_->GetColumnFamilySet()->FreeDeadColumnFamilies();
@@ -134,7 +126,7 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,

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

@@ -25,12 +25,6 @@ class DBFlushDirectIOTest : public DBFlushTest,
  DBFlushDirectIOTest() : DBFlushTest() {}
 };

-class DBAtomicFlushTest : public DBFlushTest,
-                          public ::testing::WithParamInterface<bool> {
- public:
-  DBAtomicFlushTest() : DBFlushTest() {}
-};
-
 // We had issue when two background threads trying to flush at the same time,
 // only one of them get committed. The test verifies the issue is fixed.
 TEST_F(DBFlushTest, FlushWhileWritingManifest) {
@@ -41,12 +35,11 @@ TEST_F(DBFlushTest, FlushWhileWritingManifest) {
  Reopen(options);
  FlushOptions no_wait;
  no_wait.wait = false;
-  no_wait.allow_write_stall=true;

  SyncPoint::GetInstance()->LoadDependency(
      {{"VersionSet::LogAndApply:WriteManifest",
        "DBFlushTest::FlushWhileWritingManifest:1"},
-       {"MemTableList::TryInstallMemtableFlushResults:InProgress",
+       {"MemTableList::InstallMemtableFlushResults:InProgress",
        "VersionSet::LogAndApply:WriteManifestDone"}});
  SyncPoint::GetInstance()->EnableProcessing();

@@ -62,8 +55,6 @@ TEST_F(DBFlushTest, FlushWhileWritingManifest) {
 #endif  // ROCKSDB_LITE
 }

-// Disable this test temporarily on Travis as it fails intermittently.
-// Github issue: #4151
 TEST_F(DBFlushTest, SyncFail) {
  std::unique_ptr<FaultInjectionTestEnv> fault_injection_env(
      new FaultInjectionTestEnv(env_));
@@ -72,15 +63,11 @@ 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);
+  Reopen(options);
  Put("key", "value");
  auto* cfd =
      reinterpret_cast<ColumnFamilyHandleImpl*>(db_->DefaultColumnFamily())
@@ -91,10 +78,6 @@ TEST_F(DBFlushTest, SyncFail) {
  // 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");
@@ -110,30 +93,6 @@ TEST_F(DBFlushTest, SyncFail) {
  Destroy(options);
 }

-TEST_F(DBFlushTest, SyncSkip) {
-  Options options = CurrentOptions();
-
-  SyncPoint::GetInstance()->LoadDependency(
-      {{"DBFlushTest::SyncSkip:1", "DBImpl::SyncClosedLogs:Skip"},
-       {"DBImpl::SyncClosedLogs:Skip", "DBFlushTest::SyncSkip:2"}});
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  Reopen(options);
-  Put("key", "value");
-
-  FlushOptions flush_options;
-  flush_options.wait = false;
-  ASSERT_OK(dbfull()->Flush(flush_options));
-
-  TEST_SYNC_POINT("DBFlushTest::SyncSkip:1");
-  TEST_SYNC_POINT("DBFlushTest::SyncSkip:2");
-
-  // Now the background job will do the flush; wait for it.
-  dbfull()->TEST_WaitForFlushMemTable();
-
-  Destroy(options);
-}
-
 TEST_F(DBFlushTest, FlushInLowPriThreadPool) {
  // Verify setting an empty high-pri (flush) thread pool causes flushes to be
  // scheduled in the low-pri (compaction) thread pool.
@@ -250,249 +209,9 @@ TEST_F(DBFlushTest, FlushError) {
  ASSERT_NE(s, Status::OK());
 }

-TEST_F(DBFlushTest, ManualFlushFailsInReadOnlyMode) {
-  // Regression test for bug where manual flush hangs forever when the DB
-  // is in read-only mode. Verify it now at least returns, despite failing.
-  Options options;
-  std::unique_ptr<FaultInjectionTestEnv> fault_injection_env(
-      new FaultInjectionTestEnv(env_));
-  options.env = fault_injection_env.get();
-  options.max_write_buffer_number = 2;
-  Reopen(options);
-
-  // Trigger a first flush but don't let it run
-  ASSERT_OK(db_->PauseBackgroundWork());
-  ASSERT_OK(Put("key1", "value1"));
-  FlushOptions flush_opts;
-  flush_opts.wait = false;
-  ASSERT_OK(db_->Flush(flush_opts));
-
-  // Write a key to the second memtable so we have something to flush later
-  // after the DB is in read-only mode.
-  ASSERT_OK(Put("key2", "value2"));
-
-  // Let the first flush continue, hit an error, and put the DB in read-only
-  // mode.
-  fault_injection_env->SetFilesystemActive(false);
-  ASSERT_OK(db_->ContinueBackgroundWork());
-  dbfull()->TEST_WaitForFlushMemTable();
-#ifndef ROCKSDB_LITE
-  uint64_t num_bg_errors;
-  ASSERT_TRUE(db_->GetIntProperty(DB::Properties::kBackgroundErrors,
-                                  &num_bg_errors));
-  ASSERT_GT(num_bg_errors, 0);
-#endif  // ROCKSDB_LITE
-
-  // In the bug scenario, triggering another flush would cause the second flush
-  // to hang forever. After the fix we expect it to return an error.
-  ASSERT_NOK(db_->Flush(FlushOptions()));
-
-  Close();
-}
-
-TEST_P(DBAtomicFlushTest, ManualAtomicFlush) {
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.atomic_flush = GetParam();
-  options.write_buffer_size = (static_cast<size_t>(64) << 20);
-
-  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
-  size_t num_cfs = handles_.size();
-  ASSERT_EQ(3, num_cfs);
-  WriteOptions wopts;
-  wopts.disableWAL = true;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    ASSERT_OK(Put(static_cast<int>(i) /*cf*/, "key", "value", wopts));
-  }
-  std::vector<int> cf_ids;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    cf_ids.emplace_back(static_cast<int>(i));
-  }
-  ASSERT_OK(Flush(cf_ids));
-  for (size_t i = 0; i != num_cfs; ++i) {
-    auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
-    ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
-    ASSERT_TRUE(cfh->cfd()->mem()->IsEmpty());
-  }
-}
-
-TEST_P(DBAtomicFlushTest, AtomicFlushTriggeredByMemTableFull) {
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.atomic_flush = GetParam();
-  // 4KB so that we can easily trigger auto flush.
-  options.write_buffer_size = 4096;
-
-  SyncPoint::GetInstance()->LoadDependency(
-      {{"DBImpl::BackgroundCallFlush:FlushFinish:0",
-        "DBAtomicFlushTest::AtomicFlushTriggeredByMemTableFull:BeforeCheck"}});
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
-  size_t num_cfs = handles_.size();
-  ASSERT_EQ(3, num_cfs);
-  WriteOptions wopts;
-  wopts.disableWAL = true;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    ASSERT_OK(Put(static_cast<int>(i) /*cf*/, "key", "value", wopts));
-  }
-  // Keep writing to one of them column families to trigger auto flush.
-  for (int i = 0; i != 4000; ++i) {
-    ASSERT_OK(Put(static_cast<int>(num_cfs) - 1 /*cf*/,
-                  "key" + std::to_string(i), "value" + std::to_string(i),
-                  wopts));
-  }
-
-  TEST_SYNC_POINT(
-      "DBAtomicFlushTest::AtomicFlushTriggeredByMemTableFull:BeforeCheck");
-  if (options.atomic_flush) {
-    for (size_t i = 0; i != num_cfs - 1; ++i) {
-      auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
-      ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
-      ASSERT_TRUE(cfh->cfd()->mem()->IsEmpty());
-    }
-  } else {
-    for (size_t i = 0; i != num_cfs - 1; ++i) {
-      auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
-      ASSERT_EQ(0, cfh->cfd()->imm()->NumNotFlushed());
-      ASSERT_FALSE(cfh->cfd()->mem()->IsEmpty());
-    }
-  }
-  SyncPoint::GetInstance()->DisableProcessing();
-}
-
-TEST_P(DBAtomicFlushTest, AtomicFlushRollbackSomeJobs) {
-  bool atomic_flush = GetParam();
-  if (!atomic_flush) {
-    return;
-  }
-  std::unique_ptr<FaultInjectionTestEnv> fault_injection_env(
-      new FaultInjectionTestEnv(env_));
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.atomic_flush = atomic_flush;
-  options.env = fault_injection_env.get();
-  SyncPoint::GetInstance()->DisableProcessing();
-  SyncPoint::GetInstance()->LoadDependency(
-      {{"DBImpl::AtomicFlushMemTablesToOutputFiles:SomeFlushJobsComplete:1",
-        "DBAtomicFlushTest::AtomicFlushRollbackSomeJobs:1"},
-       {"DBAtomicFlushTest::AtomicFlushRollbackSomeJobs:2",
-        "DBImpl::AtomicFlushMemTablesToOutputFiles:SomeFlushJobsComplete:2"}});
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
-  size_t num_cfs = handles_.size();
-  ASSERT_EQ(3, num_cfs);
-  WriteOptions wopts;
-  wopts.disableWAL = true;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    int cf_id = static_cast<int>(i);
-    ASSERT_OK(Put(cf_id, "key", "value", wopts));
-  }
-  FlushOptions flush_opts;
-  flush_opts.wait = false;
-  ASSERT_OK(dbfull()->Flush(flush_opts, handles_));
-  TEST_SYNC_POINT("DBAtomicFlushTest::AtomicFlushRollbackSomeJobs:1");
-  fault_injection_env->SetFilesystemActive(false);
-  TEST_SYNC_POINT("DBAtomicFlushTest::AtomicFlushRollbackSomeJobs:2");
-  for (auto* cfh : handles_) {
-    dbfull()->TEST_WaitForFlushMemTable(cfh);
-  }
-  for (size_t i = 0; i != num_cfs; ++i) {
-    auto cfh = static_cast<ColumnFamilyHandleImpl*>(handles_[i]);
-    ASSERT_EQ(1, cfh->cfd()->imm()->NumNotFlushed());
-    ASSERT_TRUE(cfh->cfd()->mem()->IsEmpty());
-  }
-  fault_injection_env->SetFilesystemActive(true);
-  Destroy(options);
-}
-
-TEST_P(DBAtomicFlushTest, FlushMultipleCFs_DropSomeBeforeRequestFlush) {
-  bool atomic_flush = GetParam();
-  if (!atomic_flush) {
-    return;
-  }
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.atomic_flush = atomic_flush;
-  SyncPoint::GetInstance()->DisableProcessing();
-  SyncPoint::GetInstance()->ClearAllCallBacks();
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
-  size_t num_cfs = handles_.size();
-  ASSERT_EQ(3, num_cfs);
-  WriteOptions wopts;
-  wopts.disableWAL = true;
-  std::vector<int> cf_ids;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    int cf_id = static_cast<int>(i);
-    ASSERT_OK(Put(cf_id, "key", "value", wopts));
-    cf_ids.push_back(cf_id);
-  }
-  ASSERT_OK(dbfull()->DropColumnFamily(handles_[1]));
-  ASSERT_TRUE(Flush(cf_ids).IsShutdownInProgress());
-  Destroy(options);
-}
-
-TEST_P(DBAtomicFlushTest,
-       FlushMultipleCFs_DropSomeAfterScheduleFlushBeforeFlushJobRun) {
-  bool atomic_flush = GetParam();
-  if (!atomic_flush) {
-    return;
-  }
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.atomic_flush = atomic_flush;
-
-  CreateAndReopenWithCF({"pikachu", "eevee"}, options);
-
-  SyncPoint::GetInstance()->DisableProcessing();
-  SyncPoint::GetInstance()->ClearAllCallBacks();
-  SyncPoint::GetInstance()->LoadDependency(
-      {{"DBImpl::AtomicFlushMemTables:AfterScheduleFlush",
-        "DBAtomicFlushTest::BeforeDropCF"},
-       {"DBAtomicFlushTest::AfterDropCF",
-        "DBImpl::BackgroundCallFlush:start"}});
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  size_t num_cfs = handles_.size();
-  ASSERT_EQ(3, num_cfs);
-  WriteOptions wopts;
-  wopts.disableWAL = true;
-  for (size_t i = 0; i != num_cfs; ++i) {
-    int cf_id = static_cast<int>(i);
-    ASSERT_OK(Put(cf_id, "key", "value", wopts));
-  }
-  port::Thread user_thread([&]() {
-    TEST_SYNC_POINT("DBAtomicFlushTest::BeforeDropCF");
-    ASSERT_OK(dbfull()->DropColumnFamily(handles_[1]));
-    TEST_SYNC_POINT("DBAtomicFlushTest::AfterDropCF");
-  });
-  FlushOptions flush_opts;
-  flush_opts.wait = true;
-  ASSERT_OK(dbfull()->Flush(flush_opts, handles_));
-  user_thread.join();
-  for (size_t i = 0; i != num_cfs; ++i) {
-    int cf_id = static_cast<int>(i);
-    ASSERT_EQ("value", Get(cf_id, "key"));
-  }
-
-  ReopenWithColumnFamilies({kDefaultColumnFamilyName, "eevee"}, options);
-  num_cfs = handles_.size();
-  ASSERT_EQ(2, num_cfs);
-  for (size_t i = 0; i != num_cfs; ++i) {
-    int cf_id = static_cast<int>(i);
-    ASSERT_EQ("value", Get(cf_id, "key"));
-  }
-  Destroy(options);
-}
-
 INSTANTIATE_TEST_CASE_P(DBFlushDirectIOTest, DBFlushDirectIOTest,
                        testing::Bool());

-INSTANTIATE_TEST_CASE_P(DBAtomicFlushTest, DBAtomicFlushTest, testing::Bool());
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -45,7 +45,7 @@
 #include "db/memtable_list.h"
 #include "db/merge_context.h"
 #include "db/merge_helper.h"
-#include "db/range_tombstone_fragmenter.h"
+#include "db/range_del_aggregator.h"
 #include "db/table_cache.h"
 #include "db/table_properties_collector.h"
 #include "db/transaction_log_impl.h"
@@ -215,12 +215,9 @@ DBImpl::DBImpl(const DBOptions& options, const std::string& dbname,
      // requires a custom gc for compaction, we use that to set use_custom_gc_
      // as well.
      use_custom_gc_(seq_per_batch),
-      shutdown_initiated_(false),
-      own_sfm_(options.sst_file_manager == nullptr),
      preserve_deletes_(options.preserve_deletes),
      closed_(false),
-      error_handler_(this, immutable_db_options_, &mutex_),
-      atomic_flush_install_cv_(&mutex_) {
+      error_handler_(immutable_db_options_, &mutex_) {
  // !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_);
@@ -262,85 +259,16 @@ Status DBImpl::Resume() {
    return Status::OK();
  }

-  if (error_handler_.IsRecoveryInProgress()) {
-    // Don't allow a mix of manual and automatic recovery
-    return Status::Busy();
-  }
-
-  mutex_.Unlock();
-  Status s = error_handler_.RecoverFromBGError(true);
-  mutex_.Lock();
-  return s;
-}
-
-// This function implements the guts of recovery from a background error. It
-// is eventually called for both manual as well as automatic recovery. It does
-// the following -
-// 1. Wait for currently scheduled background flush/compaction to exit, in
-//    order to inadvertently causing an error and thinking recovery failed
-// 2. Flush memtables if there's any data for all the CFs. This may result
-//    another error, which will be saved by error_handler_ and reported later
-//    as the recovery status
-// 3. Find and delete any obsolete files
-// 4. Schedule compactions if needed for all the CFs. This is needed as the
-//    flush in the prior step might have been a no-op for some CFs, which
-//    means a new super version wouldn't have been installed
-Status DBImpl::ResumeImpl() {
-  mutex_.AssertHeld();
-  WaitForBackgroundWork();
-
-  Status bg_error = error_handler_.GetBGError();
-  Status s;
-  if (shutdown_initiated_) {
-    // Returning shutdown status to SFM during auto recovery will cause it
-    // to abort the recovery and allow the shutdown to progress
-    s = Status::ShutdownInProgress();
-  }
-  if (s.ok() && bg_error.severity() > Status::Severity::kHardError) {
+  Status s = error_handler_.GetBGError();
+  if (s.severity() > Status::Severity::kHardError) {
    ROCKS_LOG_INFO(immutable_db_options_.info_log,
        "DB resume requested but failed due to Fatal/Unrecoverable error");
-    s = bg_error;
-  }
-
-  // We cannot guarantee consistency of the WAL. So force flush Memtables of
-  // all the column families
-  if (s.ok()) {
-    FlushOptions flush_opts;
-    // We allow flush to stall write since we are trying to resume from error.
-    flush_opts.allow_write_stall = true;
-    if (immutable_db_options_.atomic_flush) {
-      autovector<ColumnFamilyData*> cfds;
-      SelectColumnFamiliesForAtomicFlush(&cfds);
-      mutex_.Unlock();
-      s = AtomicFlushMemTables(cfds, flush_opts, FlushReason::kErrorRecovery);
-      mutex_.Lock();
-    } else {
-      for (auto cfd : *versions_->GetColumnFamilySet()) {
-        if (cfd->IsDropped()) {
-          continue;
-        }
-        cfd->Ref();
-        mutex_.Unlock();
-        s = FlushMemTable(cfd, flush_opts, FlushReason::kErrorRecovery);
-        mutex_.Lock();
-        cfd->Unref();
-        if (!s.ok()) {
-          break;
-        }
-      }
-    }
-    if (!s.ok()) {
-      ROCKS_LOG_INFO(immutable_db_options_.info_log,
-                     "DB resume requested but failed due to Flush failure [%s]",
-                     s.ToString().c_str());
-    }
+    return s;
  }

  JobContext job_context(0);
  FindObsoleteFiles(&job_context, true);
-  if (s.ok()) {
-    s = error_handler_.ClearBGError();
-  }
+  error_handler_.ClearBGError();
  mutex_.Unlock();

  job_context.manifest_file_number = 1;
@@ -349,72 +277,32 @@ Status DBImpl::ResumeImpl() {
  }
  job_context.Clean();

-  if (s.ok()) {
-    ROCKS_LOG_INFO(immutable_db_options_.info_log, "Successfully resumed DB");
-  }
+  ROCKS_LOG_INFO(immutable_db_options_.info_log, "Successfully resumed DB");
  mutex_.Lock();
-  // Check for shutdown again before scheduling further compactions,
-  // since we released and re-acquired the lock above
-  if (shutdown_initiated_) {
-    s = Status::ShutdownInProgress();
-  }
-  if (s.ok()) {
-    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      SchedulePendingCompaction(cfd);
-    }
-    MaybeScheduleFlushOrCompaction();
-  }
-
-  // Wake up any waiters - in this case, it could be the shutdown thread
-  bg_cv_.SignalAll();
+  MaybeScheduleFlushOrCompaction();

  // No need to check BGError again. If something happened, event listener would be
  // notified and the operation causing it would have failed
-  return s;
-}
-
-void DBImpl::WaitForBackgroundWork() {
-  // Wait for background work to finish
-  while (bg_bottom_compaction_scheduled_ || bg_compaction_scheduled_ ||
-         bg_flush_scheduled_) {
-    bg_cv_.Wait();
-  }
+  return Status::OK();
 }

 // Will lock the mutex_,  will wait for completion if wait is true
 void DBImpl::CancelAllBackgroundWork(bool wait) {
+  InstrumentedMutexLock l(&mutex_);

  ROCKS_LOG_INFO(immutable_db_options_.info_log,
                 "Shutdown: canceling all background work");

-  InstrumentedMutexLock l(&mutex_);
-  // To avoid deadlock, `thread_dump_stats_->cancel()` needs to be called
-  // before grabbing db mutex because the actual worker function
-  // `DBImpl::DumpStats()` also holds db mutex
-  if (thread_dump_stats_ != nullptr) {
-    mutex_.Unlock();
-    thread_dump_stats_->cancel();
-    mutex_.Lock();
-    thread_dump_stats_.reset();
-  }
  if (!shutting_down_.load(std::memory_order_acquire) &&
      has_unpersisted_data_.load(std::memory_order_relaxed) &&
      !mutable_db_options_.avoid_flush_during_shutdown) {
-    if (immutable_db_options_.atomic_flush) {
-      autovector<ColumnFamilyData*> cfds;
-      SelectColumnFamiliesForAtomicFlush(&cfds);
-      mutex_.Unlock();
-      AtomicFlushMemTables(cfds, FlushOptions(), FlushReason::kShutDown);
-      mutex_.Lock();
-    } else {
-      for (auto cfd : *versions_->GetColumnFamilySet()) {
-        if (!cfd->IsDropped() && cfd->initialized() && !cfd->mem()->IsEmpty()) {
-          cfd->Ref();
-          mutex_.Unlock();
-          FlushMemTable(cfd, FlushOptions(), FlushReason::kShutDown);
-          mutex_.Lock();
-          cfd->Unref();
-        }
+    for (auto cfd : *versions_->GetColumnFamilySet()) {
+      if (!cfd->IsDropped() && cfd->initialized() && !cfd->mem()->IsEmpty()) {
+        cfd->Ref();
+        mutex_.Unlock();
+        FlushMemTable(cfd, FlushOptions(), FlushReason::kShutDown);
+        mutex_.Lock();
+        cfd->Unref();
      }
    }
    versions_->GetColumnFamilySet()->FreeDeadColumnFamilies();
@@ -425,20 +313,14 @@ void DBImpl::CancelAllBackgroundWork(bool wait) {
  if (!wait) {
    return;
  }
-  WaitForBackgroundWork();
+  // Wait for background work to finish
+  while (bg_bottom_compaction_scheduled_ || bg_compaction_scheduled_ ||
+         bg_flush_scheduled_) {
+    bg_cv_.Wait();
+  }
 }

 Status DBImpl::CloseHelper() {
-  // Guarantee that there is no background error recovery in progress before
-  // continuing with the shutdown
-  mutex_.Lock();
-  shutdown_initiated_ = true;
-  error_handler_.CancelErrorRecovery();
-  while (error_handler_.IsRecoveryInProgress()) {
-    bg_cv_.Wait();
-  }
-  mutex_.Unlock();
-
  // 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))
@@ -456,8 +338,7 @@ Status DBImpl::CloseHelper() {
  // Wait for background work to finish
  while (bg_bottom_compaction_scheduled_ || bg_compaction_scheduled_ ||
         bg_flush_scheduled_ || bg_purge_scheduled_ ||
-         pending_purge_obsolete_files_ ||
-         error_handler_.IsRecoveryInProgress()) {
+         pending_purge_obsolete_files_) {
    TEST_SYNC_POINT("DBImpl::~DBImpl:WaitJob");
    bg_cv_.Wait();
  }
@@ -467,12 +348,9 @@ Status DBImpl::CloseHelper() {
  flush_scheduler_.Clear();

  while (!flush_queue_.empty()) {
-    const FlushRequest& flush_req = PopFirstFromFlushQueue();
-    for (const auto& iter : flush_req) {
-      ColumnFamilyData* cfd = iter.first;
-      if (cfd->Unref()) {
-        delete cfd;
-      }
+    auto cfd = PopFirstFromFlushQueue();
+    if (cfd->Unref()) {
+      delete cfd;
    }
  }
  while (!compaction_queue_.empty()) {
@@ -562,17 +440,6 @@ Status DBImpl::CloseHelper() {
  ROCKS_LOG_INFO(immutable_db_options_.info_log, "Shutdown complete");
  LogFlush(immutable_db_options_.info_log);

-#ifndef ROCKSDB_LITE
-  // If the sst_file_manager was allocated by us during DB::Open(), ccall
-  // Close() on it before closing the info_log. Otherwise, background thread
-  // in SstFileManagerImpl might try to log something
-  if (immutable_db_options_.sst_file_manager && own_sfm_) {
-    auto sfm = static_cast<SstFileManagerImpl*>(
-        immutable_db_options_.sst_file_manager.get());
-    sfm->Close();
-  }
-#endif // ROCKSDB_LITE
-
  if (immutable_db_options_.info_log && own_info_log_) {
    Status s = immutable_db_options_.info_log->Close();
    if (ret.ok()) {
@@ -618,68 +485,66 @@ void DBImpl::PrintStatistics() {
  }
 }

-void DBImpl::StartTimedTasks() {
-  unsigned int stats_dump_period_sec = 0;
-  {
-    InstrumentedMutexLock l(&mutex_);
-    stats_dump_period_sec = mutable_db_options_.stats_dump_period_sec;
-    if (stats_dump_period_sec > 0) {
-      if (!thread_dump_stats_) {
-        thread_dump_stats_.reset(new rocksdb::RepeatableThread(
-            [this]() { DBImpl::DumpStats(); }, "dump_st", env_,
-            stats_dump_period_sec * 1000000));
-      }
-    }
-  }
-}
+void DBImpl::MaybeDumpStats() {
+  mutex_.Lock();
+  unsigned int stats_dump_period_sec =
+      mutable_db_options_.stats_dump_period_sec;
+  mutex_.Unlock();
+  if (stats_dump_period_sec == 0) return;
+
+  const uint64_t now_micros = env_->NowMicros();
+
+  if (last_stats_dump_time_microsec_ + stats_dump_period_sec * 1000000 <=
+      now_micros) {
+    // Multiple threads could race in here simultaneously.
+    // However, the last one will update last_stats_dump_time_microsec_
+    // atomically. We could see more than one dump during one dump
+    // period in rare cases.
+    last_stats_dump_time_microsec_ = now_micros;

-void DBImpl::DumpStats() {
-  TEST_SYNC_POINT("DBImpl::DumpStats:1");
 #ifndef ROCKSDB_LITE
-  const DBPropertyInfo* cf_property_info =
-      GetPropertyInfo(DB::Properties::kCFStats);
-  assert(cf_property_info != nullptr);
-  const DBPropertyInfo* db_property_info =
-      GetPropertyInfo(DB::Properties::kDBStats);
-  assert(db_property_info != nullptr);
+    const DBPropertyInfo* cf_property_info =
+        GetPropertyInfo(DB::Properties::kCFStats);
+    assert(cf_property_info != nullptr);
+    const DBPropertyInfo* db_property_info =
+        GetPropertyInfo(DB::Properties::kDBStats);
+    assert(db_property_info != nullptr);

-  std::string stats;
-  if (shutdown_initiated_) {
-    return;
-  }
-  {
-    InstrumentedMutexLock l(&mutex_);
-    default_cf_internal_stats_->GetStringProperty(
-        *db_property_info, DB::Properties::kDBStats, &stats);
-    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      if (cfd->initialized()) {
-        cfd->internal_stats()->GetStringProperty(
-            *cf_property_info, DB::Properties::kCFStatsNoFileHistogram, &stats);
+    std::string stats;
+    {
+      InstrumentedMutexLock l(&mutex_);
+      default_cf_internal_stats_->GetStringProperty(
+          *db_property_info, DB::Properties::kDBStats, &stats);
+      for (auto cfd : *versions_->GetColumnFamilySet()) {
+        if (cfd->initialized()) {
+          cfd->internal_stats()->GetStringProperty(
+              *cf_property_info, DB::Properties::kCFStatsNoFileHistogram,
+              &stats);
+        }
+      }
+      for (auto cfd : *versions_->GetColumnFamilySet()) {
+        if (cfd->initialized()) {
+          cfd->internal_stats()->GetStringProperty(
+              *cf_property_info, DB::Properties::kCFFileHistogram, &stats);
+        }
      }
    }
-    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      if (cfd->initialized()) {
-        cfd->internal_stats()->GetStringProperty(
-            *cf_property_info, DB::Properties::kCFFileHistogram, &stats);
+    ROCKS_LOG_WARN(immutable_db_options_.info_log,
+                   "------- DUMPING STATS -------");
+    ROCKS_LOG_WARN(immutable_db_options_.info_log, "%s", stats.c_str());
+    if (immutable_db_options_.dump_malloc_stats) {
+      stats.clear();
+      DumpMallocStats(&stats);
+      if (!stats.empty()) {
+        ROCKS_LOG_WARN(immutable_db_options_.info_log,
+                       "------- Malloc STATS -------");
+        ROCKS_LOG_WARN(immutable_db_options_.info_log, "%s", stats.c_str());
      }
    }
-  }
-  TEST_SYNC_POINT("DBImpl::DumpStats:2");
-  ROCKS_LOG_WARN(immutable_db_options_.info_log,
-                 "------- DUMPING STATS -------");
-  ROCKS_LOG_WARN(immutable_db_options_.info_log, "%s", stats.c_str());
-  if (immutable_db_options_.dump_malloc_stats) {
-    stats.clear();
-    DumpMallocStats(&stats);
-    if (!stats.empty()) {
-      ROCKS_LOG_WARN(immutable_db_options_.info_log,
-                     "------- Malloc STATS -------");
-      ROCKS_LOG_WARN(immutable_db_options_.info_log, "%s", stats.c_str());
-    }
-  }
 #endif  // !ROCKSDB_LITE

-  PrintStatistics();
+    PrintStatistics();
+  }
 }

 void DBImpl::ScheduleBgLogWriterClose(JobContext* job_context) {
@@ -730,6 +595,7 @@ Status DBImpl::SetOptions(
  MutableCFOptions new_options;
  Status s;
  Status persist_options_status;
+  WriteThread::Writer w;
  SuperVersionContext sv_context(/* create_superversion */ true);
  {
    InstrumentedMutexLock l(&mutex_);
@@ -791,6 +657,7 @@ Status DBImpl::SetDBOptions(
  Status s;
  Status persist_options_status;
  bool wal_changed = false;
+  WriteThread::Writer w;
  WriteContext write_context;
  {
    InstrumentedMutexLock l(&mutex_);
@@ -803,22 +670,7 @@ Status DBImpl::SetDBOptions(
            new_options.max_background_compactions, Env::Priority::LOW);
        MaybeScheduleFlushOrCompaction();
      }
-      if (new_options.stats_dump_period_sec !=
-          mutable_db_options_.stats_dump_period_sec) {
-          if (thread_dump_stats_) {
-            mutex_.Unlock();
-            thread_dump_stats_->cancel();
-            mutex_.Lock();
-          }
-          if (new_options.stats_dump_period_sec > 0) {
-            thread_dump_stats_.reset(new rocksdb::RepeatableThread(
-                [this]() { DBImpl::DumpStats(); }, "dump_st", env_,
-                new_options.stats_dump_period_sec * 1000000));
-          }
-          else {
-            thread_dump_stats_.reset();
-          }
-      }
+
      write_controller_.set_max_delayed_write_rate(
          new_options.delayed_write_rate);
      table_cache_.get()->SetCapacity(new_options.max_open_files == -1
@@ -839,7 +691,6 @@ Status DBImpl::SetDBOptions(
          env_options_for_compaction_, immutable_db_options_);
      env_options_for_compaction_.compaction_readahead_size =
          mutable_db_options_.compaction_readahead_size;
-      WriteThread::Writer w;
      write_thread_.EnterUnbatched(&w, &mutex_);
      if (total_log_size_ > GetMaxTotalWalSize() || wal_changed) {
        Status purge_wal_status = SwitchWAL(&write_context);
@@ -1032,7 +883,7 @@ bool DBImpl::SetPreserveDeletesSequenceNumber(SequenceNumber seqnum) {
 }

 InternalIterator* DBImpl::NewInternalIterator(
-    Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
+    Arena* arena, RangeDelAggregator* range_del_agg,
    ColumnFamilyHandle* column_family) {
  ColumnFamilyData* cfd;
  if (column_family == nullptr) {
@@ -1046,8 +897,8 @@ InternalIterator* DBImpl::NewInternalIterator(
  SuperVersion* super_version = cfd->GetSuperVersion()->Ref();
  mutex_.Unlock();
  ReadOptions roptions;
-  return NewInternalIterator(roptions, cfd, super_version, arena, range_del_agg,
-                             sequence);
+  return NewInternalIterator(roptions, cfd, super_version, arena,
+                             range_del_agg);
 }

 void DBImpl::SchedulePurge() {
@@ -1149,12 +1000,10 @@ static void CleanupIteratorState(void* arg1, void* /*arg2*/) {
 }
 }  // namespace

-InternalIterator* DBImpl::NewInternalIterator(const ReadOptions& read_options,
-                                              ColumnFamilyData* cfd,
-                                              SuperVersion* super_version,
-                                              Arena* arena,
-                                              RangeDelAggregator* range_del_agg,
-                                              SequenceNumber sequence) {
+InternalIterator* DBImpl::NewInternalIterator(
+    const ReadOptions& read_options, ColumnFamilyData* cfd,
+    SuperVersion* super_version, Arena* arena,
+    RangeDelAggregator* range_del_agg) {
  InternalIterator* internal_iter;
  assert(arena != nullptr);
  assert(range_del_agg != nullptr);
@@ -1166,12 +1015,12 @@ InternalIterator* DBImpl::NewInternalIterator(const ReadOptions& read_options,
  // Collect iterator for mutable mem
  merge_iter_builder.AddIterator(
      super_version->mem->NewIterator(read_options, arena));
-  std::unique_ptr<FragmentedRangeTombstoneIterator> range_del_iter;
+  std::unique_ptr<InternalIterator> range_del_iter;
  Status s;
  if (!read_options.ignore_range_deletions) {
    range_del_iter.reset(
-        super_version->mem->NewRangeTombstoneIterator(read_options, sequence));
-    range_del_agg->AddTombstones(std::move(range_del_iter));
+        super_version->mem->NewRangeTombstoneIterator(read_options));
+    s = range_del_agg->AddTombstones(std::move(range_del_iter));
  }
  // Collect all needed child iterators for immutable memtables
  if (s.ok()) {
@@ -1198,7 +1047,7 @@ InternalIterator* DBImpl::NewInternalIterator(const ReadOptions& read_options,
  } else {
    CleanupSuperVersion(super_version);
  }
-  return NewErrorInternalIterator<Slice>(s, arena);
+  return NewErrorInternalIterator(s, arena);
 }

 ColumnFamilyHandle* DBImpl::DefaultColumnFamily() const {
@@ -1222,15 +1071,6 @@ Status DBImpl::GetImpl(const ReadOptions& read_options,
  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
  auto cfd = cfh->cfd();

-  if (tracer_) {
-    // TODO: This mutex should be removed later, to improve performance when
-    // tracing is enabled.
-    InstrumentedMutexLock lock(&trace_mutex_);
-    if (tracer_) {
-      tracer_->Get(column_family, key);
-    }
-  }
-
  // Acquire SuperVersion
  SuperVersion* sv = GetAndRefSuperVersion(cfd);

@@ -1272,7 +1112,7 @@ Status DBImpl::GetImpl(const ReadOptions& read_options,

  // Prepare to store a list of merge operations if merge occurs.
  MergeContext merge_context;
-  SequenceNumber max_covering_tombstone_seq = 0;
+  RangeDelAggregator range_del_agg(cfd->internal_comparator(), snapshot);

  Status s;
  // First look in the memtable, then in the immutable memtable (if any).
@@ -1286,14 +1126,13 @@ Status DBImpl::GetImpl(const ReadOptions& read_options,
  bool done = false;
  if (!skip_memtable) {
    if (sv->mem->Get(lkey, pinnable_val->GetSelf(), &s, &merge_context,
-                     &max_covering_tombstone_seq, read_options, callback,
-                     is_blob_index)) {
+                     &range_del_agg, read_options, callback, is_blob_index)) {
      done = true;
      pinnable_val->PinSelf();
      RecordTick(stats_, MEMTABLE_HIT);
    } else if ((s.ok() || s.IsMergeInProgress()) &&
               sv->imm->Get(lkey, pinnable_val->GetSelf(), &s, &merge_context,
-                            &max_covering_tombstone_seq, read_options, callback,
+                            &range_del_agg, read_options, callback,
                            is_blob_index)) {
      done = true;
      pinnable_val->PinSelf();
@@ -1307,8 +1146,8 @@ Status DBImpl::GetImpl(const ReadOptions& read_options,
  if (!done) {
    PERF_TIMER_GUARD(get_from_output_files_time);
    sv->current->Get(read_options, lkey, pinnable_val, &s, &merge_context,
-                     &max_covering_tombstone_seq, value_found, nullptr, nullptr,
-                     callback, is_blob_index);
+                     &range_del_agg, value_found, nullptr, nullptr, callback,
+                     is_blob_index);
    RecordTick(stats_, MEMTABLE_MISS);
  }

@@ -1390,7 +1229,8 @@ std::vector<Status> DBImpl::MultiGet(

    LookupKey lkey(keys[i], snapshot);
    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family[i]);
-    SequenceNumber max_covering_tombstone_seq = 0;
+    RangeDelAggregator range_del_agg(cfh->cfd()->internal_comparator(),
+                                     snapshot);
    auto mgd_iter = multiget_cf_data.find(cfh->cfd()->GetID());
    assert(mgd_iter != multiget_cf_data.end());
    auto mgd = mgd_iter->second;
@@ -1401,12 +1241,11 @@ std::vector<Status> DBImpl::MultiGet(
    bool done = false;
    if (!skip_memtable) {
      if (super_version->mem->Get(lkey, value, &s, &merge_context,
-                                  &max_covering_tombstone_seq, read_options)) {
+                                  &range_del_agg, read_options)) {
        done = true;
        RecordTick(stats_, MEMTABLE_HIT);
      } else if (super_version->imm->Get(lkey, value, &s, &merge_context,
-                                         &max_covering_tombstone_seq,
-                                         read_options)) {
+                                         &range_del_agg, read_options)) {
        done = true;
        RecordTick(stats_, MEMTABLE_HIT);
      }
@@ -1415,7 +1254,7 @@ std::vector<Status> DBImpl::MultiGet(
      PinnableSlice pinnable_val;
      PERF_TIMER_GUARD(get_from_output_files_time);
      super_version->current->Get(read_options, lkey, &pinnable_val, &s,
-                                  &merge_context, &max_covering_tombstone_seq);
+                                  &merge_context, &range_del_agg);
      value->assign(pinnable_val.data(), pinnable_val.size());
      RecordTick(stats_, MEMTABLE_MISS);
    }
@@ -1770,8 +1609,8 @@ if (read_options.tailing) {
    result = NewDBIterator(
        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
        cfd->user_comparator(), iter, kMaxSequenceNumber,
-        sv->mutable_cf_options.max_sequential_skip_in_iterations, read_callback,
-        this, cfd);
+        sv->mutable_cf_options.max_sequential_skip_in_iterations,
+        read_callback);
 #endif
  } else {
    // Note: no need to consider the special case of
@@ -1838,12 +1677,13 @@ ArenaWrappedDBIter* DBImpl::NewIteratorImpl(const ReadOptions& read_options,
  ArenaWrappedDBIter* db_iter = NewArenaWrappedDbIterator(
      env_, read_options, *cfd->ioptions(), sv->mutable_cf_options, snapshot,
      sv->mutable_cf_options.max_sequential_skip_in_iterations,
-      sv->version_number, read_callback, this, cfd, allow_blob,
-      ((read_options.snapshot != nullptr) ? false : allow_refresh));
+      sv->version_number, read_callback,
+      ((read_options.snapshot != nullptr) ? nullptr : this), cfd, allow_blob,
+      allow_refresh);

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

  return db_iter;
@@ -1876,7 +1716,7 @@ Status DBImpl::NewIterators(
          env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
          cfd->user_comparator(), iter, kMaxSequenceNumber,
          sv->mutable_cf_options.max_sequential_skip_in_iterations,
-          read_callback, this, cfd));
+          read_callback));
    }
 #endif
  } else {
@@ -2214,18 +2054,17 @@ ColumnFamilyHandle* DBImpl::GetColumnFamilyHandle(uint32_t column_family_id) {
 }

 // REQUIRED: mutex is NOT held.
-std::unique_ptr<ColumnFamilyHandle> DBImpl::GetColumnFamilyHandleUnlocked(
+ColumnFamilyHandle* DBImpl::GetColumnFamilyHandleUnlocked(
    uint32_t column_family_id) {
+  ColumnFamilyMemTables* cf_memtables = column_family_memtables_.get();
+
  InstrumentedMutexLock l(&mutex_);

-  auto* cfd =
-      versions_->GetColumnFamilySet()->GetColumnFamily(column_family_id);
-  if (cfd == nullptr) {
+  if (!cf_memtables->Seek(column_family_id)) {
    return nullptr;
  }

-  return std::unique_ptr<ColumnFamilyHandleImpl>(
-      new ColumnFamilyHandleImpl(cfd, this, &mutex_));
+  return cf_memtables->GetColumnFamilyHandle();
 }

 void DBImpl::GetApproximateMemTableStats(ColumnFamilyHandle* column_family,
@@ -2296,8 +2135,9 @@ void DBImpl::ReleaseFileNumberFromPendingOutputs(

 #ifndef ROCKSDB_LITE
 Status DBImpl::GetUpdatesSince(
-    SequenceNumber seq, std::unique_ptr<TransactionLogIterator>* iter,
+    SequenceNumber seq, unique_ptr<TransactionLogIterator>* iter,
    const TransactionLogIterator::ReadOptions& read_options) {
+
  RecordTick(stats_, GET_UPDATES_SINCE_CALLS);
  if (seq > versions_->LastSequence()) {
    return Status::NotFound("Requested sequence not yet written in the db");
@@ -2387,9 +2227,9 @@ Status DBImpl::DeleteFile(std::string name) {
    status = versions_->LogAndApply(cfd, *cfd->GetLatestMutableCFOptions(),
                                    &edit, &mutex_, directories_.GetDbDir());
    if (status.ok()) {
-      InstallSuperVersionAndScheduleWork(
-          cfd, &job_context.superversion_contexts[0],
-          *cfd->GetLatestMutableCFOptions(), FlushReason::kDeleteFiles);
+      InstallSuperVersionAndScheduleWork(cfd, &job_context.superversion_context,
+                                         *cfd->GetLatestMutableCFOptions(),
+                                         FlushReason::kDeleteFiles);
    }
    FindObsoleteFiles(&job_context, false);
  }  // lock released here
@@ -2472,9 +2312,9 @@ Status DBImpl::DeleteFilesInRanges(ColumnFamilyHandle* column_family,
    status = versions_->LogAndApply(cfd, *cfd->GetLatestMutableCFOptions(),
                                    &edit, &mutex_, directories_.GetDbDir());
    if (status.ok()) {
-      InstallSuperVersionAndScheduleWork(
-          cfd, &job_context.superversion_contexts[0],
-          *cfd->GetLatestMutableCFOptions(), FlushReason::kDeleteFiles);
+      InstallSuperVersionAndScheduleWork(cfd, &job_context.superversion_context,
+                                         *cfd->GetLatestMutableCFOptions(),
+                                         FlushReason::kDeleteFiles);
    }
    for (auto* deleted_file : deleted_files) {
      deleted_file->being_compacted = false;
@@ -2545,10 +2385,10 @@ Status DBImpl::CheckConsistency() {
 Status DBImpl::GetDbIdentity(std::string& identity) const {
  std::string idfilename = IdentityFileName(dbname_);
  const EnvOptions soptions;
-  std::unique_ptr<SequentialFileReader> id_file_reader;
+  unique_ptr<SequentialFileReader> id_file_reader;
  Status s;
  {
-    std::unique_ptr<SequentialFile> idfile;
+    unique_ptr<SequentialFile> idfile;
    s = env_->NewSequentialFile(idfilename, &idfile, soptions);
    if (!s.ok()) {
      return s;
@@ -2562,7 +2402,7 @@ Status DBImpl::GetDbIdentity(std::string& identity) const {
  if (!s.ok()) {
    return s;
  }
-  char* buffer = reinterpret_cast<char*>(alloca(static_cast<size_t>(file_size)));
+  char* buffer = reinterpret_cast<char*>(alloca(file_size));
  Slice id;
  s = id_file_reader->Read(static_cast<size_t>(file_size), &id, buffer);
  if (!s.ok()) {
@@ -2945,7 +2785,8 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
                                       bool* is_blob_index) {
  Status s;
  MergeContext merge_context;
-  SequenceNumber max_covering_tombstone_seq = 0;
+  RangeDelAggregator range_del_agg(sv->mem->GetInternalKeyComparator(),
+                                   kMaxSequenceNumber);

  ReadOptions read_options;
  SequenceNumber current_seq = versions_->LastSequence();
@@ -2955,8 +2796,8 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  *found_record_for_key = false;

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

  if (!(s.ok() || s.IsNotFound() || s.IsMergeInProgress())) {
    // unexpected error reading memtable.
@@ -2974,8 +2815,8 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  }

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

  if (!(s.ok() || s.IsNotFound() || s.IsMergeInProgress())) {
    // unexpected error reading memtable.
@@ -2993,9 +2834,8 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  }

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

  if (!(s.ok() || s.IsNotFound() || s.IsMergeInProgress())) {
    // unexpected error reading memtable.
@@ -3018,7 +2858,7 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
  if (!cache_only) {
    // Check tables
    sv->current->Get(read_options, lkey, nullptr, &s, &merge_context,
-                     &max_covering_tombstone_seq, nullptr /* value_found */,
+                     &range_del_agg, nullptr /* value_found */,
                     found_record_for_key, seq, nullptr /*read_callback*/,
                     is_blob_index);

@@ -3027,20 +2867,18 @@ Status DBImpl::GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
      ROCKS_LOG_ERROR(immutable_db_options_.info_log,
                      "Unexpected status returned from Version::Get: %s\n",
                      s.ToString().c_str());
+
+      return s;
    }
  }

-  return s;
+  return Status::OK();
 }

 Status DBImpl::IngestExternalFile(
    ColumnFamilyHandle* column_family,
    const std::vector<std::string>& external_files,
    const IngestExternalFileOptions& ingestion_options) {
-  if (external_files.empty()) {
-    return Status::InvalidArgument("external_files is empty");
-  }
-
  Status status;
  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
  auto cfd = cfh->cfd();
@@ -3058,9 +2896,6 @@ Status DBImpl::IngestExternalFile(
                                            immutable_db_options_, env_options_,
                                            &snapshots_, ingestion_options);

-  SuperVersionContext dummy_sv_ctx(/* create_superversion */ true);
-  VersionEdit dummy_edit;
-  uint64_t next_file_number = 0;
  std::list<uint64_t>::iterator pending_output_elem;
  {
    InstrumentedMutexLock l(&mutex_);
@@ -3071,29 +2906,10 @@ Status DBImpl::IngestExternalFile(

    // Make sure that bg cleanup wont delete the files that we are ingesting
    pending_output_elem = CaptureCurrentFileNumberInPendingOutputs();
-
-    // If crash happen after a hard link established, Recover function may
-    // reuse the file number that has already assigned to the internal file,
-    // and this will overwrite the external file. To protect the external
-    // file, we have to make sure the file number will never being reused.
-    next_file_number = versions_->FetchAddFileNumber(external_files.size());
-    auto cf_options = cfd->GetLatestMutableCFOptions();
-    status = versions_->LogAndApply(cfd, *cf_options, &dummy_edit, &mutex_,
-                                    directories_.GetDbDir());
-    if (status.ok()) {
-      InstallSuperVersionAndScheduleWork(cfd, &dummy_sv_ctx, *cf_options);
-    }
-  }
-  dummy_sv_ctx.Clean();
-  if (!status.ok()) {
-    InstrumentedMutexLock l(&mutex_);
-    ReleaseFileNumberFromPendingOutputs(pending_output_elem);
-    return status;
  }

  SuperVersion* super_version = cfd->GetReferencedSuperVersion(&mutex_);
-  status =
-      ingestion_job.Prepare(external_files, next_file_number, super_version);
+  status = ingestion_job.Prepare(external_files, super_version);
  CleanupSuperVersion(super_version);
  if (!status.ok()) {
    InstrumentedMutexLock l(&mutex_);
@@ -3117,7 +2933,6 @@ Status DBImpl::IngestExternalFile(
    }

    num_running_ingest_file_++;
-    TEST_SYNC_POINT("DBImpl::IngestExternalFile:AfterIncIngestFileCounter");

    // We cannot ingest a file into a dropped CF
    if (cfd->IsDropped()) {
@@ -3132,21 +2947,10 @@ Status DBImpl::IngestExternalFile(
      TEST_SYNC_POINT_CALLBACK("DBImpl::IngestExternalFile:NeedFlush",
                               &need_flush);
      if (status.ok() && need_flush) {
-        FlushOptions flush_opts;
-        flush_opts.allow_write_stall = true;
-        if (immutable_db_options_.atomic_flush) {
-          autovector<ColumnFamilyData*> cfds;
-          SelectColumnFamiliesForAtomicFlush(&cfds);
-          mutex_.Unlock();
-          status = AtomicFlushMemTables(cfds, flush_opts,
-                                        FlushReason::kExternalFileIngestion,
-                                        true /* writes_stopped */);
-        } else {
-          mutex_.Unlock();
-          status = FlushMemTable(cfd, flush_opts,
-                                 FlushReason::kExternalFileIngestion,
-                                 true /* writes_stopped */);
-        }
+        mutex_.Unlock();
+        status = FlushMemTable(cfd, FlushOptions(),
+                               FlushReason::kExternalFileIngestion,
+                               true /* writes_stopped */);
        mutex_.Lock();
      }
    }
@@ -3256,6 +3060,7 @@ Status DBImpl::VerifyChecksum() {

 void DBImpl::NotifyOnExternalFileIngested(
    ColumnFamilyData* cfd, const ExternalSstFileIngestionJob& ingestion_job) {
+#ifndef ROCKSDB_LITE
  if (immutable_db_options_.listeners.empty()) {
    return;
  }
@@ -3271,6 +3076,8 @@ void DBImpl::NotifyOnExternalFileIngested(
      listener->OnExternalFileIngested(this, info);
    }
  }
+
+#endif
 }

 void DBImpl::WaitForIngestFile() {
@@ -3280,43 +3087,5 @@ 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)));
-  return Status::OK();
-}
-
-Status DBImpl::EndTrace() {
-  InstrumentedMutexLock lock(&trace_mutex_);
-  Status s = tracer_->Close();
-  tracer_.reset();
-  return s;
-}
-
-Status DBImpl::TraceIteratorSeek(const uint32_t& cf_id, const Slice& key) {
-  Status s;
-  if (tracer_) {
-    InstrumentedMutexLock lock(&trace_mutex_);
-    if (tracer_) {
-      s = tracer_->IteratorSeek(cf_id, key);
-    }
-  }
-  return s;
-}
-
-Status DBImpl::TraceIteratorSeekForPrev(const uint32_t& cf_id,
-                                        const Slice& key) {
-  Status s;
-  if (tracer_) {
-    InstrumentedMutexLock lock(&trace_mutex_);
-    if (tracer_) {
-      s = tracer_->IteratorSeekForPrev(cf_id, key);
-    }
-  }
-  return s;
-}
-
 #endif  // ROCKSDB_LITE
-
 }  // namespace rocksdb
@@ -22,16 +22,15 @@
 #include "db/column_family.h"
 #include "db/compaction_job.h"
 #include "db/dbformat.h"
+#include "db/external_sst_file_ingestion_job.h"
 #include "db/error_handler.h"
 #include "db/event_helpers.h"
-#include "db/external_sst_file_ingestion_job.h"
 #include "db/flush_job.h"
 #include "db/flush_scheduler.h"
 #include "db/internal_stats.h"
 #include "db/log_writer.h"
 #include "db/logs_with_prep_tracker.h"
 #include "db/pre_release_callback.h"
-#include "db/range_del_aggregator.h"
 #include "db/read_callback.h"
 #include "db/snapshot_checker.h"
 #include "db/snapshot_impl.h"
@@ -47,17 +46,14 @@
 #include "rocksdb/env.h"
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/status.h"
-#include "rocksdb/trace_reader_writer.h"
 #include "rocksdb/transaction_log.h"
 #include "rocksdb/write_buffer_manager.h"
 #include "table/scoped_arena_iterator.h"
 #include "util/autovector.h"
 #include "util/event_logger.h"
 #include "util/hash.h"
-#include "util/repeatable_thread.h"
 #include "util/stop_watch.h"
 #include "util/thread_local.h"
-#include "util/trace_replay.h"

 namespace rocksdb {

@@ -229,9 +225,6 @@ class DBImpl : public DB {
  using DB::Flush;
  virtual Status Flush(const FlushOptions& options,
                       ColumnFamilyHandle* column_family) override;
-  virtual Status Flush(
-      const FlushOptions& options,
-      const std::vector<ColumnFamilyHandle*>& column_families) override;
  virtual Status FlushWAL(bool sync) override;
  bool TEST_WALBufferIsEmpty();
  virtual Status SyncWAL() override;
@@ -260,9 +253,9 @@ class DBImpl : public DB {
  virtual Status GetSortedWalFiles(VectorLogPtr& files) override;

  virtual Status GetUpdatesSince(
-      SequenceNumber seq_number, std::unique_ptr<TransactionLogIterator>* iter,
-      const TransactionLogIterator::ReadOptions& read_options =
-          TransactionLogIterator::ReadOptions()) override;
+      SequenceNumber seq_number, unique_ptr<TransactionLogIterator>* iter,
+      const TransactionLogIterator::ReadOptions&
+          read_options = TransactionLogIterator::ReadOptions()) override;
  virtual Status DeleteFile(std::string name) override;
  Status DeleteFilesInRanges(ColumnFamilyHandle* column_family,
                             const RangePtr* ranges, size_t n,
@@ -340,15 +333,6 @@ class DBImpl : public DB {

  virtual Status VerifyChecksum() override;

-  using DB::StartTrace;
-  virtual Status StartTrace(
-      const TraceOptions& options,
-      std::unique_ptr<TraceWriter>&& trace_writer) override;
-
-  using DB::EndTrace;
-  virtual Status EndTrace() override;
-  Status TraceIteratorSeek(const uint32_t& cf_id, const Slice& key);
-  Status TraceIteratorSeekForPrev(const uint32_t& cf_id, const Slice& key);
 #endif  // ROCKSDB_LITE

  // Similar to GetSnapshot(), but also lets the db know that this snapshot
@@ -374,7 +358,7 @@ class DBImpl : public DB {
  // The keys of this iterator are internal keys (see format.h).
  // The returned iterator should be deleted when no longer needed.
  InternalIterator* NewInternalIterator(
-      Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence,
+      Arena* arena, RangeDelAggregator* range_del_agg,
      ColumnFamilyHandle* column_family = nullptr);

  LogsWithPrepTracker* logs_with_prep_tracker() {
@@ -401,7 +385,7 @@ class DBImpl : public DB {
  Status TEST_SwitchMemtable(ColumnFamilyData* cfd = nullptr);

  // Force current memtable contents to be flushed.
-  Status TEST_FlushMemTable(bool wait = true, bool allow_write_stall = false,
+  Status TEST_FlushMemTable(bool wait = true,
                            ColumnFamilyHandle* cfh = nullptr);

  // Wait for memtable compaction
@@ -469,8 +453,6 @@ class DBImpl : public DB {

  int TEST_BGCompactionsAllowed() const;
  int TEST_BGFlushesAllowed() const;
-  size_t TEST_GetWalPreallocateBlockSize(uint64_t write_buffer_size) const;
-  void TEST_WaitForTimedTaskRun(std::function<void()> callback) const;

 #endif  // NDEBUG

@@ -492,14 +474,6 @@ class DBImpl : public DB {

  uint64_t MinLogNumberToKeep();

-  // Returns the lower bound file number for SSTs that won't be deleted, even if
-  // they're obsolete. This lower bound is used internally to prevent newly
-  // created flush/compaction output files from being deleted before they're
-  // installed. This technique avoids the need for tracking the exact numbers of
-  // files pending creation, although it prevents more files than necessary from
-  // being deleted.
-  uint64_t MinObsoleteSstNumberToKeep();
-
  // Returns the list of live files in 'live' and the list
  // of all files in the filesystem in 'candidate_files'.
  // If force == false and the last call was less than
@@ -559,8 +533,7 @@ class DBImpl : public DB {
  ColumnFamilyHandle* GetColumnFamilyHandle(uint32_t column_family_id);

  // Same as above, should called without mutex held and not on write thread.
-  std::unique_ptr<ColumnFamilyHandle> GetColumnFamilyHandleUnlocked(
-      uint32_t column_family_id);
+  ColumnFamilyHandle* GetColumnFamilyHandleUnlocked(uint32_t column_family_id);

  // Returns the number of currently running flushes.
  // REQUIREMENT: mutex_ must be held when calling this function.
@@ -578,9 +551,11 @@ class DBImpl : public DB {

  const WriteController& write_controller() { return write_controller_; }

-  InternalIterator* NewInternalIterator(
-      const ReadOptions&, ColumnFamilyData* cfd, SuperVersion* super_version,
-      Arena* arena, RangeDelAggregator* range_del_agg, SequenceNumber sequence);
+  InternalIterator* NewInternalIterator(const ReadOptions&,
+                                        ColumnFamilyData* cfd,
+                                        SuperVersion* super_version,
+                                        Arena* arena,
+                                        RangeDelAggregator* range_del_agg);

  // hollow transactions shell used for recovery.
  // these will then be passed to TransactionDB so that
@@ -713,7 +688,7 @@ class DBImpl : public DB {
 protected:
  Env* const env_;
  const std::string dbname_;
-  std::unique_ptr<VersionSet> versions_;
+  unique_ptr<VersionSet> versions_;
  // Flag to check whether we allocated and own the info log file
  bool own_info_log_;
  const DBOptions initial_db_options_;
@@ -722,8 +697,6 @@ class DBImpl : public DB {
  Statistics* stats_;
  std::unordered_map<std::string, RecoveredTransaction*>
      recovered_transactions_;
-  std::unique_ptr<Tracer> tracer_;
-  InstrumentedMutex trace_mutex_;

  // Except in DB::Open(), WriteOptionsFile can only be called when:
  // Persist options to options file.
@@ -745,11 +718,6 @@ class DBImpl : public DB {
                              const MutableCFOptions& mutable_cf_options,
                              int job_id, TableProperties prop);

-  void NotifyOnCompactionBegin(ColumnFamilyData* cfd,
-                               Compaction *c, const Status &st,
-                               const CompactionJobStats& job_stats,
-                               int job_id);
-
  void NotifyOnCompactionCompleted(ColumnFamilyData* cfd,
                                   Compaction *c, const Status &st,
                                   const CompactionJobStats& job_stats,
@@ -814,7 +782,6 @@ class DBImpl : public DB {

 private:
  friend class DB;
-  friend class ErrorHandler;
  friend class InternalStats;
  friend class PessimisticTransaction;
  friend class TransactionBaseImpl;
@@ -823,21 +790,18 @@ class DBImpl : public DB {
  friend class WritePreparedTxnDB;
  friend class WriteBatchWithIndex;
  friend class WriteUnpreparedTxnDB;
-  friend class WriteUnpreparedTxn;
-
 #ifndef ROCKSDB_LITE
  friend class ForwardIterator;
 #endif
  friend struct SuperVersion;
  friend class CompactedDBImpl;
  friend class DBTest_ConcurrentFlushWAL_Test;
-  friend class DBTest_MixedSlowdownOptionsStop_Test;
 #ifndef NDEBUG
  friend class DBTest2_ReadCallbackTest_Test;
  friend class WriteCallbackTest_WriteWithCallbackTest_Test;
  friend class XFTransactionWriteHandler;
  friend class DBBlobIndexTest;
-  friend class WriteUnpreparedTransactionTest_RecoveryTest_Test;
+  friend class WriteUnpreparedTransactionTest_RecoveryRollbackUnprepared_Test;
 #endif
  struct CompactionState;

@@ -866,8 +830,6 @@ class DBImpl : public DB {
                 bool read_only = false, bool error_if_log_file_exist = false,
                 bool error_if_data_exists_in_logs = false);

-  Status ResumeImpl();
-
  void MaybeIgnoreError(Status* s) const;

  const Status CreateArchivalDirectory();
@@ -907,45 +869,12 @@ class DBImpl : public DB {
  Status SyncClosedLogs(JobContext* job_context);

  // Flush the in-memory write buffer to storage.  Switches to a new
-  // log-file/memtable and writes a new descriptor iff successful. Then
-  // installs a new super version for the column family.
+  // log-file/memtable and writes a new descriptor iff successful.
  Status FlushMemTableToOutputFile(ColumnFamilyData* cfd,
                                   const MutableCFOptions& mutable_cf_options,
                                   bool* madeProgress, JobContext* job_context,
-                                   SuperVersionContext* superversion_context,
                                   LogBuffer* log_buffer);

-  // Argument required by background flush thread.
-  struct BGFlushArg {
-    BGFlushArg()
-        : cfd_(nullptr), max_memtable_id_(0), superversion_context_(nullptr) {}
-    BGFlushArg(ColumnFamilyData* cfd, uint64_t max_memtable_id,
-               SuperVersionContext* superversion_context)
-        : cfd_(cfd),
-          max_memtable_id_(max_memtable_id),
-          superversion_context_(superversion_context) {}
-
-    // Column family to flush.
-    ColumnFamilyData* cfd_;
-    // Maximum ID of memtable to flush. In this column family, memtables with
-    // IDs smaller than this value must be flushed before this flush completes.
-    uint64_t max_memtable_id_;
-    // Pointer to a SuperVersionContext object. After flush completes, RocksDB
-    // installs a new superversion for the column family. This operation
-    // requires a SuperVersionContext object (currently embedded in JobContext).
-    SuperVersionContext* superversion_context_;
-  };
-
-  // Flush the memtables of (multiple) column families to multiple files on
-  // persistent storage.
-  Status FlushMemTablesToOutputFiles(
-      const autovector<BGFlushArg>& bg_flush_args, bool* made_progress,
-      JobContext* job_context, LogBuffer* log_buffer);
-
-  Status AtomicFlushMemTablesToOutputFiles(
-      const autovector<BGFlushArg>& bg_flush_args, bool* made_progress,
-      JobContext* job_context, LogBuffer* log_buffer);
-
  // REQUIRES: log_numbers are sorted in ascending order
  Status RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
                         SequenceNumber* next_sequence, bool read_only);
@@ -958,12 +887,6 @@ class DBImpl : public DB {
  Status WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
                                     MemTable* mem, VersionEdit* edit);

-  // Restore alive_log_files_ and total_log_size_ after recovery.
-  // It needs to run only when there's no flush during recovery
-  // (e.g. avoid_flush_during_recovery=true). May also trigger flush
-  // in case total_log_size > max_total_wal_size.
-  Status RestoreAliveLogFiles(const std::vector<uint64_t>& log_numbers);
-
  // num_bytes: for slowdown case, delay time is calculated based on
  //            `num_bytes` going through.
  Status DelayWrite(uint64_t num_bytes, const WriteOptions& write_options);
@@ -973,43 +896,19 @@ class DBImpl : public DB {

  Status ScheduleFlushes(WriteContext* context);

-  Status SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context);
-
-  void SelectColumnFamiliesForAtomicFlush(autovector<ColumnFamilyData*>* cfds);
+  Status SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context,
+                        FlushReason flush_reason = FlushReason::kOthers);

  // Force current memtable contents to be flushed.
  Status FlushMemTable(ColumnFamilyData* cfd, const FlushOptions& options,
                       FlushReason flush_reason, bool writes_stopped = false);

-  Status AtomicFlushMemTables(
-      const autovector<ColumnFamilyData*>& column_family_datas,
-      const FlushOptions& options, FlushReason flush_reason,
-      bool writes_stopped = false);
-
-  // Wait until flushing this column family won't stall writes
-  Status WaitUntilFlushWouldNotStallWrites(ColumnFamilyData* cfd,
-                                           bool* flush_needed);
-
  // Wait for memtable flushed.
  // If flush_memtable_id is non-null, wait until the memtable with the ID
  // gets flush. Otherwise, wait until the column family don't have any
  // memtable pending flush.
-  // resuming_from_bg_err indicates whether the caller is attempting to resume
-  // from background error.
  Status WaitForFlushMemTable(ColumnFamilyData* cfd,
-                              const uint64_t* flush_memtable_id = nullptr,
-                              bool resuming_from_bg_err = false) {
-    return WaitForFlushMemTables({cfd}, {flush_memtable_id},
-                                 resuming_from_bg_err);
-  }
-  // Wait for memtables to be flushed for multiple column families.
-  Status WaitForFlushMemTables(
-      const autovector<ColumnFamilyData*>& cfds,
-      const autovector<const uint64_t*>& flush_memtable_ids,
-      bool resuming_from_bg_err);
-
-  // REQUIRES: mutex locked and in write thread.
-  void AssignAtomicFlushSeq(const autovector<ColumnFamilyData*>& cfds);
+                              const uint64_t* flush_memtable_id = nullptr);

  // REQUIRES: mutex locked
  Status SwitchWAL(WriteContext* write_context);
@@ -1065,20 +964,7 @@ class DBImpl : public DB {
  ColumnFamilyData* GetColumnFamilyDataByName(const std::string& cf_name);

  void MaybeScheduleFlushOrCompaction();
-
-  // A flush request specifies the column families to flush as well as the
-  // largest memtable id to persist for each column family. Once all the
-  // memtables whose IDs are smaller than or equal to this per-column-family
-  // specified value, this flush request is considered to have completed its
-  // work of flushing this column family. After completing the work for all
-  // column families in this request, this flush is considered complete.
-  typedef std::vector<std::pair<ColumnFamilyData*, uint64_t>> FlushRequest;
-
-  void GenerateFlushRequest(const autovector<ColumnFamilyData*>& cfds,
-                            FlushRequest* req);
-
-  void SchedulePendingFlush(const FlushRequest& req, FlushReason flush_reason);
-
+  void SchedulePendingFlush(ColumnFamilyData* cfd, FlushReason flush_reason);
  void SchedulePendingCompaction(ColumnFamilyData* cfd);
  void SchedulePendingPurge(std::string fname, std::string dir_to_sync,
                            FileType type, uint64_t number, int job_id);
@@ -1097,19 +983,15 @@ class DBImpl : public DB {
                              LogBuffer* log_buffer,
                              PrepickedCompaction* prepicked_compaction);
  Status BackgroundFlush(bool* madeProgress, JobContext* job_context,
-                         LogBuffer* log_buffer, FlushReason* reason);
+                         LogBuffer* log_buffer);

-  bool EnoughRoomForCompaction(ColumnFamilyData* cfd,
-                               const std::vector<CompactionInputFiles>& inputs,
+  bool EnoughRoomForCompaction(const std::vector<CompactionInputFiles>& inputs,
                               bool* sfm_bookkeeping, LogBuffer* log_buffer);

-  // Schedule background tasks
-  void StartTimedTasks();
-
  void PrintStatistics();

  // dump rocksdb.stats to LOG
-  void DumpStats();
+  void MaybeDumpStats();

  // Return the minimum empty level that could hold the total data in the
  // input level. Return the input level, if such level could not be found.
@@ -1124,7 +1006,8 @@ class DBImpl : public DB {
  // helper functions for adding and removing from flush & compaction queues
  void AddToCompactionQueue(ColumnFamilyData* cfd);
  ColumnFamilyData* PopFirstFromCompactionQueue();
-  FlushRequest PopFirstFromFlushQueue();
+  void AddToFlushQueue(ColumnFamilyData* cfd, FlushReason flush_reason);
+  ColumnFamilyData* PopFirstFromFlushQueue();

  // helper function to call after some of the logs_ were synced
  void MarkLogsSynced(uint64_t up_to, bool synced_dir, const Status& status);
@@ -1137,8 +1020,6 @@ class DBImpl : public DB {

  Status CloseHelper();

-  void WaitForBackgroundWork();
-
  // table_cache_ provides its own synchronization
  std::shared_ptr<Cache> table_cache_;

@@ -1189,7 +1070,7 @@ class DBImpl : public DB {
  bool log_empty_;
  ColumnFamilyHandleImpl* default_cf_handle_;
  InternalStats* default_cf_internal_stats_;
-  std::unique_ptr<ColumnFamilyMemTablesImpl> column_family_memtables_;
+  unique_ptr<ColumnFamilyMemTablesImpl> column_family_memtables_;
  struct LogFileNumberSize {
    explicit LogFileNumberSize(uint64_t _number)
        : number(_number) {}
@@ -1217,7 +1098,7 @@ class DBImpl : public DB {

    uint64_t number;
    // Visual Studio doesn't support deque's member to be noncopyable because
-    // of a std::unique_ptr as a member.
+    // of a unique_ptr as a member.
    log::Writer* writer;  // own
    // true for some prefix of logs_
    bool getting_synced = false;
@@ -1303,7 +1184,7 @@ class DBImpl : public DB {

  WriteController write_controller_;

-  std::unique_ptr<RateLimiter> low_pri_write_rate_limiter_;
+  unique_ptr<RateLimiter> low_pri_write_rate_limiter_;

  // Size of the last batch group. In slowdown mode, next write needs to
  // sleep if it uses up the quota.
@@ -1359,7 +1240,7 @@ class DBImpl : public DB {
  // in MaybeScheduleFlushOrCompaction()
  // invariant(column family present in flush_queue_ <==>
  // ColumnFamilyData::pending_flush_ == true)
-  std::deque<FlushRequest> flush_queue_;
+  std::deque<ColumnFamilyData*> flush_queue_;
  // invariant(column family present in compaction_queue_ <==>
  // ColumnFamilyData::pending_compaction_ == true)
  std::deque<ColumnFamilyData*> compaction_queue_;
@@ -1510,10 +1391,6 @@ class DBImpl : public DB {
  // Only to be set during initialization
  std::unique_ptr<PreReleaseCallback> recoverable_state_pre_release_callback_;

-  // handle for scheduling jobs at fixed intervals
-  // REQUIRES: mutex locked
-  std::unique_ptr<rocksdb::RepeatableThread> thread_dump_stats_;
-
  // No copying allowed
  DBImpl(const DBImpl&);
  void operator=(const DBImpl&);
@@ -1587,16 +1464,6 @@ class DBImpl : public DB {
  // flush/compaction and if it is not provided vis SnapshotChecker, we should
  // disable gc to be safe.
  const bool use_custom_gc_;
-  // Flag to indicate that the DB instance shutdown has been initiated. This
-  // different from shutting_down_ atomic in that it is set at the beginning
-  // of shutdown sequence, specifically in order to prevent any background
-  // error recovery from going on in parallel. The latter, shutting_down_,
-  // is set a little later during the shutdown after scheduling memtable
-  // flushes
-  std::atomic<bool> shutdown_initiated_;
-  // Flag to indicate whether sst_file_manager object was allocated in
-  // DB::Open() or passed to us
-  bool own_sfm_;

  // Clients must periodically call SetPreserveDeletesSequenceNumber()
  // to advance this seqnum. Default value is 0 which means ALL deletes are
@@ -1609,17 +1476,6 @@ class DBImpl : public DB {
  bool closed_;

  ErrorHandler error_handler_;
-
-  // Conditional variable to coordinate installation of atomic flush results.
-  // With atomic flush, each bg thread installs the result of flushing multiple
-  // column families, and different threads can flush different column
-  // families. It's difficult to rely on one thread to perform batch
-  // installation for all threads. This is different from the non-atomic flush
-  // case.
-  // atomic_flush_install_cv_ makes sure that threads install atomic flush
-  // results sequentially. Flush results of memtables with lower IDs get
-  // installed to MANIFEST first.
-  InstrumentedCondVar atomic_flush_install_cv_;
 };

 extern Options SanitizeOptions(const std::string& db,
@@ -100,11 +100,9 @@ Status DBImpl::TEST_SwitchMemtable(ColumnFamilyData* cfd) {
  return SwitchMemtable(cfd, &write_context);
 }

-Status DBImpl::TEST_FlushMemTable(bool wait, bool allow_write_stall,
-    ColumnFamilyHandle* cfh) {
+Status DBImpl::TEST_FlushMemTable(bool wait, ColumnFamilyHandle* cfh) {
  FlushOptions fo;
  fo.wait = wait;
-  fo.allow_write_stall = allow_write_stall;
  ColumnFamilyData* cfd;
  if (cfh == nullptr) {
    cfd = default_cf_handle_->cfd();
@@ -123,7 +121,7 @@ Status DBImpl::TEST_WaitForFlushMemTable(ColumnFamilyHandle* column_family) {
    auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
    cfd = cfh->cfd();
  }
-  return WaitForFlushMemTable(cfd, nullptr, false);
+  return WaitForFlushMemTable(cfd);
 }

 Status DBImpl::TEST_WaitForCompact(bool wait_unscheduled) {
@@ -137,7 +135,7 @@ Status DBImpl::TEST_WaitForCompact(bool wait_unscheduled) {
  while ((bg_bottom_compaction_scheduled_ || bg_compaction_scheduled_ ||
          bg_flush_scheduled_ ||
          (wait_unscheduled && unscheduled_compactions_)) &&
-         (error_handler_.GetBGError() == Status::OK())) {
+         !error_handler_.IsDBStopped()) {
    bg_cv_.Wait();
  }
  return error_handler_.GetBGError();
@@ -237,16 +235,5 @@ SequenceNumber DBImpl::TEST_GetLastVisibleSequence() const {
  }
 }

-size_t DBImpl::TEST_GetWalPreallocateBlockSize(
-    uint64_t write_buffer_size) const {
-  InstrumentedMutexLock l(&mutex_);
-  return GetWalPreallocateBlockSize(write_buffer_size);
-}
-
-void DBImpl::TEST_WaitForTimedTaskRun(std::function<void()> callback) const {
-  if (thread_dump_stats_ != nullptr) {
-    thread_dump_stats_->TEST_WaitForRun(callback);
-  }
-}
 }  // namespace rocksdb
 #endif  // NDEBUG
@@ -131,16 +131,16 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
      edit.DeleteFile(0, f->fd.GetNumber());
      edit.AddFile(target_level, f->fd.GetNumber(), f->fd.GetPathId(),
                   f->fd.GetFileSize(), f->smallest, f->largest,
-                   f->fd.smallest_seqno, f->fd.largest_seqno,
+                   f->smallest_seqno, f->largest_seqno,
                   f->marked_for_compaction);
    }

    status = versions_->LogAndApply(cfd, *cfd->GetLatestMutableCFOptions(),
                                    &edit, &mutex_, directories_.GetDbDir());
    if (status.ok()) {
-      InstallSuperVersionAndScheduleWork(cfd,
-                                         &job_context.superversion_contexts[0],
-                                         *cfd->GetLatestMutableCFOptions());
+      InstallSuperVersionAndScheduleWork(
+          cfd, &job_context.superversion_context,
+          *cfd->GetLatestMutableCFOptions());
    }
  }  // lock released here
  LogFlush(immutable_db_options_.info_log);
@@ -20,7 +20,6 @@
 #include "util/sst_file_manager_impl.h"

 namespace rocksdb {
-
 uint64_t DBImpl::MinLogNumberToKeep() {
  if (allow_2pc()) {
    return versions_->min_log_number_to_keep_2pc();
@@ -29,14 +28,6 @@ uint64_t DBImpl::MinLogNumberToKeep() {
  }
 }

-uint64_t DBImpl::MinObsoleteSstNumberToKeep() {
-  mutex_.AssertHeld();
-  if (!pending_outputs_.empty()) {
-    return *pending_outputs_.begin();
-  }
-  return std::numeric_limits<uint64_t>::max();
-}
-
 // * Returns the list of live files in 'sst_live'
 // If it's doing full scan:
 // * Returns the list of all files in the filesystem in
@@ -465,13 +456,7 @@ void DBImpl::PurgeObsoleteFiles(JobContext& state, bool schedule_only) {
    } else {
      dir_to_sync =
          (type == kLogFile) ? immutable_db_options_.wal_dir : dbname_;
-      fname = dir_to_sync
-            + (
-                (!dir_to_sync.empty() && dir_to_sync.back() == '/') ||
-                (!to_delete.empty() && to_delete.front() == '/')
-                ? "" : "/"
-              )
-            + to_delete;
+      fname = dir_to_sync + "/" + to_delete;
    }

 #ifndef ROCKSDB_LITE
@@ -23,7 +23,8 @@
 #include "util/sync_point.h"

 namespace rocksdb {
-Options SanitizeOptions(const std::string& dbname, const Options& src) {
+Options SanitizeOptions(const std::string& dbname,
+                        const Options& src) {
  auto db_options = SanitizeOptions(dbname, DBOptions(src));
  ImmutableDBOptions immutable_db_options(db_options);
  auto cf_options =
@@ -55,9 +56,10 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
    result.write_buffer_manager.reset(
        new WriteBufferManager(result.db_write_buffer_size));
  }
-  auto bg_job_limits = DBImpl::GetBGJobLimits(
-      result.max_background_flushes, result.max_background_compactions,
-      result.max_background_jobs, true /* parallelize_compactions */);
+  auto bg_job_limits = DBImpl::GetBGJobLimits(result.max_background_flushes,
+                                              result.max_background_compactions,
+                                              result.max_background_jobs,
+                                              true /* parallelize_compactions */);
  result.env->IncBackgroundThreadsIfNeeded(bg_job_limits.max_compactions,
                                           Env::Priority::LOW);
  result.env->IncBackgroundThreadsIfNeeded(bg_job_limits.max_flushes,
@@ -105,12 +107,14 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
    result.db_paths.emplace_back(dbname, std::numeric_limits<uint64_t>::max());
  }

-  if (result.use_direct_reads && result.compaction_readahead_size == 0) {
+  if (result.use_direct_reads &&
+      result.compaction_readahead_size == 0) {
    TEST_SYNC_POINT_CALLBACK("SanitizeOptions:direct_io", nullptr);
    result.compaction_readahead_size = 1024 * 1024 * 2;
  }

-  if (result.compaction_readahead_size > 0 || result.use_direct_reads) {
+  if (result.compaction_readahead_size > 0 ||
+      result.use_direct_reads) {
    result.new_table_reader_for_compaction_inputs = true;
  }

@@ -130,15 +134,8 @@ DBOptions SanitizeOptions(const std::string& dbname, const DBOptions& src) {
  for (size_t i = 0; i < result.db_paths.size(); i++) {
    DeleteScheduler::CleanupDirectory(result.env, sfm, result.db_paths[i].path);
  }
-
-  // Create a default SstFileManager for purposes of tracking compaction size
-  // and facilitating recovery from out of space errors.
-  if (result.sst_file_manager.get() == nullptr) {
-    std::shared_ptr<SstFileManager> sst_file_manager(
-        NewSstFileManager(result.env, result.info_log));
-    result.sst_file_manager = sst_file_manager;
-  }
 #endif
+
  return result;
 }

@@ -214,7 +211,7 @@ static Status ValidateOptions(

  return Status::OK();
 }
-}  // namespace
+} // namespace
 Status DBImpl::NewDB() {
  VersionEdit new_db;
  new_db.SetLogNumber(0);
@@ -226,7 +223,7 @@ Status DBImpl::NewDB() {
  ROCKS_LOG_INFO(immutable_db_options_.info_log, "Creating manifest 1 \n");
  const std::string manifest = DescriptorFileName(dbname_, 1);
  {
-    std::unique_ptr<WritableFile> file;
+    unique_ptr<WritableFile> file;
    EnvOptions env_options = env_->OptimizeForManifestWrite(env_options_);
    s = NewWritableFile(env_, manifest, &file, env_options);
    if (!s.ok()) {
@@ -234,9 +231,8 @@ Status DBImpl::NewDB() {
    }
    file->SetPreallocationBlockSize(
        immutable_db_options_.manifest_preallocation_size);
-    std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-        std::move(file), manifest, env_options, nullptr /* stats */,
-        immutable_db_options_.listeners));
+    unique_ptr<WritableFileWriter> file_writer(
+        new WritableFileWriter(std::move(file), env_options));
    log::Writer log(std::move(file_writer), 0, false);
    std::string record;
    new_db.EncodeTo(&record);
@@ -254,8 +250,9 @@ Status DBImpl::NewDB() {
  return s;
 }

-Status DBImpl::CreateAndNewDirectory(Env* env, const std::string& dirname,
-                                     std::unique_ptr<Directory>* directory) {
+Status DBImpl::CreateAndNewDirectory(
+    Env* env, const std::string& dirname,
+    std::unique_ptr<Directory>* directory) {
  // We call CreateDirIfMissing() as the directory may already exist (if we
  // are reopening a DB), when this happens we don't want creating the
  // directory to cause an error. However, we need to check if creating the
@@ -336,8 +333,8 @@ Status DBImpl::Recover(
      }
    } else if (s.ok()) {
      if (immutable_db_options_.error_if_exists) {
-        return Status::InvalidArgument(dbname_,
-                                       "exists (error_if_exists is true)");
+        return Status::InvalidArgument(
+            dbname_, "exists (error_if_exists is true)");
      }
    } else {
      // Unexpected error reading file
@@ -357,14 +354,14 @@ Status DBImpl::Recover(
    }
    // Verify compatibility of env_options_ and filesystem
    {
-      std::unique_ptr<RandomAccessFile> idfile;
+      unique_ptr<RandomAccessFile> idfile;
      EnvOptions customized_env(env_options_);
      customized_env.use_direct_reads |=
          immutable_db_options_.use_direct_io_for_flush_and_compaction;
      s = env_->NewRandomAccessFile(IdentityFileName(dbname_), &idfile,
                                    customized_env);
      if (!s.ok()) {
-        std::string error_str = s.ToString();
+        const char* error_msg = s.ToString().c_str();
        // Check if unsupported Direct I/O is the root cause
        customized_env.use_direct_reads = false;
        s = env_->NewRandomAccessFile(IdentityFileName(dbname_), &idfile,
@@ -374,7 +371,7 @@ Status DBImpl::Recover(
              "Direct I/O is not supported by the specified DB.");
        } else {
          return Status::InvalidArgument(
-              "Found options incompatible with filesystem", error_str.c_str());
+              "Found options incompatible with filesystem", error_msg);
        }
      }
    }
@@ -392,16 +389,6 @@ Status DBImpl::Recover(
      }
    }
  }
-
-  // Initial max_total_in_memory_state_ before recovery logs. Log recovery
-  // may check this value to decide whether to flush.
-  max_total_in_memory_state_ = 0;
-  for (auto cfd : *versions_->GetColumnFamilySet()) {
-    auto* mutable_cf_options = cfd->GetLatestMutableCFOptions();
-    max_total_in_memory_state_ += mutable_cf_options->write_buffer_size *
-                                  mutable_cf_options->max_write_buffer_number;
-  }
-
  if (s.ok()) {
    SequenceNumber next_sequence(kMaxSequenceNumber);
    default_cf_handle_ = new ColumnFamilyHandleImpl(
@@ -474,26 +461,12 @@ Status DBImpl::Recover(
    }
  }

-  if (read_only) {
-    // If we are opening as read-only, we need to update options_file_number_
-    // to reflect the most recent OPTIONS file. It does not matter for regular
-    // read-write db instance because options_file_number_ will later be
-    // updated to versions_->NewFileNumber() in RenameTempFileToOptionsFile.
-    std::vector<std::string> file_names;
-    if (s.ok()) {
-      s = env_->GetChildren(GetName(), &file_names);
-    }
-    if (s.ok()) {
-      uint64_t number = 0;
-      uint64_t options_file_number = 0;
-      FileType type;
-      for (const auto& fname : file_names) {
-        if (ParseFileName(fname, &number, &type) && type == kOptionsFile) {
-          options_file_number = std::max(number, options_file_number);
-        }
-      }
-      versions_->options_file_number_ = options_file_number;
-    }
+  // Initial value
+  max_total_in_memory_state_ = 0;
+  for (auto cfd : *versions_->GetColumnFamilySet()) {
+    auto* mutable_cf_options = cfd->GetLatestMutableCFOptions();
+    max_total_in_memory_state_ += mutable_cf_options->write_buffer_size *
+                                  mutable_cf_options->max_write_buffer_number;
  }

  return s;
@@ -544,9 +517,10 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
    std::map<std::string, uint32_t> cf_name_id_map;
    std::map<uint32_t, uint64_t> cf_lognumber_map;
    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      cf_name_id_map.insert(std::make_pair(cfd->GetName(), cfd->GetID()));
+      cf_name_id_map.insert(
+        std::make_pair(cfd->GetName(), cfd->GetID()));
      cf_lognumber_map.insert(
-          std::make_pair(cfd->GetID(), cfd->GetLogNumber()));
+        std::make_pair(cfd->GetID(), cfd->GetLogNumber()));
    }

    immutable_db_options_.wal_filter->ColumnFamilyLogNumberMap(cf_lognumber_map,
@@ -589,9 +563,9 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
      continue;
    }

-    std::unique_ptr<SequentialFileReader> file_reader;
+    unique_ptr<SequentialFileReader> file_reader;
    {
-      std::unique_ptr<SequentialFile> file;
+      unique_ptr<SequentialFile> file;
      status = env_->NewSequentialFile(fname, &file,
                                       env_->OptimizeForLogRead(env_options_));
      if (!status.ok()) {
@@ -624,8 +598,8 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
    // to be skipped instead of propagating bad information (like overly
    // large sequence numbers).
    log::Reader reader(immutable_db_options_.info_log, std::move(file_reader),
-                       &reporter, true /*checksum*/, log_number,
-                       false /* retry_after_eof */);
+                       &reporter, true /*checksum*/, 0 /*initial_offset*/,
+                       log_number);

    // Determine if we should tolerate incomplete records at the tail end of the
    // Read all the records and add to a memtable
@@ -896,8 +870,8 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
      // VersionSet::next_file_number_ always to be strictly greater than any
      // log number
      versions_->MarkFileNumberUsed(max_log_number + 1);
-      status = versions_->LogAndApply(cfd, *cfd->GetLatestMutableCFOptions(),
-                                      edit, &mutex_);
+      status = versions_->LogAndApply(
+          cfd, *cfd->GetLatestMutableCFOptions(), edit, &mutex_);
      if (!status.ok()) {
        // Recovery failed
        break;
@@ -905,8 +879,18 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
    }
  }

-  if (status.ok() && data_seen && !flushed) {
-    status = RestoreAliveLogFiles(log_numbers);
+  if (data_seen && !flushed) {
+    // Mark these as alive so they'll be considered for deletion later by
+    // FindObsoleteFiles()
+    if (two_write_queues_) {
+      log_write_mutex_.Lock();
+    }
+    for (auto log_number : log_numbers) {
+      alive_log_files_.push_back(LogFileNumberSize(log_number));
+    }
+    if (two_write_queues_) {
+      log_write_mutex_.Unlock();
+    }
  }

  event_logger_.Log() << "job" << job_id << "event"
@@ -915,60 +899,6 @@ Status DBImpl::RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
  return status;
 }

-Status DBImpl::RestoreAliveLogFiles(const std::vector<uint64_t>& log_numbers) {
-  if (log_numbers.empty()) {
-    return Status::OK();
-  }
-  Status s;
-  mutex_.AssertHeld();
-  assert(immutable_db_options_.avoid_flush_during_recovery);
-  if (two_write_queues_) {
-    log_write_mutex_.Lock();
-  }
-  // Mark these as alive so they'll be considered for deletion later by
-  // FindObsoleteFiles()
-  total_log_size_ = 0;
-  log_empty_ = false;
-  for (auto log_number : log_numbers) {
-    LogFileNumberSize log(log_number);
-    std::string fname = LogFileName(immutable_db_options_.wal_dir, log_number);
-    // This gets the appear size of the logs, not including preallocated space.
-    s = env_->GetFileSize(fname, &log.size);
-    if (!s.ok()) {
-      break;
-    }
-    total_log_size_ += log.size;
-    alive_log_files_.push_back(log);
-    // We preallocate space for logs, but then after a crash and restart, those
-    // preallocated space are not needed anymore. It is likely only the last
-    // log has such preallocated space, so we only truncate for the last log.
-    if (log_number == log_numbers.back()) {
-      std::unique_ptr<WritableFile> last_log;
-      Status truncate_status = env_->ReopenWritableFile(
-          fname, &last_log,
-          env_->OptimizeForLogWrite(
-              env_options_,
-              BuildDBOptions(immutable_db_options_, mutable_db_options_)));
-      if (truncate_status.ok()) {
-        truncate_status = last_log->Truncate(log.size);
-      }
-      if (truncate_status.ok()) {
-        truncate_status = last_log->Close();
-      }
-      // Not a critical error if fail to truncate.
-      if (!truncate_status.ok()) {
-        ROCKS_LOG_WARN(immutable_db_options_.info_log,
-                       "Failed to truncate log #%" PRIu64 ": %s", log_number,
-                       truncate_status.ToString().c_str());
-      }
-    }
-  }
-  if (two_write_queues_) {
-    log_write_mutex_.Unlock();
-  }
-  return s;
-}
-
 Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
                                           MemTable* mem, VersionEdit* edit) {
  mutex_.AssertHeld();
@@ -1010,17 +940,11 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
      if (use_custom_gc_ && snapshot_checker == nullptr) {
        snapshot_checker = DisableGCSnapshotChecker::Instance();
      }
-      std::vector<std::unique_ptr<FragmentedRangeTombstoneIterator>>
-          range_del_iters;
-      auto range_del_iter =
-          mem->NewRangeTombstoneIterator(ro, kMaxSequenceNumber);
-      if (range_del_iter != nullptr) {
-        range_del_iters.emplace_back(range_del_iter);
-      }
      s = BuildTable(
          dbname_, env_, *cfd->ioptions(), mutable_cf_options,
          env_options_for_compaction_, cfd->table_cache(), iter.get(),
-          std::move(range_del_iters), &meta, cfd->internal_comparator(),
+          std::unique_ptr<InternalIterator>(mem->NewRangeTombstoneIterator(ro)),
+          &meta, cfd->internal_comparator(),
          cfd->int_tbl_prop_collector_factories(), cfd->GetID(), cfd->GetName(),
          snapshot_seqs, earliest_write_conflict_snapshot, snapshot_checker,
          GetCompressionFlush(*cfd->ioptions(), mutable_cf_options),
@@ -1045,7 +969,7 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
  if (s.ok() && meta.fd.GetFileSize() > 0) {
    edit->AddFile(level, meta.fd.GetNumber(), meta.fd.GetPathId(),
                  meta.fd.GetFileSize(), meta.smallest, meta.largest,
-                  meta.fd.smallest_seqno, meta.fd.largest_seqno,
+                  meta.smallest_seqno, meta.largest_seqno,
                  meta.marked_for_compaction);
  }

@@ -1054,8 +978,8 @@ Status DBImpl::WriteLevel0TableForRecovery(int job_id, ColumnFamilyData* cfd,
  stats.bytes_written = meta.fd.GetFileSize();
  stats.num_output_files = 1;
  cfd->internal_stats()->AddCompactionStats(level, stats);
-  cfd->internal_stats()->AddCFStats(InternalStats::BYTES_FLUSHED,
-                                    meta.fd.GetFileSize());
+  cfd->internal_stats()->AddCFStats(
+      InternalStats::BYTES_FLUSHED, meta.fd.GetFileSize());
  RecordTick(stats_, COMPACT_WRITE_BYTES, meta.fd.GetFileSize());
  return s;
 }
@@ -1127,12 +1051,6 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
        break;
      }
    }
-
-    // For recovery from NoSpace() error, we can only handle
-    // the case where the database is stored in a single path
-    if (paths.size() <= 1) {
-      impl->error_handler_.EnableAutoRecovery();
-    }
  }

  if (!s.ok()) {
@@ -1151,16 +1069,16 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
  s = impl->Recover(column_families);
  if (s.ok()) {
    uint64_t new_log_number = impl->versions_->NewFileNumber();
-    std::unique_ptr<WritableFile> lfile;
+    unique_ptr<WritableFile> lfile;
    EnvOptions soptions(db_options);
    EnvOptions opt_env_options =
        impl->immutable_db_options_.env->OptimizeForLogWrite(
            soptions, BuildDBOptions(impl->immutable_db_options_,
                                     impl->mutable_db_options_));
-    std::string log_fname =
-        LogFileName(impl->immutable_db_options_.wal_dir, new_log_number);
-    s = NewWritableFile(impl->immutable_db_options_.env, log_fname, &lfile,
-                        opt_env_options);
+    s = NewWritableFile(
+        impl->immutable_db_options_.env,
+        LogFileName(impl->immutable_db_options_.wal_dir, new_log_number),
+        &lfile, opt_env_options);
    if (s.ok()) {
      lfile->SetWriteLifeTimeHint(write_hint);
      lfile->SetPreallocationBlockSize(
@@ -1168,10 +1086,8 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
      {
        InstrumentedMutexLock wl(&impl->log_write_mutex_);
        impl->logfile_number_ = new_log_number;
-        const auto& listeners = impl->immutable_db_options_.listeners;
-        std::unique_ptr<WritableFileWriter> file_writer(
-            new WritableFileWriter(std::move(lfile), log_fname, opt_env_options,
-                                   nullptr /* stats */, listeners));
+        unique_ptr<WritableFileWriter> file_writer(
+            new WritableFileWriter(std::move(lfile), opt_env_options));
        impl->logs_.emplace_back(
            new_log_number,
            new log::Writer(
@@ -1248,8 +1164,7 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
          !cfd->mem()->IsMergeOperatorSupported()) {
        s = Status::InvalidArgument(
            "The memtable of column family %s does not support merge operator "
-            "its options.merge_operator is non-null",
-            cfd->GetName().c_str());
+            "its options.merge_operator is non-null", cfd->GetName().c_str());
      }
      if (!s.ok()) {
        break;
@@ -1299,14 +1214,6 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
        }
      }
    }
-
-    // Reserve some disk buffer space. This is a heuristic - when we run out
-    // of disk space, this ensures that there is atleast write_buffer_size
-    // amount of free space before we resume DB writes. In low disk space
-    // conditions, we want to avoid a lot of small L0 files due to frequent
-    // WAL write failures and resultant forced flushes
-    sfm->ReserveDiskBuffer(max_write_buffer_size,
-                           impl->immutable_db_options_.db_paths[0].path);
  }
 #endif  // !ROCKSDB_LITE

@@ -1322,9 +1229,6 @@ Status DBImpl::Open(const DBOptions& db_options, const std::string& dbname,
          persist_options_status.ToString());
    }
  }
-  if (s.ok()) {
-    impl->StartTimedTasks();
-  }
  if (!s.ok()) {
    for (auto* h : *handles) {
      delete h;
@@ -9,6 +9,7 @@
 #include "db/db_impl.h"
 #include "db/db_iter.h"
 #include "db/merge_context.h"
+#include "db/range_del_aggregator.h"
 #include "monitoring/perf_context_imp.h"

 namespace rocksdb {
@@ -30,38 +31,22 @@ Status DBImplReadOnly::Get(const ReadOptions& read_options,
                           ColumnFamilyHandle* column_family, const Slice& key,
                           PinnableSlice* pinnable_val) {
  assert(pinnable_val != nullptr);
-  // TODO: stopwatch DB_GET needed?, perf timer needed?
-  PERF_TIMER_GUARD(get_snapshot_time);
  Status s;
  SequenceNumber snapshot = versions_->LastSequence();
  auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
  auto cfd = cfh->cfd();
-  if (tracer_) {
-    InstrumentedMutexLock lock(&trace_mutex_);
-    if (tracer_) {
-      tracer_->Get(column_family, key);
-    }
-  }
  SuperVersion* super_version = cfd->GetSuperVersion();
  MergeContext merge_context;
-  SequenceNumber max_covering_tombstone_seq = 0;
+  RangeDelAggregator range_del_agg(cfd->internal_comparator(), snapshot);
  LookupKey lkey(key, snapshot);
-  PERF_TIMER_STOP(get_snapshot_time);
  if (super_version->mem->Get(lkey, pinnable_val->GetSelf(), &s, &merge_context,
-                              &max_covering_tombstone_seq, read_options)) {
+                              &range_del_agg, read_options)) {
    pinnable_val->PinSelf();
-    RecordTick(stats_, MEMTABLE_HIT);
  } else {
    PERF_TIMER_GUARD(get_from_output_files_time);
    super_version->current->Get(read_options, lkey, pinnable_val, &s,
-                                &merge_context, &max_covering_tombstone_seq);
-    RecordTick(stats_, MEMTABLE_MISS);
+                                &merge_context, &range_del_agg);
  }
-  RecordTick(stats_, NUMBER_KEYS_READ);
-  size_t size = pinnable_val->size();
-  RecordTick(stats_, BYTES_READ, size);
-  MeasureTime(stats_, BYTES_PER_READ, size);
-  PERF_COUNTER_ADD(get_read_bytes, size);
  return s;
 }

@@ -71,20 +56,18 @@ Iterator* DBImplReadOnly::NewIterator(const ReadOptions& read_options,
  auto cfd = cfh->cfd();
  SuperVersion* super_version = cfd->GetSuperVersion()->Ref();
  SequenceNumber latest_snapshot = versions_->LastSequence();
-  SequenceNumber read_seq =
-      read_options.snapshot != nullptr
-          ? reinterpret_cast<const SnapshotImpl*>(read_options.snapshot)
-                ->number_
-          : latest_snapshot;
  ReadCallback* read_callback = nullptr;  // No read callback provided.
  auto db_iter = NewArenaWrappedDbIterator(
      env_, read_options, *cfd->ioptions(), super_version->mutable_cf_options,
-      read_seq,
+      (read_options.snapshot != nullptr
+           ? reinterpret_cast<const SnapshotImpl*>(read_options.snapshot)
+                 ->number_
+           : latest_snapshot),
      super_version->mutable_cf_options.max_sequential_skip_in_iterations,
      super_version->version_number, read_callback);
  auto internal_iter =
      NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
-                          db_iter->GetRangeDelAggregator(), read_seq);
+                          db_iter->GetRangeDelAggregator());
  db_iter->SetIterUnderDBIter(internal_iter);
  return db_iter;
 }
@@ -100,22 +83,21 @@ Status DBImplReadOnly::NewIterators(
  iterators->clear();
  iterators->reserve(column_families.size());
  SequenceNumber latest_snapshot = versions_->LastSequence();
-  SequenceNumber read_seq =
-      read_options.snapshot != nullptr
-          ? reinterpret_cast<const SnapshotImpl*>(read_options.snapshot)
-                ->number_
-          : latest_snapshot;

  for (auto cfh : column_families) {
    auto* cfd = reinterpret_cast<ColumnFamilyHandleImpl*>(cfh)->cfd();
    auto* sv = cfd->GetSuperVersion()->Ref();
    auto* db_iter = NewArenaWrappedDbIterator(
-        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options, read_seq,
+        env_, read_options, *cfd->ioptions(), sv->mutable_cf_options,
+        (read_options.snapshot != nullptr
+             ? reinterpret_cast<const SnapshotImpl*>(read_options.snapshot)
+                   ->number_
+             : latest_snapshot),
        sv->mutable_cf_options.max_sequential_skip_in_iterations,
        sv->version_number, read_callback);
    auto* internal_iter =
        NewInternalIterator(read_options, cfd, sv, db_iter->GetArena(),
-                            db_iter->GetRangeDelAggregator(), read_seq);
+                            db_iter->GetRangeDelAggregator());
    db_iter->SetIterUnderDBIter(internal_iter);
    iterators->push_back(db_iter);
  }
@@ -89,11 +89,10 @@ class DBImplReadOnly : public DBImpl {
  virtual Status EnableFileDeletions(bool /*force*/) override {
    return Status::NotSupported("Not supported operation in read only mode.");
  }
-  virtual Status GetLiveFiles(std::vector<std::string>& ret,
-                              uint64_t* manifest_file_size,
-                              bool /*flush_memtable*/) override {
-    return DBImpl::GetLiveFiles(ret, manifest_file_size,
-                                false /* flush_memtable */);
+  virtual Status GetLiveFiles(std::vector<std::string>&,
+                              uint64_t* /*manifest_file_size*/,
+                              bool /*flush_memtable*/ = true) override {
+    return Status::NotSupported("Not supported operation in read only mode.");
  }

  using DBImpl::Flush;
@@ -76,12 +76,6 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
  if (my_batch == nullptr) {
    return Status::Corruption("Batch is nullptr!");
  }
-  if (tracer_) {
-    InstrumentedMutexLock lock(&trace_mutex_);
-    if (tracer_) {
-      tracer_->Write(my_batch);
-    }
-  }
  if (write_options.sync && write_options.disableWAL) {
    return Status::InvalidArgument("Sync writes has to enable WAL.");
  }
@@ -265,19 +259,18 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
    // We're optimistic, updating the stats before we successfully
    // commit.  That lets us release our leader status early.
    auto stats = default_cf_internal_stats_;
-    stats->AddDBStats(InternalStats::kIntStatsNumKeysWritten, total_count,
+    stats->AddDBStats(InternalStats::NUMBER_KEYS_WRITTEN, total_count,
                      concurrent_update);
    RecordTick(stats_, NUMBER_KEYS_WRITTEN, total_count);
-    stats->AddDBStats(InternalStats::kIntStatsBytesWritten, total_byte_size,
+    stats->AddDBStats(InternalStats::BYTES_WRITTEN, total_byte_size,
                      concurrent_update);
    RecordTick(stats_, BYTES_WRITTEN, total_byte_size);
-    stats->AddDBStats(InternalStats::kIntStatsWriteDoneBySelf, 1,
-                      concurrent_update);
+    stats->AddDBStats(InternalStats::WRITE_DONE_BY_SELF, 1, concurrent_update);
    RecordTick(stats_, WRITE_DONE_BY_SELF);
    auto write_done_by_other = write_group.size - 1;
    if (write_done_by_other > 0) {
-      stats->AddDBStats(InternalStats::kIntStatsWriteDoneByOther,
-                        write_done_by_other, concurrent_update);
+      stats->AddDBStats(InternalStats::WRITE_DONE_BY_OTHER, write_done_by_other,
+                        concurrent_update);
      RecordTick(stats_, WRITE_DONE_BY_OTHER, write_done_by_other);
    }
    MeasureTime(stats_, BYTES_PER_WRITE, total_byte_size);
@@ -318,8 +311,7 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
        w.status = WriteBatchInternal::InsertInto(
            write_group, current_sequence, column_family_memtables_.get(),
            &flush_scheduler_, write_options.ignore_missing_column_families,
-            0 /*recovery_log_number*/, this, parallel, seq_per_batch_,
-            batch_per_txn_);
+            0 /*recovery_log_number*/, this, parallel, seq_per_batch_);
      } else {
        SequenceNumber next_sequence = current_sequence;
        // Note: the logic for advancing seq here must be consistent with the
@@ -339,6 +331,8 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
          }
        }
        write_group.last_sequence = last_sequence;
+        write_group.running.store(static_cast<uint32_t>(write_group.size),
+                                  std::memory_order_relaxed);
        write_thread_.LaunchParallelMemTableWriters(&write_group);
        in_parallel_group = true;

@@ -352,7 +346,7 @@ Status DBImpl::WriteImpl(const WriteOptions& write_options,
              &w, w.sequence, &column_family_memtables, &flush_scheduler_,
              write_options.ignore_missing_column_families, 0 /*log_number*/,
              this, true /*concurrent_memtable_writes*/, seq_per_batch_,
-              w.batch_cnt, batch_per_txn_);
+              w.batch_cnt);
        }
      }
      if (seq_used != nullptr) {
@@ -468,9 +462,9 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
    }

    auto stats = default_cf_internal_stats_;
-    stats->AddDBStats(InternalStats::kIntStatsNumKeysWritten, total_count);
+    stats->AddDBStats(InternalStats::NUMBER_KEYS_WRITTEN, total_count);
    RecordTick(stats_, NUMBER_KEYS_WRITTEN, total_count);
-    stats->AddDBStats(InternalStats::kIntStatsBytesWritten, total_byte_size);
+    stats->AddDBStats(InternalStats::BYTES_WRITTEN, total_byte_size);
    RecordTick(stats_, BYTES_WRITTEN, total_byte_size);
    MeasureTime(stats_, BYTES_PER_WRITE, total_byte_size);

@@ -478,10 +472,10 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,

    if (w.ShouldWriteToWAL()) {
      PERF_TIMER_GUARD(write_wal_time);
-      stats->AddDBStats(InternalStats::kIntStatsWriteDoneBySelf, 1);
+      stats->AddDBStats(InternalStats::WRITE_DONE_BY_SELF, 1);
      RecordTick(stats_, WRITE_DONE_BY_SELF, 1);
      if (wal_write_group.size > 1) {
-        stats->AddDBStats(InternalStats::kIntStatsWriteDoneByOther,
+        stats->AddDBStats(InternalStats::WRITE_DONE_BY_OTHER,
                          wal_write_group.size - 1);
        RecordTick(stats_, WRITE_DONE_BY_OTHER, wal_write_group.size - 1);
      }
@@ -514,8 +508,7 @@ Status DBImpl::PipelinedWriteImpl(const WriteOptions& write_options,
      memtable_write_group.status = WriteBatchInternal::InsertInto(
          memtable_write_group, w.sequence, column_family_memtables_.get(),
          &flush_scheduler_, write_options.ignore_missing_column_families,
-          0 /*log_number*/, this, false /*concurrent_memtable_writes*/,
-          seq_per_batch_, batch_per_txn_);
+          0 /*log_number*/, this, seq_per_batch_);
      versions_->SetLastSequence(memtable_write_group.last_sequence);
      write_thread_.ExitAsMemTableWriter(&w, memtable_write_group);
    }
@@ -572,6 +565,7 @@ Status DBImpl::WriteImplWALOnly(const WriteOptions& write_options,
  }
  // else we are the leader of the write batch group
  assert(w.state == WriteThread::STATE_GROUP_LEADER);
+  WriteContext write_context;
  WriteThread::WriteGroup write_group;
  uint64_t last_sequence;
  nonmem_write_thread_.EnterAsBatchGroupLeader(&w, &write_group);
@@ -592,16 +586,15 @@ Status DBImpl::WriteImplWALOnly(const WriteOptions& write_options,
  // We're optimistic, updating the stats before we successfully
  // commit.  That lets us release our leader status early.
  auto stats = default_cf_internal_stats_;
-  stats->AddDBStats(InternalStats::kIntStatsBytesWritten, total_byte_size,
+  stats->AddDBStats(InternalStats::BYTES_WRITTEN, total_byte_size,
                    concurrent_update);
  RecordTick(stats_, BYTES_WRITTEN, total_byte_size);
-  stats->AddDBStats(InternalStats::kIntStatsWriteDoneBySelf, 1,
-                    concurrent_update);
+  stats->AddDBStats(InternalStats::WRITE_DONE_BY_SELF, 1, concurrent_update);
  RecordTick(stats_, WRITE_DONE_BY_SELF);
  auto write_done_by_other = write_group.size - 1;
  if (write_done_by_other > 0) {
-    stats->AddDBStats(InternalStats::kIntStatsWriteDoneByOther,
-                      write_done_by_other, concurrent_update);
+    stats->AddDBStats(InternalStats::WRITE_DONE_BY_OTHER, write_done_by_other,
+                      concurrent_update);
    RecordTick(stats_, WRITE_DONE_BY_OTHER, write_done_by_other);
  }
  MeasureTime(stats_, BYTES_PER_WRITE, total_byte_size);
@@ -710,10 +703,6 @@ Status DBImpl::PreprocessWrite(const WriteOptions& write_options,
  assert(write_context != nullptr && need_log_sync != nullptr);
  Status status;

-  if (error_handler_.IsDBStopped()) {
-    status = error_handler_.GetBGError();
-  }
-
  PERF_TIMER_GUARD(write_scheduling_flushes_compactions_time);

  assert(!single_column_family_mode_ ||
@@ -732,6 +721,10 @@ Status DBImpl::PreprocessWrite(const WriteOptions& write_options,
    status = HandleWriteBufferFull(write_context);
  }

+  if (UNLIKELY(status.ok())) {
+    status = error_handler_.GetBGError();
+  }
+
  if (UNLIKELY(status.ok() && !flush_scheduler_.Empty())) {
    status = ScheduleFlushes(write_context);
  }
@@ -910,12 +903,12 @@ Status DBImpl::WriteToWAL(const WriteThread::WriteGroup& write_group,
  if (status.ok()) {
    auto stats = default_cf_internal_stats_;
    if (need_log_sync) {
-      stats->AddDBStats(InternalStats::kIntStatsWalFileSynced, 1);
+      stats->AddDBStats(InternalStats::WAL_FILE_SYNCED, 1);
      RecordTick(stats_, WAL_FILE_SYNCED);
    }
-    stats->AddDBStats(InternalStats::kIntStatsWalFileBytes, log_size);
+    stats->AddDBStats(InternalStats::WAL_FILE_BYTES, log_size);
    RecordTick(stats_, WAL_FILE_BYTES, log_size);
-    stats->AddDBStats(InternalStats::kIntStatsWriteWithWal, write_with_wal);
+    stats->AddDBStats(InternalStats::WRITE_WITH_WAL, write_with_wal);
    RecordTick(stats_, WRITE_WITH_WAL, write_with_wal);
  }
  return status;
@@ -961,10 +954,9 @@ Status DBImpl::ConcurrentWriteToWAL(const WriteThread::WriteGroup& write_group,
  if (status.ok()) {
    const bool concurrent = true;
    auto stats = default_cf_internal_stats_;
-    stats->AddDBStats(InternalStats::kIntStatsWalFileBytes, log_size,
-                      concurrent);
+    stats->AddDBStats(InternalStats::WAL_FILE_BYTES, log_size, concurrent);
    RecordTick(stats_, WAL_FILE_BYTES, log_size);
-    stats->AddDBStats(InternalStats::kIntStatsWriteWithWal, write_with_wal,
+    stats->AddDBStats(InternalStats::WRITE_WITH_WAL, write_with_wal,
                      concurrent);
    RecordTick(stats_, WRITE_WITH_WAL, write_with_wal);
  }
@@ -1017,28 +1009,6 @@ Status DBImpl::WriteRecoverableState() {
  return Status::OK();
 }

-void DBImpl::SelectColumnFamiliesForAtomicFlush(
-    autovector<ColumnFamilyData*>* cfds) {
-  for (ColumnFamilyData* cfd : *versions_->GetColumnFamilySet()) {
-    if (cfd->IsDropped()) {
-      continue;
-    }
-    if (cfd->imm()->NumNotFlushed() != 0 || !cfd->mem()->IsEmpty() ||
-        !cached_recoverable_state_empty_.load()) {
-      cfds->push_back(cfd);
-    }
-  }
-}
-
-// Assign sequence number for atomic flush.
-void DBImpl::AssignAtomicFlushSeq(const autovector<ColumnFamilyData*>& cfds) {
-  assert(immutable_db_options_.atomic_flush);
-  auto seq = versions_->LastSequence();
-  for (auto cfd : cfds) {
-    cfd->imm()->AssignAtomicFlushSeq(seq);
-  }
-}
-
 Status DBImpl::SwitchWAL(WriteContext* write_context) {
  mutex_.AssertHeld();
  assert(write_context != nullptr);
@@ -1087,39 +1057,20 @@ Status DBImpl::SwitchWAL(WriteContext* write_context) {
                 oldest_alive_log, total_log_size_.load(), GetMaxTotalWalSize());
  // no need to refcount because drop is happening in write thread, so can't
  // happen while we're in the write thread
-  autovector<ColumnFamilyData*> cfds;
-  if (immutable_db_options_.atomic_flush) {
-    SelectColumnFamiliesForAtomicFlush(&cfds);
-  } else {
-    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      if (cfd->IsDropped()) {
-        continue;
+  for (auto cfd : *versions_->GetColumnFamilySet()) {
+    if (cfd->IsDropped()) {
+      continue;
+    }
+    if (cfd->OldestLogToKeep() <= oldest_alive_log) {
+      status = SwitchMemtable(cfd, write_context);
+      if (!status.ok()) {
+        break;
      }
-      if (cfd->OldestLogToKeep() <= oldest_alive_log) {
-        cfds.push_back(cfd);
-      }
-    }
-  }
-  for (const auto cfd : cfds) {
-    cfd->Ref();
-    status = SwitchMemtable(cfd, write_context);
-    cfd->Unref();
-    if (!status.ok()) {
-      break;
-    }
-  }
-  if (status.ok()) {
-    if (immutable_db_options_.atomic_flush) {
-      AssignAtomicFlushSeq(cfds);
-    }
-    for (auto cfd : cfds) {
      cfd->imm()->FlushRequested();
+      SchedulePendingFlush(cfd, FlushReason::kWriteBufferManager);
    }
-    FlushRequest flush_req;
-    GenerateFlushRequest(cfds, &flush_req);
-    SchedulePendingFlush(flush_req, FlushReason::kWriteBufferManager);
-    MaybeScheduleFlushOrCompaction();
  }
+  MaybeScheduleFlushOrCompaction();
  return status;
 }

@@ -1141,54 +1092,31 @@ Status DBImpl::HandleWriteBufferFull(WriteContext* write_context) {
      write_buffer_manager_->buffer_size());
  // no need to refcount because drop is happening in write thread, so can't
  // happen while we're in the write thread
-  autovector<ColumnFamilyData*> cfds;
-  if (immutable_db_options_.atomic_flush) {
-    SelectColumnFamiliesForAtomicFlush(&cfds);
-  } else {
-    ColumnFamilyData* cfd_picked = nullptr;
-    SequenceNumber seq_num_for_cf_picked = kMaxSequenceNumber;
+  ColumnFamilyData* cfd_picked = nullptr;
+  SequenceNumber seq_num_for_cf_picked = kMaxSequenceNumber;

-    for (auto cfd : *versions_->GetColumnFamilySet()) {
-      if (cfd->IsDropped()) {
-        continue;
-      }
-      if (!cfd->mem()->IsEmpty()) {
-        // We only consider active mem table, hoping immutable memtable is
-        // already in the process of flushing.
-        uint64_t seq = cfd->mem()->GetCreationSeq();
-        if (cfd_picked == nullptr || seq < seq_num_for_cf_picked) {
-          cfd_picked = cfd;
-          seq_num_for_cf_picked = seq;
-        }
-      }
-    }
-    if (cfd_picked != nullptr) {
-      cfds.push_back(cfd_picked);
-    }
-  }
-
-  for (const auto cfd : cfds) {
-    if (cfd->mem()->IsEmpty()) {
+  for (auto cfd : *versions_->GetColumnFamilySet()) {
+    if (cfd->IsDropped()) {
      continue;
    }
-    cfd->Ref();
-    status = SwitchMemtable(cfd, write_context);
-    cfd->Unref();
-    if (!status.ok()) {
-      break;
+    if (!cfd->mem()->IsEmpty()) {
+      // We only consider active mem table, hoping immutable memtable is
+      // already in the process of flushing.
+      uint64_t seq = cfd->mem()->GetCreationSeq();
+      if (cfd_picked == nullptr || seq < seq_num_for_cf_picked) {
+        cfd_picked = cfd;
+        seq_num_for_cf_picked = seq;
+      }
    }
  }
-  if (status.ok()) {
-    if (immutable_db_options_.atomic_flush) {
-      AssignAtomicFlushSeq(cfds);
+  if (cfd_picked != nullptr) {
+    status = SwitchMemtable(cfd_picked, write_context,
+                            FlushReason::kWriteBufferFull);
+    if (status.ok()) {
+      cfd_picked->imm()->FlushRequested();
+      SchedulePendingFlush(cfd_picked, FlushReason::kWriteBufferFull);
+      MaybeScheduleFlushOrCompaction();
    }
-    for (const auto cfd : cfds) {
-      cfd->imm()->FlushRequested();
-    }
-    FlushRequest flush_req;
-    GenerateFlushRequest(cfds, &flush_req);
-    SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
-    MaybeScheduleFlushOrCompaction();
  }
  return status;
 }
@@ -1211,14 +1139,10 @@ Status DBImpl::DelayWrite(uint64_t num_bytes,
    uint64_t delay = write_controller_.GetDelay(env_, num_bytes);
    if (delay > 0) {
      if (write_options.no_slowdown) {
-        return Status::Incomplete("Write stall");
+        return Status::Incomplete();
      }
      TEST_SYNC_POINT("DBImpl::DelayWrite:Sleep");

-      // Notify write_thread_ about the stall so it can setup a barrier and
-      // fail any pending writers with no_slowdown
-      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
@@ -1235,47 +1159,25 @@ Status DBImpl::DelayWrite(uint64_t num_bytes,
        env_->SleepForMicroseconds(kDelayInterval);
      }
      mutex_.Lock();
-      write_thread_.EndWriteStall();
    }

-    // Don't wait if there's a background error, even if its a soft error. We
-    // might wait here indefinitely as the background compaction may never
-    // finish successfully, resulting in the stall condition lasting
-    // indefinitely
-    while (error_handler_.GetBGError().ok() && write_controller_.IsStopped()) {
+    while (!error_handler_.IsDBStopped() && write_controller_.IsStopped()) {
      if (write_options.no_slowdown) {
-        return Status::Incomplete("Write stall");
+        return Status::Incomplete();
      }
      delayed = true;
-
-      // Notify write_thread_ about the stall so it can setup a barrier and
-      // fail any pending writers with no_slowdown
-      write_thread_.BeginWriteStall();
      TEST_SYNC_POINT("DBImpl::DelayWrite:Wait");
      bg_cv_.Wait();
-      write_thread_.EndWriteStall();
    }
  }
  assert(!delayed || !write_options.no_slowdown);
  if (delayed) {
-    default_cf_internal_stats_->AddDBStats(
-        InternalStats::kIntStatsWriteStallMicros, time_delayed);
+    default_cf_internal_stats_->AddDBStats(InternalStats::WRITE_STALL_MICROS,
+                                           time_delayed);
    RecordTick(stats_, STALL_MICROS, time_delayed);
  }

-  // If DB is not in read-only mode and write_controller is not stopping
-  // writes, we can ignore any background errors and allow the write to
-  // proceed
-  Status s;
-  if (write_controller_.IsStopped()) {
-    // If writes are still stopped, it means we bailed due to a background
-    // error
-    s = Status::Incomplete(error_handler_.GetBGError().ToString());
-  }
-  if (error_handler_.IsDBStopped()) {
-    s = error_handler_.GetBGError();
-  }
-  return s;
+  return error_handler_.GetBGError();
 }

 Status DBImpl::ThrottleLowPriWritesIfNeeded(const WriteOptions& write_options,
@@ -1309,42 +1211,17 @@ Status DBImpl::ThrottleLowPriWritesIfNeeded(const WriteOptions& write_options,
 }

 Status DBImpl::ScheduleFlushes(WriteContext* context) {
-  autovector<ColumnFamilyData*> cfds;
-  if (immutable_db_options_.atomic_flush) {
-    SelectColumnFamiliesForAtomicFlush(&cfds);
-    for (auto cfd : cfds) {
-      cfd->Ref();
-    }
-    flush_scheduler_.Clear();
-  } else {
-    ColumnFamilyData* tmp_cfd;
-    while ((tmp_cfd = flush_scheduler_.TakeNextColumnFamily()) != nullptr) {
-      cfds.push_back(tmp_cfd);
-    }
-  }
-  Status status;
-  for (auto& cfd : cfds) {
-    if (!cfd->mem()->IsEmpty()) {
-      status = SwitchMemtable(cfd, context);
-    }
+  ColumnFamilyData* cfd;
+  while ((cfd = flush_scheduler_.TakeNextColumnFamily()) != nullptr) {
+    auto status = SwitchMemtable(cfd, context, FlushReason::kWriteBufferFull);
    if (cfd->Unref()) {
      delete cfd;
-      cfd = nullptr;
    }
    if (!status.ok()) {
-      break;
+      return status;
    }
  }
-  if (status.ok()) {
-    if (immutable_db_options_.atomic_flush) {
-      AssignAtomicFlushSeq(cfds);
-    }
-    FlushRequest flush_req;
-    GenerateFlushRequest(cfds, &flush_req);
-    SchedulePendingFlush(flush_req, FlushReason::kWriteBufferFull);
-    MaybeScheduleFlushOrCompaction();
-  }
-  return status;
+  return Status::OK();
 }

 #ifndef ROCKSDB_LITE
@@ -1365,7 +1242,8 @@ void DBImpl::NotifyOnMemTableSealed(ColumnFamilyData* /*cfd*/,

 // REQUIRES: mutex_ is held
 // REQUIRES: this thread is currently at the front of the writer queue
-Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
+Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context,
+                              FlushReason flush_reason) {
  mutex_.AssertHeld();
  WriteThread::Writer nonmem_w;
  if (two_write_queues_) {
@@ -1374,7 +1252,7 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
    nonmem_write_thread_.EnterUnbatched(&nonmem_w, &mutex_);
  }

-  std::unique_ptr<WritableFile> lfile;
+  unique_ptr<WritableFile> lfile;
  log::Writer* new_log = nullptr;
  MemTable* new_mem = nullptr;

@@ -1434,8 +1312,6 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
  auto write_hint = CalculateWALWriteHint();
  mutex_.Unlock();
  {
-    std::string log_fname =
-        LogFileName(immutable_db_options_.wal_dir, new_log_number);
    if (creating_new_log) {
      EnvOptions opt_env_opt =
          env_->OptimizeForLogWrite(env_options_, db_options);
@@ -1443,12 +1319,14 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
        ROCKS_LOG_INFO(immutable_db_options_.info_log,
                       "reusing log %" PRIu64 " from recycle list\n",
                       recycle_log_number);
-        std::string old_log_fname =
-            LogFileName(immutable_db_options_.wal_dir, recycle_log_number);
-        s = env_->ReuseWritableFile(log_fname, old_log_fname, &lfile,
-                                    opt_env_opt);
+        s = env_->ReuseWritableFile(
+            LogFileName(immutable_db_options_.wal_dir, new_log_number),
+            LogFileName(immutable_db_options_.wal_dir, recycle_log_number),
+            &lfile, opt_env_opt);
      } else {
-        s = NewWritableFile(env_, log_fname, &lfile, opt_env_opt);
+        s = NewWritableFile(
+            env_, LogFileName(immutable_db_options_.wal_dir, new_log_number),
+            &lfile, opt_env_opt);
      }
      if (s.ok()) {
        // Our final size should be less than write_buffer_size
@@ -1458,9 +1336,8 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
        // of calling GetWalPreallocateBlockSize()
        lfile->SetPreallocationBlockSize(preallocate_block_size);
        lfile->SetWriteLifeTimeHint(write_hint);
-        std::unique_ptr<WritableFileWriter> file_writer(new WritableFileWriter(
-            std::move(lfile), log_fname, opt_env_opt, nullptr /* stats */,
-            immutable_db_options_.listeners));
+        unique_ptr<WritableFileWriter> file_writer(
+            new WritableFileWriter(std::move(lfile), opt_env_opt));
        new_log = new log::Writer(
            std::move(file_writer), new_log_number,
            immutable_db_options_.recycle_log_file_num > 0, manual_wal_flush_);
@@ -1538,7 +1415,7 @@ Status DBImpl::SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context) {
  new_mem->Ref();
  cfd->SetMemtable(new_mem);
  InstallSuperVersionAndScheduleWork(cfd, &context->superversion_context,
-                                     mutable_cf_options);
+                                     mutable_cf_options, flush_reason);
  if (two_write_queues_) {
    nonmem_write_thread_.ExitUnbatched(&nonmem_w);
  }
@@ -42,7 +42,7 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
          "Error when reading %s dir\n", dbname.c_str());
  }
  std::sort(files.begin(), files.end());
-  for (const std::string& file : files) {
+  for (std::string file : files) {
    if (!ParseFileName(file, &number, &type)) {
      continue;
    }
@@ -85,7 +85,7 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
        continue;
      }
      std::sort(files.begin(), files.end());
-      for (const std::string& file : files) {
+      for (std::string file : files) {
        if (ParseFileName(file, &number, &type)) {
          if (type == kTableFile && ++file_num < 10) {
            file_info.append(file).append(" ");
@@ -109,7 +109,7 @@ void DumpDBFileSummary(const ImmutableDBOptions& options,
      return;
    }
    wal_info.clear();
-    for (const std::string& file : files) {
+    for (std::string file : files) {
      if (ParseFileName(file, &number, &type)) {
        if (type == kLogFile) {
          env->GetFileSize(options.wal_dir + "/" + file, &file_size);
@@ -88,6 +88,7 @@ TEST_F(DBIOFailureTest, DropWritesFlush) {
    env_->drop_writes_.store(false, std::memory_order_release);
  } while (ChangeCompactOptions());
 }
+#endif  // ROCKSDB_LITE

 // Check that CompactRange() returns failure if there is not enough space left
 // on device
@@ -115,7 +116,6 @@ TEST_F(DBIOFailureTest, NoSpaceCompactRange) {
    env_->no_space_.store(false, std::memory_order_release);
  } while (ChangeCompactOptions());
 }
-#endif  // ROCKSDB_LITE

 TEST_F(DBIOFailureTest, NonWritableFileSystem) {
  do {
@@ -27,7 +27,6 @@
 #include "util/logging.h"
 #include "util/mutexlock.h"
 #include "util/string_util.h"
-#include "util/trace_replay.h"

 namespace rocksdb {

@@ -115,8 +114,7 @@ class DBIter final: public Iterator {
         const MutableCFOptions& mutable_cf_options, const Comparator* cmp,
         InternalIterator* iter, SequenceNumber s, bool arena_mode,
         uint64_t max_sequential_skip_in_iterations,
-         ReadCallback* read_callback, DBImpl* db_impl, ColumnFamilyData* cfd,
-         bool allow_blob)
+         ReadCallback* read_callback, bool allow_blob)
      : arena_mode_(arena_mode),
        env_(_env),
        logger_(cf_options.info_log),
@@ -134,14 +132,13 @@ class DBIter final: public Iterator {
        prefix_same_as_start_(read_options.prefix_same_as_start),
        pin_thru_lifetime_(read_options.pin_data),
        total_order_seek_(read_options.total_order_seek),
-        range_del_agg_(&cf_options.internal_comparator, s),
+        range_del_agg_(cf_options.internal_comparator, s,
+                       true /* collapse_deletions */),
        read_callback_(read_callback),
-        db_impl_(db_impl),
-        cfd_(cfd),
        allow_blob_(allow_blob),
        is_blob_(false),
        start_seqnum_(read_options.iter_start_seqnum) {
-    RecordTick(statistics_, NO_ITERATOR_CREATED);
+    RecordTick(statistics_, NO_ITERATORS);
    prefix_extractor_ = mutable_cf_options.prefix_extractor.get();
    max_skip_ = max_sequential_skip_in_iterations;
    max_skippable_internal_keys_ = read_options.max_skippable_internal_keys;
@@ -157,7 +154,9 @@ class DBIter final: public Iterator {
    if (pinned_iters_mgr_.PinningEnabled()) {
      pinned_iters_mgr_.ReleasePinnedData();
    }
-    RecordTick(statistics_, NO_ITERATOR_DELETED);
+    // Compiler warning issue filed:
+    // https://github.com/facebook/rocksdb/issues/3013
+    RecordTick(statistics_, NO_ITERATORS, uint64_t(-1));
    ResetInternalKeysSkippedCounter();
    local_stats_.BumpGlobalStatistics(statistics_);
    if (!arena_mode_) {
@@ -171,7 +170,7 @@ class DBIter final: public Iterator {
    iter_ = iter;
    iter_->SetPinnedItersMgr(&pinned_iters_mgr_);
  }
-  virtual ReadRangeDelAggregator* GetRangeDelAggregator() {
+  virtual RangeDelAggregator* GetRangeDelAggregator() {
    return &range_del_agg_;
  }

@@ -229,7 +228,7 @@ class DBIter final: public Iterator {
      *prop = saved_key_.GetUserKey().ToString();
      return Status::OK();
    }
-    return Status::InvalidArgument("Unidentified property.");
+    return Status::InvalidArgument("Undentified property.");
  }

  virtual void Next() override;
@@ -341,12 +340,10 @@ class DBIter final: public Iterator {
  const bool total_order_seek_;
  // List of operands for merge operator.
  MergeContext merge_context_;
-  ReadRangeDelAggregator range_del_agg_;
+  RangeDelAggregator range_del_agg_;
  LocalStatistics local_stats_;
  PinnedIteratorsManager pinned_iters_mgr_;
  ReadCallback* read_callback_;
-  DBImpl* db_impl_;
-  ColumnFamilyData* cfd_;
  bool allow_blob_;
  bool is_blob_;
  // for diff snapshots we want the lower bound on the seqnum;
@@ -1112,7 +1109,7 @@ bool DBIter::FindValueForCurrentKeyUsingSeek() {

    if (ikey.type == kTypeDeletion || ikey.type == kTypeSingleDeletion ||
        range_del_agg_.ShouldDelete(
-            ikey, RangeDelPositioningMode::kForwardTraversal)) {
+            ikey, RangeDelPositioningMode::kBackwardTraversal)) {
      break;
    } else if (ikey.type == kTypeValue) {
      const Slice val = iter_->value();
@@ -1270,12 +1267,6 @@ void DBIter::Seek(const Slice& target) {
  saved_key_.Clear();
  saved_key_.SetInternalKey(target, seq);

-#ifndef ROCKSDB_LITE
-  if (db_impl_ != nullptr && cfd_ != nullptr) {
-    db_impl_->TraceIteratorSeek(cfd_->GetID(), target);
-  }
-#endif  // ROCKSDB_LITE
-
  if (iterate_lower_bound_ != nullptr &&
      user_comparator_->Compare(saved_key_.GetUserKey(),
                                *iterate_lower_bound_) < 0) {
@@ -1340,12 +1331,6 @@ void DBIter::SeekForPrev(const Slice& target) {
    range_del_agg_.InvalidateRangeDelMapPositions();
  }

-#ifndef ROCKSDB_LITE
-  if (db_impl_ != nullptr && cfd_ != nullptr) {
-    db_impl_->TraceIteratorSeekForPrev(cfd_->GetID(), target);
-  }
-#endif  // ROCKSDB_LITE
-
  RecordTick(statistics_, NUMBER_DB_SEEK);
  if (iter_->Valid()) {
    if (prefix_extractor_ && prefix_same_as_start_) {
@@ -1468,18 +1453,17 @@ Iterator* NewDBIterator(Env* env, const ReadOptions& read_options,
                        InternalIterator* internal_iter,
                        const SequenceNumber& sequence,
                        uint64_t max_sequential_skip_in_iterations,
-                        ReadCallback* read_callback, DBImpl* db_impl,
-                        ColumnFamilyData* cfd, bool allow_blob) {
-  DBIter* db_iter = new DBIter(
-      env, read_options, cf_options, mutable_cf_options, user_key_comparator,
-      internal_iter, sequence, false, max_sequential_skip_in_iterations,
-      read_callback, db_impl, cfd, allow_blob);
+                        ReadCallback* read_callback, bool allow_blob) {
+  DBIter* db_iter =
+      new DBIter(env, read_options, cf_options, mutable_cf_options,
+                 user_key_comparator, internal_iter, sequence, false,
+                 max_sequential_skip_in_iterations, read_callback, allow_blob);
  return db_iter;
 }

 ArenaWrappedDBIter::~ArenaWrappedDBIter() { db_iter_->~DBIter(); }

-ReadRangeDelAggregator* ArenaWrappedDBIter::GetRangeDelAggregator() {
+RangeDelAggregator* ArenaWrappedDBIter::GetRangeDelAggregator() {
  return db_iter_->GetRangeDelAggregator();
 }

@@ -1520,14 +1504,13 @@ void ArenaWrappedDBIter::Init(Env* env, const ReadOptions& read_options,
                              const SequenceNumber& sequence,
                              uint64_t max_sequential_skip_in_iteration,
                              uint64_t version_number,
-                              ReadCallback* read_callback, DBImpl* db_impl,
-                              ColumnFamilyData* cfd, bool allow_blob,
+                              ReadCallback* read_callback, bool allow_blob,
                              bool allow_refresh) {
  auto mem = arena_.AllocateAligned(sizeof(DBIter));
-  db_iter_ = new (mem) DBIter(env, read_options, cf_options, mutable_cf_options,
-                              cf_options.user_comparator, nullptr, sequence,
-                              true, max_sequential_skip_in_iteration,
-                              read_callback, db_impl, cfd, allow_blob);
+  db_iter_ = new (mem)
+      DBIter(env, read_options, cf_options, mutable_cf_options,
+             cf_options.user_comparator, nullptr, sequence, true,
+             max_sequential_skip_in_iteration, read_callback, allow_blob);
  sv_number_ = version_number;
  allow_refresh_ = allow_refresh;
 }
@@ -1551,12 +1534,10 @@ Status ArenaWrappedDBIter::Refresh() {
    SuperVersion* sv = cfd_->GetReferencedSuperVersion(db_impl_->mutex());
    Init(env, read_options_, *(cfd_->ioptions()), sv->mutable_cf_options,
         latest_seq, sv->mutable_cf_options.max_sequential_skip_in_iterations,
-         cur_sv_number, read_callback_, db_impl_, cfd_, allow_blob_,
-         allow_refresh_);
+         cur_sv_number, read_callback_, allow_blob_, allow_refresh_);

    InternalIterator* internal_iter = db_impl_->NewInternalIterator(
-        read_options_, cfd_, sv, &arena_, db_iter_->GetRangeDelAggregator(),
-        latest_seq);
+        read_options_, cfd_, sv, &arena_, db_iter_->GetRangeDelAggregator());
    SetIterUnderDBIter(internal_iter);
  } else {
    db_iter_->set_sequence(latest_seq);
@@ -1575,7 +1556,7 @@ ArenaWrappedDBIter* NewArenaWrappedDbIterator(
  ArenaWrappedDBIter* iter = new ArenaWrappedDBIter();
  iter->Init(env, read_options, cf_options, mutable_cf_options, sequence,
             max_sequential_skip_in_iterations, version_number, read_callback,
-             db_impl, cfd, allow_blob, allow_refresh);
+             allow_blob, allow_refresh);
  if (db_impl != nullptr && cfd != nullptr && allow_refresh) {
    iter->StoreRefreshInfo(read_options, db_impl, cfd, read_callback,
                           allow_blob);
@@ -23,18 +23,20 @@ namespace rocksdb {

 class Arena;
 class DBIter;
+class InternalIterator;

 // Return a new iterator that converts internal keys (yielded by
 // "*internal_iter") that were live at the specified "sequence" number
 // into appropriate user keys.
-extern Iterator* NewDBIterator(
-    Env* env, const ReadOptions& read_options,
-    const ImmutableCFOptions& cf_options,
-    const MutableCFOptions& mutable_cf_options,
-    const Comparator* user_key_comparator, InternalIterator* internal_iter,
-    const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iterations,
-    ReadCallback* read_callback, DBImpl* db_impl = nullptr,
-    ColumnFamilyData* cfd = nullptr, bool allow_blob = false);
+extern Iterator* NewDBIterator(Env* env, const ReadOptions& read_options,
+                               const ImmutableCFOptions& cf_options,
+                               const MutableCFOptions& mutable_cf_options,
+                               const Comparator* user_key_comparator,
+                               InternalIterator* internal_iter,
+                               const SequenceNumber& sequence,
+                               uint64_t max_sequential_skip_in_iterations,
+                               ReadCallback* read_callback,
+                               bool allow_blob = false);

 // A wrapper iterator which wraps DB Iterator and the arena, with which the DB
 // iterator is supposed be allocated. This class is used as an entry point of
@@ -48,7 +50,7 @@ class ArenaWrappedDBIter : public Iterator {
  // Get the arena to be used to allocate memory for DBIter to be wrapped,
  // as well as child iterators in it.
  virtual Arena* GetArena() { return &arena_; }
-  virtual ReadRangeDelAggregator* GetRangeDelAggregator();
+  virtual RangeDelAggregator* GetRangeDelAggregator();

  // Set the internal iterator wrapped inside the DB Iterator. Usually it is
  // a merging iterator.
@@ -73,8 +75,7 @@ class ArenaWrappedDBIter : public Iterator {
            const MutableCFOptions& mutable_cf_options,
            const SequenceNumber& sequence,
            uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
-            ReadCallback* read_callback, DBImpl* db_impl, ColumnFamilyData* cfd,
-            bool allow_blob, bool allow_refresh);
+            ReadCallback* read_callback, bool allow_blob, bool allow_refresh);

  void StoreRefreshInfo(const ReadOptions& read_options, DBImpl* db_impl,
                        ColumnFamilyData* cfd, ReadCallback* read_callback,
@@ -85,7 +85,7 @@ TEST_P(DBIteratorTest, IteratorProperty) {
  ReadOptions ropt;
  ropt.pin_data = false;
  {
-    std::unique_ptr<Iterator> iter(NewIterator(ropt, handles_[1]));
+    unique_ptr<Iterator> iter(NewIterator(ropt, handles_[1]));
    iter->SeekToFirst();
    std::string prop_value;
    ASSERT_NOK(iter->GetProperty("non_existing.value", &prop_value));
@@ -1951,7 +1951,6 @@ TEST_P(DBIteratorTest, ReadAhead) {
  size_t bytes_read = env_->random_read_bytes_counter_;
  delete iter;

-  int64_t num_file_closes = TestGetTickerCount(options, NO_FILE_CLOSES);
  env_->random_read_bytes_counter_ = 0;
  options.statistics->setTickerCount(NO_FILE_OPENS, 0);
  read_options.readahead_size = 1024 * 10;
@@ -1960,10 +1959,7 @@ TEST_P(DBIteratorTest, ReadAhead) {
  int64_t num_file_opens_readahead = TestGetTickerCount(options, NO_FILE_OPENS);
  size_t bytes_read_readahead = env_->random_read_bytes_counter_;
  delete iter;
-  int64_t num_file_closes_readahead =
-      TestGetTickerCount(options, NO_FILE_CLOSES);
  ASSERT_EQ(num_file_opens + 3, num_file_opens_readahead);
-  ASSERT_EQ(num_file_closes + 3, num_file_closes_readahead);
  ASSERT_GT(bytes_read_readahead, bytes_read);
  ASSERT_GT(bytes_read_readahead, read_options.readahead_size * 3);

@@ -2097,34 +2093,6 @@ TEST_P(DBIteratorTest, Refresh) {
  iter.reset();
 }

-TEST_P(DBIteratorTest, RefreshWithSnapshot) {
-  ASSERT_OK(Put("x", "y"));
-  const Snapshot* snapshot = db_->GetSnapshot();
-  ReadOptions options;
-  options.snapshot = snapshot;
-  Iterator* iter = NewIterator(options);
-
-  iter->Seek(Slice("a"));
-  ASSERT_TRUE(iter->Valid());
-  ASSERT_EQ(iter->key().compare(Slice("x")), 0);
-  iter->Next();
-  ASSERT_FALSE(iter->Valid());
-
-  ASSERT_OK(Put("c", "d"));
-
-  iter->Seek(Slice("a"));
-  ASSERT_TRUE(iter->Valid());
-  ASSERT_EQ(iter->key().compare(Slice("x")), 0);
-  iter->Next();
-  ASSERT_FALSE(iter->Valid());
-
-  Status s;
-  s = iter->Refresh();
-  ASSERT_TRUE(s.IsNotSupported());
-  db_->ReleaseSnapshot(snapshot);
-  delete iter;
-}
-
 TEST_P(DBIteratorTest, CreationFailure) {
  SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::NewInternalIterator:StatusCallback", [](void* arg) {
@@ -2377,7 +2345,7 @@ TEST_P(DBIteratorTest, SeekAfterHittingManyInternalKeys) {
  Delete("5");
  Put("6", "val_6");

-  std::unique_ptr<Iterator> iter(NewIterator(ropts));
+  unique_ptr<Iterator> iter(NewIterator(ropts));
  iter->SeekToFirst();

  ASSERT_TRUE(iter->Valid());
@@ -2396,7 +2364,7 @@ TEST_P(DBIteratorTest, SeekAfterHittingManyInternalKeys) {
  ASSERT_EQ("4", prop_value);

  // Create a new iterator to seek to the internal key.
-  std::unique_ptr<Iterator> iter2(NewIterator(ropts));
+  unique_ptr<Iterator> iter2(NewIterator(ropts));
  iter2->Seek(prop_value);
  ASSERT_TRUE(iter2->Valid());
  ASSERT_OK(iter2->status());
@@ -2424,7 +2392,7 @@ TEST_P(DBIteratorTest, NonBlockingIterationBugRepro) {
  // Create a nonblocking iterator before writing to memtable.
  ReadOptions ropt;
  ropt.read_tier = kBlockCacheTier;
-  std::unique_ptr<Iterator> iter(NewIterator(ropt));
+  unique_ptr<Iterator> iter(NewIterator(ropt));

  // Overwrite a key in memtable many times to hit
  // max_sequential_skip_in_iterations (which is 8 by default).
@@ -2434,7 +2402,7 @@ TEST_P(DBIteratorTest, NonBlockingIterationBugRepro) {

  // Load the second block in sst file into the block cache.
  {
-    std::unique_ptr<Iterator> iter2(NewIterator(ReadOptions()));
+    unique_ptr<Iterator> iter2(NewIterator(ReadOptions()));
    iter2->Seek("d");
  }

@@ -23,7 +23,7 @@ class DBTestXactLogIterator : public DBTestBase {

  std::unique_ptr<TransactionLogIterator> OpenTransactionLogIter(
      const SequenceNumber seq) {
-    std::unique_ptr<TransactionLogIterator> iter;
+    unique_ptr<TransactionLogIterator> iter;
    Status status = dbfull()->GetUpdatesSince(seq, &iter);
    EXPECT_OK(status);
    EXPECT_TRUE(iter->Valid());
@@ -8,7 +8,6 @@

 #include "db/db_test_util.h"
 #include "db/memtable.h"
-#include "db/range_del_aggregator.h"
 #include "port/stack_trace.h"
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/slice_transform.h"
@@ -136,8 +135,7 @@ TEST_F(DBMemTableTest, DuplicateSeq) {
  MergeContext merge_context;
  Options options;
  InternalKeyComparator ikey_cmp(options.comparator);
-  ReadRangeDelAggregator range_del_agg(&ikey_cmp,
-                                       kMaxSequenceNumber /* upper_bound */);
+  RangeDelAggregator range_del_agg(ikey_cmp, {} /* snapshots */);

  // Create a MemTable
  InternalKeyComparator cmp(BytewiseComparator());
@@ -514,33 +514,6 @@ TEST_F(DBOptionsTest, SetStatsDumpPeriodSec) {
  }
 }

-TEST_F(DBOptionsTest, RunStatsDumpPeriodSec) {
-  Options options;
-  options.create_if_missing = true;
-  options.stats_dump_period_sec = 5;
-  std::unique_ptr<rocksdb::MockTimeEnv> mock_env;
-  mock_env.reset(new rocksdb::MockTimeEnv(env_));
-  mock_env->set_current_time(0); // in seconds
-  options.env = mock_env.get();
-  int counter = 0;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::DumpStats:1", [&](void* /*arg*/) {
-        counter++;
-      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-  Reopen(options);
-  ASSERT_EQ(5, dbfull()->GetDBOptions().stats_dump_period_sec);
-  dbfull()->TEST_WaitForTimedTaskRun([&] { mock_env->set_current_time(5); });
-  ASSERT_GE(counter, 1);
-
-  // Test cacel job through SetOptions
-  ASSERT_OK(dbfull()->SetDBOptions({{"stats_dump_period_sec", "0"}}));
-  int old_val = counter;
-  env_->SleepForMicroseconds(10000000);
-  ASSERT_EQ(counter, old_val);
-  Close();
-}
-
 static void assert_candidate_files_empty(DBImpl* dbfull, const bool empty) {
  dbfull->TEST_LockMutex();
  JobContext job_context(0);
@@ -170,8 +170,6 @@ void ResetTableProperties(TableProperties* tp) {
  tp->raw_value_size = 0;
  tp->num_data_blocks = 0;
  tp->num_entries = 0;
-  tp->num_deletions = 0;
-  tp->num_merge_operands = 0;
  tp->num_range_deletions = 0;
 }

@@ -181,19 +179,18 @@ void ParseTablePropertiesString(std::string tp_string, TableProperties* tp) {
  std::replace(tp_string.begin(), tp_string.end(), '=', ' ');
  ResetTableProperties(tp);
  sscanf(tp_string.c_str(),
-         "# data blocks %" SCNu64 " # entries %" SCNu64 " # deletions %" SCNu64
-         " # merge operands %" SCNu64 " # range deletions %" SCNu64
+         "# data blocks %" SCNu64 " # entries %" SCNu64
+         " # range deletions %" SCNu64
         " raw key size %" SCNu64
         " raw average key size %lf "
         " raw value size %" SCNu64
         " raw average value size %lf "
         " data block size %" SCNu64 " index block size (user-key? %" SCNu64
-         ", delta-value? %" SCNu64 ") %" SCNu64 " filter block size %" SCNu64,
-         &tp->num_data_blocks, &tp->num_entries, &tp->num_deletions,
-         &tp->num_merge_operands, &tp->num_range_deletions, &tp->raw_key_size,
-         &dummy_double, &tp->raw_value_size, &dummy_double, &tp->data_size,
-         &tp->index_key_is_user_key, &tp->index_value_is_delta_encoded,
-         &tp->index_size, &tp->filter_size);
+         ") %" SCNu64 " filter block size %" SCNu64,
+         &tp->num_data_blocks, &tp->num_entries, &tp->num_range_deletions,
+         &tp->raw_key_size, &dummy_double, &tp->raw_value_size, &dummy_double,
+         &tp->data_size, &tp->index_key_is_user_key, &tp->index_size,
+         &tp->filter_size);
 }

 void VerifySimilar(uint64_t a, uint64_t b, double bias) {
@@ -221,51 +218,37 @@ void VerifyTableProperties(const TableProperties& base_tp,
  VerifySimilar(base_tp.filter_size, new_tp.filter_size, filter_size_bias);
  VerifySimilar(base_tp.num_data_blocks, new_tp.num_data_blocks,
                num_data_blocks_bias);
-
  ASSERT_EQ(base_tp.raw_key_size, new_tp.raw_key_size);
  ASSERT_EQ(base_tp.raw_value_size, new_tp.raw_value_size);
  ASSERT_EQ(base_tp.num_entries, new_tp.num_entries);
-  ASSERT_EQ(base_tp.num_deletions, new_tp.num_deletions);
  ASSERT_EQ(base_tp.num_range_deletions, new_tp.num_range_deletions);
-
-  // Merge operands may become Puts, so we only have an upper bound the exact
-  // number of merge operands.
-  ASSERT_GE(base_tp.num_merge_operands, new_tp.num_merge_operands);
 }

-void GetExpectedTableProperties(
-    TableProperties* expected_tp, const int kKeySize, const int kValueSize,
-    const int kPutsPerTable, const int kDeletionsPerTable,
-    const int kMergeOperandsPerTable, const int kRangeDeletionsPerTable,
-    const int kTableCount, const int kBloomBitsPerKey, const size_t kBlockSize,
-    const bool index_key_is_user_key, const bool value_delta_encoding) {
-  const int kKeysPerTable =
-      kPutsPerTable + kDeletionsPerTable + kMergeOperandsPerTable;
-  const int kPutCount = kTableCount * kPutsPerTable;
-  const int kDeletionCount = kTableCount * kDeletionsPerTable;
-  const int kMergeCount = kTableCount * kMergeOperandsPerTable;
+void GetExpectedTableProperties(TableProperties* expected_tp,
+                                const int kKeySize, const int kValueSize,
+                                const int kKeysPerTable,
+                                const int kRangeDeletionsPerTable,
+                                const int kTableCount,
+                                const int kBloomBitsPerKey,
+                                const size_t kBlockSize,
+                                const bool index_key_is_user_key) {
+  const int kKeyCount = kTableCount * kKeysPerTable;
  const int kRangeDeletionCount = kTableCount * kRangeDeletionsPerTable;
-  const int kKeyCount = kPutCount + kDeletionCount + kMergeCount;
  const int kAvgSuccessorSize = kKeySize / 5;
  const int kEncodingSavePerKey = kKeySize / 4;
-  expected_tp->raw_key_size =
-      (kKeyCount + kRangeDeletionCount) * (kKeySize + 8);
-  expected_tp->raw_value_size =
-      (kPutCount + kMergeCount + kRangeDeletionCount) * kValueSize;
+  expected_tp->raw_key_size = (kKeyCount + kRangeDeletionCount) * (kKeySize + 8);
+  expected_tp->raw_value_size = (kKeyCount + kRangeDeletionCount) * kValueSize;
  expected_tp->num_entries = kKeyCount;
-  expected_tp->num_deletions = kDeletionCount;
-  expected_tp->num_merge_operands = kMergeCount;
  expected_tp->num_range_deletions = kRangeDeletionCount;
  expected_tp->num_data_blocks =
-      kTableCount * (kKeysPerTable * (kKeySize - kEncodingSavePerKey + kValueSize)) /
+      kTableCount *
+      (kKeysPerTable * (kKeySize - kEncodingSavePerKey + kValueSize)) /
      kBlockSize;
  expected_tp->data_size =
      kTableCount * (kKeysPerTable * (kKeySize + 8 + kValueSize));
  expected_tp->index_size =
      expected_tp->num_data_blocks *
-      (kAvgSuccessorSize + (index_key_is_user_key ? 0 : 8) -
-       // discount 1 byte as value size is not encoded in value delta encoding
-       (value_delta_encoding ? 1 : 0));
+      (kAvgSuccessorSize + (index_key_is_user_key ? 0 : 8));
  expected_tp->filter_size =
      kTableCount * (kKeysPerTable * kBloomBitsPerKey / 8);
 }
@@ -317,10 +300,8 @@ TEST_F(DBPropertiesTest, ValidateSampleNumber) {

 TEST_F(DBPropertiesTest, AggregatedTableProperties) {
  for (int kTableCount = 40; kTableCount <= 100; kTableCount += 30) {
-    const int kDeletionsPerTable = 5;
-    const int kMergeOperandsPerTable = 15;
    const int kRangeDeletionsPerTable = 5;
-    const int kPutsPerTable = 100;
+    const int kKeysPerTable = 100;
    const int kKeySize = 80;
    const int kValueSize = 200;
    const int kBloomBitsPerKey = 20;
@@ -329,8 +310,6 @@ TEST_F(DBPropertiesTest, AggregatedTableProperties) {
    options.level0_file_num_compaction_trigger = 8;
    options.compression = kNoCompression;
    options.create_if_missing = true;
-    options.preserve_deletes = true;
-    options.merge_operator.reset(new TestPutOperator());

    BlockBasedTableOptions table_options;
    table_options.filter_policy.reset(
@@ -346,17 +325,10 @@ TEST_F(DBPropertiesTest, AggregatedTableProperties) {

    Random rnd(5632);
    for (int table = 1; table <= kTableCount; ++table) {
-      for (int i = 0; i < kPutsPerTable; ++i) {
+      for (int i = 0; i < kKeysPerTable; ++i) {
        db_->Put(WriteOptions(), RandomString(&rnd, kKeySize),
                 RandomString(&rnd, kValueSize));
      }
-      for (int i = 0; i < kDeletionsPerTable; i++) {
-        db_->Delete(WriteOptions(), RandomString(&rnd, kKeySize));
-      }
-      for (int i = 0; i < kMergeOperandsPerTable; i++) {
-        db_->Merge(WriteOptions(), RandomString(&rnd, kKeySize),
-                   RandomString(&rnd, kValueSize));
-      }
      for (int i = 0; i < kRangeDeletionsPerTable; i++) {
        std::string start = RandomString(&rnd, kKeySize);
        std::string end = start;
@@ -370,14 +342,12 @@ TEST_F(DBPropertiesTest, AggregatedTableProperties) {
    TableProperties output_tp;
    ParseTablePropertiesString(property, &output_tp);
    bool index_key_is_user_key = output_tp.index_key_is_user_key > 0;
-    bool value_is_delta_encoded = output_tp.index_value_is_delta_encoded > 0;

    TableProperties expected_tp;
-    GetExpectedTableProperties(
-        &expected_tp, kKeySize, kValueSize, kPutsPerTable, kDeletionsPerTable,
-        kMergeOperandsPerTable, kRangeDeletionsPerTable, kTableCount,
-        kBloomBitsPerKey, table_options.block_size, index_key_is_user_key,
-        value_is_delta_encoded);
+    GetExpectedTableProperties(&expected_tp, kKeySize, kValueSize,
+                               kKeysPerTable, kRangeDeletionsPerTable,
+                               kTableCount, kBloomBitsPerKey,
+                               table_options.block_size, index_key_is_user_key);

    VerifyTableProperties(expected_tp, output_tp);
  }
@@ -446,7 +416,7 @@ TEST_F(DBPropertiesTest, ReadLatencyHistogramByLevel) {
  ASSERT_EQ(std::string::npos, prop.find("** Level 1 read latency histogram"));
  ASSERT_EQ(std::string::npos, prop.find("** Level 2 read latency histogram"));
  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(ReadOptions()));
+    unique_ptr<Iterator> iter(db_->NewIterator(ReadOptions()));
    for (iter->Seek(Key(0)); iter->Valid(); iter->Next()) {
    }
  }
@@ -499,10 +469,8 @@ TEST_F(DBPropertiesTest, ReadLatencyHistogramByLevel) {

 TEST_F(DBPropertiesTest, AggregatedTablePropertiesAtLevel) {
  const int kTableCount = 100;
-  const int kDeletionsPerTable = 2;
-  const int kMergeOperandsPerTable = 2;
  const int kRangeDeletionsPerTable = 2;
-  const int kPutsPerTable = 10;
+  const int kKeysPerTable = 10;
  const int kKeySize = 50;
  const int kValueSize = 400;
  const int kMaxLevel = 7;
@@ -518,8 +486,6 @@ TEST_F(DBPropertiesTest, AggregatedTablePropertiesAtLevel) {
  options.max_bytes_for_level_multiplier = 2;
  // This ensures there no compaction happening when we call GetProperty().
  options.disable_auto_compactions = true;
-  options.preserve_deletes = true;
-  options.merge_operator.reset(new TestPutOperator());

  BlockBasedTableOptions table_options;
  table_options.filter_policy.reset(
@@ -537,17 +503,10 @@ TEST_F(DBPropertiesTest, AggregatedTablePropertiesAtLevel) {
  TableProperties level_tps[kMaxLevel];
  TableProperties tp, sum_tp, expected_tp;
  for (int table = 1; table <= kTableCount; ++table) {
-    for (int i = 0; i < kPutsPerTable; ++i) {
+    for (int i = 0; i < kKeysPerTable; ++i) {
      db_->Put(WriteOptions(), RandomString(&rnd, kKeySize),
               RandomString(&rnd, kValueSize));
    }
-    for (int i = 0; i < kDeletionsPerTable; i++) {
-      db_->Delete(WriteOptions(), RandomString(&rnd, kKeySize));
-    }
-    for (int i = 0; i < kMergeOperandsPerTable; i++) {
-      db_->Merge(WriteOptions(), RandomString(&rnd, kKeySize),
-                 RandomString(&rnd, kValueSize));
-    }
    for (int i = 0; i < kRangeDeletionsPerTable; i++) {
      std::string start = RandomString(&rnd, kKeySize);
      std::string end = start;
@@ -569,14 +528,11 @@ TEST_F(DBPropertiesTest, AggregatedTablePropertiesAtLevel) {
      sum_tp.raw_value_size += level_tps[level].raw_value_size;
      sum_tp.num_data_blocks += level_tps[level].num_data_blocks;
      sum_tp.num_entries += level_tps[level].num_entries;
-      sum_tp.num_deletions += level_tps[level].num_deletions;
-      sum_tp.num_merge_operands += level_tps[level].num_merge_operands;
      sum_tp.num_range_deletions += level_tps[level].num_range_deletions;
    }
    db_->GetProperty(DB::Properties::kAggregatedTableProperties, &tp_string);
    ParseTablePropertiesString(tp_string, &tp);
    bool index_key_is_user_key = tp.index_key_is_user_key > 0;
-    bool value_is_delta_encoded = tp.index_value_is_delta_encoded > 0;
    ASSERT_EQ(sum_tp.data_size, tp.data_size);
    ASSERT_EQ(sum_tp.index_size, tp.index_size);
    ASSERT_EQ(sum_tp.filter_size, tp.filter_size);
@@ -584,15 +540,12 @@ TEST_F(DBPropertiesTest, AggregatedTablePropertiesAtLevel) {
    ASSERT_EQ(sum_tp.raw_value_size, tp.raw_value_size);
    ASSERT_EQ(sum_tp.num_data_blocks, tp.num_data_blocks);
    ASSERT_EQ(sum_tp.num_entries, tp.num_entries);
-    ASSERT_EQ(sum_tp.num_deletions, tp.num_deletions);
-    ASSERT_EQ(sum_tp.num_merge_operands, tp.num_merge_operands);
    ASSERT_EQ(sum_tp.num_range_deletions, tp.num_range_deletions);
    if (table > 3) {
      GetExpectedTableProperties(
-          &expected_tp, kKeySize, kValueSize, kPutsPerTable, kDeletionsPerTable,
-          kMergeOperandsPerTable, kRangeDeletionsPerTable, table,
-          kBloomBitsPerKey, table_options.block_size, index_key_is_user_key,
-          value_is_delta_encoded);
+          &expected_tp, kKeySize, kValueSize, kKeysPerTable,
+          kRangeDeletionsPerTable, table, kBloomBitsPerKey,
+          table_options.block_size, index_key_is_user_key);
      // Gives larger bias here as index block size, filter block size,
      // and data block size become much harder to estimate in this test.
      VerifyTableProperties(expected_tp, tp, 0.5, 0.4, 0.4, 0.25);
@@ -1529,64 +1482,6 @@ TEST_F(DBPropertiesTest, SstFilesSize) {
  ASSERT_TRUE(listener->callback_triggered);
 }

-TEST_F(DBPropertiesTest, MinObsoleteSstNumberToKeep) {
-  class TestListener : public EventListener {
-   public:
-    void OnTableFileCreated(const TableFileCreationInfo& info) override {
-      if (info.reason == TableFileCreationReason::kCompaction) {
-        // Verify the property indicates that SSTs created by a running
-        // compaction cannot be deleted.
-        uint64_t created_file_num;
-        FileType created_file_type;
-        std::string filename =
-            info.file_path.substr(info.file_path.rfind('/') + 1);
-        ASSERT_TRUE(
-            ParseFileName(filename, &created_file_num, &created_file_type));
-        ASSERT_EQ(kTableFile, created_file_type);
-
-        uint64_t keep_sst_lower_bound;
-        ASSERT_TRUE(
-            db_->GetIntProperty(DB::Properties::kMinObsoleteSstNumberToKeep,
-                                &keep_sst_lower_bound));
-
-        ASSERT_LE(keep_sst_lower_bound, created_file_num);
-        validated_ = true;
-      }
-    }
-
-    void SetDB(DB* db) { db_ = db; }
-
-    int GetNumCompactions() { return num_compactions_; }
-
-    // True if we've verified the property for at least one output file
-    bool Validated() { return validated_; }
-
-   private:
-    int num_compactions_ = 0;
-    bool validated_ = false;
-    DB* db_ = nullptr;
-  };
-
-  const int kNumL0Files = 4;
-
-  std::shared_ptr<TestListener> listener = std::make_shared<TestListener>();
-
-  Options options = CurrentOptions();
-  options.listeners.push_back(listener);
-  options.level0_file_num_compaction_trigger = kNumL0Files;
-  DestroyAndReopen(options);
-  listener->SetDB(db_);
-
-  for (int i = 0; i < kNumL0Files; ++i) {
-    // Make sure they overlap in keyspace to prevent trivial move
-    Put("key1", "val");
-    Put("key2", "val");
-    Flush();
-  }
-  dbfull()->TEST_WaitForCompact();
-  ASSERT_TRUE(listener->Validated());
-}
-
 TEST_F(DBPropertiesTest, BlockCacheProperties) {
  Options options;
  uint64_t value;
@@ -191,7 +191,7 @@ TEST_F(DBRangeDelTest, SentinelsOmittedFromOutputFile) {

  std::vector<std::vector<FileMetaData>> files;
  dbfull()->TEST_GetFilesMetaData(db_->DefaultColumnFamily(), &files);
-  ASSERT_GT(files[0][0].fd.smallest_seqno, 0);
+  ASSERT_GT(files[0][0].smallest_seqno, 0);

  db_->ReleaseSnapshot(snapshot);
 }
@@ -1041,16 +1041,11 @@ TEST_F(DBRangeDelTest, RangeTombstoneEndKeyAsSstableUpperBound) {
    //   L2:
    //     [key000000#1,1, key000000#1,1]
    //     [key000002#6,1, key000004#72057594037927935,15]
-    //
-    // At the same time, verify the compaction does not cause the key at the
-    // endpoint (key000002#6,1) to disappear.
-    ASSERT_EQ(value, Get(Key(2)));
    auto begin_str = Key(3);
    const rocksdb::Slice begin = begin_str;
    dbfull()->TEST_CompactRange(1, &begin, nullptr);
    ASSERT_EQ(1, NumTableFilesAtLevel(1));
    ASSERT_EQ(2, NumTableFilesAtLevel(2));
-    ASSERT_EQ(value, Get(Key(2)));
  }

  {
@@ -1108,386 +1103,6 @@ TEST_F(DBRangeDelTest, UnorderedTombstones) {
  ASSERT_TRUE(s.IsNotFound());
 }

-class MockMergeOperator : public MergeOperator {
-  // Mock non-associative operator. Non-associativity is expressed by lack of
-  // implementation for any `PartialMerge*` functions.
- public:
-  virtual bool FullMergeV2(const MergeOperationInput& merge_in,
-                           MergeOperationOutput* merge_out) const override {
-    assert(merge_out != nullptr);
-    merge_out->new_value = merge_in.operand_list.back().ToString();
-    return true;
-  }
-
-  virtual const char* Name() const override { return "MockMergeOperator"; }
-};
-
-TEST_F(DBRangeDelTest, KeyAtOverlappingEndpointReappears) {
-  // This test uses a non-associative merge operator since that is a convenient
-  // way to get compaction to write out files with overlapping user-keys at the
-  // endpoints. Note, however, overlapping endpoints can also occur with other
-  // value types (Put, etc.), assuming the right snapshots are present.
-  const int kFileBytes = 1 << 20;
-  const int kValueBytes = 1 << 10;
-  const int kNumFiles = 4;
-
-  Options options = CurrentOptions();
-  options.compression = kNoCompression;
-  options.disable_auto_compactions = true;
-  options.merge_operator.reset(new MockMergeOperator());
-  options.target_file_size_base = kFileBytes;
-  Reopen(options);
-
-  Random rnd(301);
-  const Snapshot* snapshot = nullptr;
-  for (int i = 0; i < kNumFiles; ++i) {
-    for (int j = 0; j < kFileBytes / kValueBytes; ++j) {
-      auto value = RandomString(&rnd, kValueBytes);
-      ASSERT_OK(db_->Merge(WriteOptions(), "key", value));
-    }
-    if (i == kNumFiles - 1) {
-      // Take snapshot to prevent covered merge operands from being dropped by
-      // compaction.
-      snapshot = db_->GetSnapshot();
-      // The DeleteRange is the last write so all merge operands are covered.
-      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                                 "key", "key_"));
-    }
-    ASSERT_OK(db_->Flush(FlushOptions()));
-  }
-  ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(0));
-  std::string value;
-  ASSERT_TRUE(db_->Get(ReadOptions(), "key", &value).IsNotFound());
-
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr /* begin_key */,
-                              nullptr /* end_key */));
-  ASSERT_EQ(0, NumTableFilesAtLevel(0));
-  // Now we have multiple files at L1 all containing a single user key, thus
-  // guaranteeing overlap in the file endpoints.
-  ASSERT_GT(NumTableFilesAtLevel(1), 1);
-
-  // Verify no merge operands reappeared after the compaction.
-  ASSERT_TRUE(db_->Get(ReadOptions(), "key", &value).IsNotFound());
-
-  // Compact and verify again. It's worthwhile because now the files have
-  // tighter endpoints, so we can verify that doesn't mess anything up.
-  dbfull()->TEST_CompactRange(1 /* level */, nullptr /* begin */,
-                              nullptr /* end */, nullptr /* column_family */,
-                              true /* disallow_trivial_move */);
-  ASSERT_GT(NumTableFilesAtLevel(2), 1);
-  ASSERT_TRUE(db_->Get(ReadOptions(), "key", &value).IsNotFound());
-
-  db_->ReleaseSnapshot(snapshot);
-}
-
-TEST_F(DBRangeDelTest, UntruncatedTombstoneDoesNotDeleteNewerKey) {
-  // Verify a key newer than a range tombstone cannot be deleted by being
-  // compacted to the bottom level (and thus having its seqnum zeroed) before
-  // the range tombstone. This used to happen when range tombstones were
-  // untruncated on reads such that they extended past their file boundaries.
-  //
-  // Test summary:
-  //
-  // - L1 is bottommost.
-  // - A couple snapshots are strategically taken to prevent seqnums from being
-  //   zeroed, range tombstone from being dropped, merge operands from being
-  //   dropped, and merge operands from being combined.
-  // - Left half of files in L1 all have same user key, ensuring their file
-  //   boundaries overlap. In the past this would cause range tombstones to be
-  //   untruncated.
-  // - Right half of L1 files all have different keys, ensuring no overlap.
-  // - A range tombstone spans all L1 keys, so it is stored in every L1 file.
-  // - Keys in the right side of the key-range are overwritten. These are
-  //   compacted down to L1 after releasing snapshots such that their seqnums
-  //   will be zeroed.
-  // - A full range scan is performed. If the tombstone in the left L1 files
-  //   were untruncated, it would now cover keys newer than it (but with zeroed
-  //   seqnums) in the right L1 files.
-  const int kFileBytes = 1 << 20;
-  const int kValueBytes = 1 << 10;
-  const int kNumFiles = 4;
-  const int kMaxKey = kNumFiles* kFileBytes / kValueBytes;
-  const int kKeysOverwritten = 10;
-
-  Options options = CurrentOptions();
-  options.compression = kNoCompression;
-  options.disable_auto_compactions = true;
-  options.merge_operator.reset(new MockMergeOperator());
-  options.num_levels = 2;
-  options.target_file_size_base = kFileBytes;
-  Reopen(options);
-
-  Random rnd(301);
-  // - snapshots[0] prevents merge operands from being combined during
-  //   compaction.
-  // - snapshots[1] prevents merge operands from being dropped due to the
-  //   covering range tombstone.
-  const Snapshot* snapshots[] = {nullptr, nullptr};
-  for (int i = 0; i < kNumFiles; ++i) {
-    for (int j = 0; j < kFileBytes / kValueBytes; ++j) {
-      auto value = RandomString(&rnd, kValueBytes);
-      std::string key;
-      if (i < kNumFiles / 2) {
-        key = Key(0);
-      } else {
-        key = Key(1 + i * kFileBytes / kValueBytes + j);
-      }
-      ASSERT_OK(db_->Merge(WriteOptions(), key, value));
-    }
-    if (i == 0) {
-      snapshots[0] = db_->GetSnapshot();
-    }
-    if (i == kNumFiles - 1) {
-      snapshots[1] = db_->GetSnapshot();
-      // The DeleteRange is the last write so all merge operands are covered.
-      ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                                 Key(0), Key(kMaxKey + 1)));
-    }
-    ASSERT_OK(db_->Flush(FlushOptions()));
-  }
-  ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(0));
-
-  auto get_key_count = [this]() -> int {
-    auto* iter = db_->NewIterator(ReadOptions());
-    iter->SeekToFirst();
-    int keys_found = 0;
-    for (; iter->Valid(); iter->Next()) {
-      ++keys_found;
-    }
-    delete iter;
-    return keys_found;
-  };
-
-  // All keys should be covered
-  ASSERT_EQ(0, get_key_count());
-
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr /* begin_key */,
-                              nullptr /* end_key */));
-  ASSERT_EQ(0, NumTableFilesAtLevel(0));
-  // Roughly the left half of L1 files should have overlapping boundary keys,
-  // while the right half should not.
-  ASSERT_GE(NumTableFilesAtLevel(1), kNumFiles);
-
-  // Now overwrite a few keys that are in L1 files that definitely don't have
-  // overlapping boundary keys.
-  for (int i = kMaxKey; i > kMaxKey - kKeysOverwritten; --i) {
-    auto value = RandomString(&rnd, kValueBytes);
-    ASSERT_OK(db_->Merge(WriteOptions(), Key(i), value));
-  }
-  ASSERT_OK(db_->Flush(FlushOptions()));
-
-  // The overwritten keys are in L0 now, so clearly aren't covered by the range
-  // tombstone in L1.
-  ASSERT_EQ(kKeysOverwritten, get_key_count());
-
-  // Release snapshots so seqnums can be zeroed when L0->L1 happens.
-  db_->ReleaseSnapshot(snapshots[0]);
-  db_->ReleaseSnapshot(snapshots[1]);
-
-  auto begin_key_storage = Key(kMaxKey - kKeysOverwritten + 1);
-  auto end_key_storage = Key(kMaxKey);
-  Slice begin_key(begin_key_storage);
-  Slice end_key(end_key_storage);
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), &begin_key, &end_key));
-  ASSERT_EQ(0, NumTableFilesAtLevel(0));
-  ASSERT_GE(NumTableFilesAtLevel(1), kNumFiles);
-
-  ASSERT_EQ(kKeysOverwritten, get_key_count());
-}
-
-TEST_F(DBRangeDelTest, DeletedMergeOperandReappearsIterPrev) {
-  // Exposes a bug where we were using
-  // `RangeDelPositioningMode::kBackwardTraversal` while scanning merge operands
-  // in the forward direction. Confusingly, this case happened during
-  // `DBIter::Prev`. It could cause assertion failure, or reappearing keys.
-  const int kFileBytes = 1 << 20;
-  const int kValueBytes = 1 << 10;
-  // Need multiple keys so we can get results when calling `Prev()` after
-  // `SeekToLast()`.
-  const int kNumKeys = 3;
-  const int kNumFiles = 4;
-
-  Options options = CurrentOptions();
-  options.compression = kNoCompression;
-  options.disable_auto_compactions = true;
-  options.merge_operator.reset(new MockMergeOperator());
-  options.target_file_size_base = kFileBytes;
-  Reopen(options);
-
-  Random rnd(301);
-  const Snapshot* snapshot = nullptr;
-  for (int i = 0; i < kNumFiles; ++i) {
-    for (int j = 0; j < kFileBytes / kValueBytes; ++j) {
-      auto value = RandomString(&rnd, kValueBytes);
-      ASSERT_OK(db_->Merge(WriteOptions(), Key(j % kNumKeys), value));
-      if (i == 0 && j == kNumKeys) {
-        // Take snapshot to prevent covered merge operands from being dropped or
-        // merged by compaction.
-        snapshot = db_->GetSnapshot();
-        // Do a DeleteRange near the beginning so only the oldest merge operand
-        // for each key is covered. This ensures the sequence of events:
-        //
-        // - `DBIter::Prev()` is called
-        // - After several same versions of the same user key are encountered,
-        //   it decides to seek using `DBIter::FindValueForCurrentKeyUsingSeek`.
-        // - Binary searches to the newest version of the key, which is in the
-        //   leftmost file containing the user key.
-        // - Scans forwards to collect all merge operands. Eventually reaches
-        //   the rightmost file containing the oldest merge operand, which
-        //   should be covered by the `DeleteRange`. If `RangeDelAggregator`
-        //   were not properly using `kForwardTraversal` here, that operand
-        //   would reappear.
-        ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                                   Key(0), Key(kNumKeys + 1)));
-      }
-    }
-    ASSERT_OK(db_->Flush(FlushOptions()));
-  }
-  ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(0));
-
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr /* begin_key */,
-                              nullptr /* end_key */));
-  ASSERT_EQ(0, NumTableFilesAtLevel(0));
-  ASSERT_GT(NumTableFilesAtLevel(1), 1);
-
-  auto* iter = db_->NewIterator(ReadOptions());
-  iter->SeekToLast();
-  int keys_found = 0;
-  for (; iter->Valid(); iter->Prev()) {
-    ++keys_found;
-  }
-  delete iter;
-  ASSERT_EQ(kNumKeys, keys_found);
-
-  db_->ReleaseSnapshot(snapshot);
-}
-
-TEST_F(DBRangeDelTest, SnapshotPreventsDroppedKeys) {
-  const int kFileBytes = 1 << 20;
-
-  Options options = CurrentOptions();
-  options.compression = kNoCompression;
-  options.disable_auto_compactions = true;
-  options.target_file_size_base = kFileBytes;
-  Reopen(options);
-
-  ASSERT_OK(Put(Key(0), "a"));
-  const Snapshot* snapshot = db_->GetSnapshot();
-
-  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0),
-                             Key(10)));
-
-  db_->Flush(FlushOptions());
-
-  ReadOptions read_opts;
-  read_opts.snapshot = snapshot;
-  auto* iter = db_->NewIterator(read_opts);
-
-  iter->SeekToFirst();
-  ASSERT_TRUE(iter->Valid());
-  ASSERT_EQ(Key(0), iter->key());
-
-  iter->Next();
-  ASSERT_FALSE(iter->Valid());
-
-  delete iter;
-  db_->ReleaseSnapshot(snapshot);
-}
-
-TEST_F(DBRangeDelTest, RangeTombstoneWrittenToMinimalSsts) {
-  // Adapted from
-  // https://github.com/cockroachdb/cockroach/blob/de8b3ea603dd1592d9dc26443c2cc92c356fbc2f/pkg/storage/engine/rocksdb_test.go#L1267-L1398.
-  // Regression test for issue where range tombstone was written to more files
-  // than necessary when it began exactly at the begin key in the next
-  // compaction output file.
-  const int kFileBytes = 1 << 20;
-  const int kValueBytes = 4 << 10;
-  Options options = CurrentOptions();
-  options.compression = kNoCompression;
-  options.disable_auto_compactions = true;
-  // Have a bit of slack in the size limits but we enforce them more strictly
-  // when manually flushing/compacting.
-  options.max_compaction_bytes = 2 * kFileBytes;
-  options.target_file_size_base = 2 * kFileBytes;
-  options.write_buffer_size = 2 * kFileBytes;
-  Reopen(options);
-
-  Random rnd(301);
-  for (char first_char : {'a', 'b', 'c'}) {
-    for (int i = 0; i < kFileBytes / kValueBytes; ++i) {
-      std::string key(1, first_char);
-      key.append(Key(i));
-      std::string value = RandomString(&rnd, kValueBytes);
-      ASSERT_OK(Put(key, value));
-    }
-    db_->Flush(FlushOptions());
-    MoveFilesToLevel(2);
-  }
-  ASSERT_EQ(0, NumTableFilesAtLevel(0));
-  ASSERT_EQ(3, NumTableFilesAtLevel(2));
-
-  // Populate the memtable lightly while spanning the whole key-space. The
-  // setting of `max_compaction_bytes` will cause the L0->L1 to output multiple
-  // files to prevent a large L1->L2 compaction later.
-  ASSERT_OK(Put("a", "val"));
-  ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
-                             "c" + Key(1), "d"));
-  // Our compaction output file cutting logic currently only considers point
-  // keys. So, in order for the range tombstone to have a chance at landing at
-  // the start of a new file, we need a point key at the range tombstone's
-  // start.
-  // TODO(ajkr): remove this `Put` after file cutting accounts for range
-  // tombstones (#3977).
-  ASSERT_OK(Put("c" + Key(1), "value"));
-  db_->Flush(FlushOptions());
-
-  // Ensure manual L0->L1 compaction cuts the outputs before the range tombstone
-  // and the range tombstone is only placed in the second SST.
-  std::string begin_key_storage("c" + Key(1));
-  Slice begin_key(begin_key_storage);
-  std::string end_key_storage("d");
-  Slice end_key(end_key_storage);
-  dbfull()->TEST_CompactRange(0 /* level */, &begin_key /* begin */,
-                              &end_key /* end */, nullptr /* column_family */,
-                              true /* disallow_trivial_move */);
-  ASSERT_EQ(2, NumTableFilesAtLevel(1));
-
-  std::vector<LiveFileMetaData> all_metadata;
-  std::vector<LiveFileMetaData> l1_metadata;
-  db_->GetLiveFilesMetaData(&all_metadata);
-  for (const auto& metadata : all_metadata) {
-    if (metadata.level == 1) {
-      l1_metadata.push_back(metadata);
-    }
-  }
-  std::sort(l1_metadata.begin(), l1_metadata.end(),
-            [&](const LiveFileMetaData& a, const LiveFileMetaData& b) {
-              return options.comparator->Compare(a.smallestkey, b.smallestkey) <
-                     0;
-            });
-  ASSERT_EQ("a", l1_metadata[0].smallestkey);
-  ASSERT_EQ("a", l1_metadata[0].largestkey);
-  ASSERT_EQ("c" + Key(1), l1_metadata[1].smallestkey);
-  ASSERT_EQ("d", l1_metadata[1].largestkey);
-
-  TablePropertiesCollection all_table_props;
-  ASSERT_OK(db_->GetPropertiesOfAllTables(&all_table_props));
-  int64_t num_range_deletions = 0;
-  for (const auto& name_and_table_props : all_table_props) {
-    const auto& name = name_and_table_props.first;
-    const auto& table_props = name_and_table_props.second;
-    // The range tombstone should only be output to the second L1 SST.
-    if (name.size() >= l1_metadata[1].name.size() &&
-        name.substr(name.size() - l1_metadata[1].name.size()).compare(l1_metadata[1].name) == 0) {
-      ASSERT_EQ(1, table_props->num_range_deletions);
-      ++num_range_deletions;
-    } else {
-      ASSERT_EQ(0, table_props->num_range_deletions);
-    }
-  }
-  ASSERT_EQ(1, num_range_deletions);
-}
-
 #endif  // ROCKSDB_LITE

 }  // namespace rocksdb
@@ -436,7 +436,7 @@ TEST_F(DBSSTTest, OpenDBWithExistingTrash) {
 // deleted from first db_path were deleted using DeleteScheduler and
 // files in the second path were not.
 TEST_F(DBSSTTest, DeleteSchedulerMultipleDBPaths) {
-  std::atomic<int> bg_delete_file(0);
+  int bg_delete_file = 0;
  rocksdb::SyncPoint::GetInstance()->SetCallBack(
      "DeleteScheduler::DeleteTrashFile:DeleteFile",
      [&](void* /*arg*/) { bg_delete_file++; });
@@ -703,19 +703,26 @@ TEST_F(DBSSTTest, DBWithMaxSpaceAllowedRandomized) {
  // When bg_error_ is set we will verify that the DB size is greater
  // than the limit.

-  std::vector<int> max_space_limits_mbs = {1, 10};
-  std::atomic<bool> bg_error_set(false);
+  std::vector<int> max_space_limits_mbs = {1, 2, 4, 8, 10};
+  decltype(max_space_limits_mbs)::value_type limit_mb_cb;
+  bool bg_error_set = false;
+  uint64_t total_sst_files_size = 0;

-  std::atomic<int> reached_max_space_on_flush(0);
-  std::atomic<int> reached_max_space_on_compaction(0);
+  std::atomic<int> estimate_multiplier(1);
+  int reached_max_space_on_flush = 0;
+  int reached_max_space_on_compaction = 0;
  rocksdb::SyncPoint::GetInstance()->SetCallBack(
      "DBImpl::FlushMemTableToOutputFile:MaxAllowedSpaceReached",
      [&](void* arg) {
        Status* bg_error = static_cast<Status*>(arg);
        bg_error_set = true;
+        GetAllSSTFiles(&total_sst_files_size);
        reached_max_space_on_flush++;
+        // low limit for size calculated using sst files
+        ASSERT_GE(total_sst_files_size, limit_mb_cb * 1024 * 1024);
        // clear error to ensure compaction callback is called
        *bg_error = Status::OK();
+        estimate_multiplier++;  // used in the main loop assert
      });

  rocksdb::SyncPoint::GetInstance()->SetCallBack(
@@ -728,11 +735,15 @@ TEST_F(DBSSTTest, DBWithMaxSpaceAllowedRandomized) {
      "CompactionJob::FinishCompactionOutputFile:MaxAllowedSpaceReached",
      [&](void* /*arg*/) {
        bg_error_set = true;
+        GetAllSSTFiles(&total_sst_files_size);
        reached_max_space_on_compaction++;
      });

  for (auto limit_mb : max_space_limits_mbs) {
    bg_error_set = false;
+    total_sst_files_size = 0;
+    estimate_multiplier = 1;
+    limit_mb_cb = limit_mb;
    rocksdb::SyncPoint::GetInstance()->ClearTrace();
    rocksdb::SyncPoint::GetInstance()->EnableProcessing();
    std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
@@ -746,17 +757,21 @@ TEST_F(DBSSTTest, DBWithMaxSpaceAllowedRandomized) {

    sfm->SetMaxAllowedSpaceUsage(limit_mb * 1024 * 1024);

-    // It is easy to detect if the test is stuck in a loop. No need for
-    // complex termination logic.
+    int keys_written = 0;
+    uint64_t estimated_db_size = 0;
    while (true) {
      auto s = Put(RandomString(&rnd, 10), RandomString(&rnd, 50));
      if (!s.ok()) {
        break;
      }
+      keys_written++;
+      // Check the estimated db size vs the db limit just to make sure we
+      // dont run into an infinite loop
+      estimated_db_size = keys_written * 60;  // ~60 bytes per key
+      ASSERT_LT(estimated_db_size,
+                estimate_multiplier * limit_mb * 1024 * 1024 * 2);
    }
    ASSERT_TRUE(bg_error_set);
-    uint64_t total_sst_files_size = 0;
-    GetAllSSTFiles(&total_sst_files_size);
    ASSERT_GE(total_sst_files_size, limit_mb * 1024 * 1024);
    rocksdb::SyncPoint::GetInstance()->DisableProcessing();
  }
@@ -252,14 +252,13 @@ TEST_F(DBTablePropertiesTest, GetColumnFamilyNameProperty) {
 }

 TEST_F(DBTablePropertiesTest, DeletionTriggeredCompactionMarking) {
-  int kNumKeys = 1000;
-  int kWindowSize = 100;
-  int kNumDelsTrigger = 90;
-  std::shared_ptr<TablePropertiesCollectorFactory> compact_on_del =
-    NewCompactOnDeletionCollectorFactory(kWindowSize, kNumDelsTrigger);
+  const int kNumKeys = 1000;
+  const int kWindowSize = 100;
+  const int kNumDelsTrigger = 90;

  Options opts = CurrentOptions();
-  opts.table_properties_collector_factories.emplace_back(compact_on_del);
+  opts.table_properties_collector_factories.emplace_back(
+      NewCompactOnDeletionCollectorFactory(kWindowSize, kNumDelsTrigger));
  Reopen(opts);

  // add an L1 file to prevent tombstones from dropping due to obsolescence
@@ -281,48 +280,6 @@ TEST_F(DBTablePropertiesTest, DeletionTriggeredCompactionMarking) {
  dbfull()->TEST_WaitForCompact();
  ASSERT_EQ(0, NumTableFilesAtLevel(0));
  ASSERT_GT(NumTableFilesAtLevel(1), 0);
-
-  // Change the window size and deletion trigger and ensure new values take
-  // effect
-  kWindowSize = 50;
-  kNumDelsTrigger = 40;
-  static_cast<CompactOnDeletionCollectorFactory*>
-    (compact_on_del.get())->SetWindowSize(kWindowSize);
-  static_cast<CompactOnDeletionCollectorFactory*>
-    (compact_on_del.get())->SetDeletionTrigger(kNumDelsTrigger);
-  for (int i = 0; i < kNumKeys; ++i) {
-    if (i >= kNumKeys - kWindowSize &&
-        i < kNumKeys - kWindowSize + kNumDelsTrigger) {
-      Delete(Key(i));
-    } else {
-      Put(Key(i), "val");
-    }
-  }
-  Flush();
-
-  dbfull()->TEST_WaitForCompact();
-  ASSERT_EQ(0, NumTableFilesAtLevel(0));
-  ASSERT_GT(NumTableFilesAtLevel(1), 0);
-
-  // Change the window size to disable delete triggered compaction
-  kWindowSize = 0;
-  static_cast<CompactOnDeletionCollectorFactory*>
-    (compact_on_del.get())->SetWindowSize(kWindowSize);
-  static_cast<CompactOnDeletionCollectorFactory*>
-    (compact_on_del.get())->SetDeletionTrigger(kNumDelsTrigger);
-  for (int i = 0; i < kNumKeys; ++i) {
-    if (i >= kNumKeys - kWindowSize &&
-        i < kNumKeys - kWindowSize + kNumDelsTrigger) {
-      Delete(Key(i));
-    } else {
-      Put(Key(i), "val");
-    }
-  }
-  Flush();
-
-  dbfull()->TEST_WaitForCompact();
-  ASSERT_EQ(1, NumTableFilesAtLevel(0));
-
 }

 }  // namespace rocksdb
@@ -93,7 +93,7 @@ class DBTestWithParam
 };

 TEST_F(DBTest, MockEnvTest) {
-  std::unique_ptr<MockEnv> env{new MockEnv(Env::Default())};
+  unique_ptr<MockEnv> env{new MockEnv(Env::Default())};
  Options options;
  options.create_if_missing = true;
  options.env = env.get();
@@ -143,7 +143,7 @@ TEST_F(DBTest, MockEnvTest) {
 // defined.
 #ifndef ROCKSDB_LITE
 TEST_F(DBTest, MemEnvTest) {
-  std::unique_ptr<Env> env{NewMemEnv(Env::Default())};
+  unique_ptr<Env> env{NewMemEnv(Env::Default())};
  Options options;
  options.create_if_missing = true;
  options.env = env.get();
@@ -262,196 +262,6 @@ TEST_F(DBTest, SkipDelay) {
  }
 }

-TEST_F(DBTest, MixedSlowdownOptions) {
-  Options options = CurrentOptions();
-  options.env = env_;
-  options.write_buffer_size = 100000;
-  CreateAndReopenWithCF({"pikachu"}, options);
-  std::vector<port::Thread> threads;
-  std::atomic<int> thread_num(0);
-
-  std::function<void()> write_slowdown_func = [&]() {
-    int a = thread_num.fetch_add(1);
-    std::string key = "foo" + std::to_string(a);
-    WriteOptions wo;
-    wo.no_slowdown = false;
-    ASSERT_OK(dbfull()->Put(wo, key, "bar"));
-  };
-  std::function<void()> write_no_slowdown_func = [&]() {
-    int a = thread_num.fetch_add(1);
-    std::string key = "foo" + std::to_string(a);
-    WriteOptions wo;
-    wo.no_slowdown = true;
-    ASSERT_NOK(dbfull()->Put(wo, key, "bar"));
-  };
-  // Use a small number to ensure a large delay that is still effective
-  // when we do Put
-  // TODO(myabandeh): this is time dependent and could potentially make
-  // the test flaky
-  auto token = dbfull()->TEST_write_controler().GetDelayToken(1);
-  std::atomic<int> sleep_count(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::DelayWrite:BeginWriteStallDone",
-      [&](void* /*arg*/) {
-        sleep_count.fetch_add(1);
-        if (threads.empty()) {
-          for (int i = 0; i < 2; ++i) {
-            threads.emplace_back(write_slowdown_func);
-          }
-          for (int i = 0; i < 2; ++i) {
-            threads.emplace_back(write_no_slowdown_func);
-          }
-        }
-      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  WriteOptions wo;
-  wo.sync = false;
-  wo.disableWAL = false;
-  wo.no_slowdown = false;
-  dbfull()->Put(wo, "foo", "bar");
-  // 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_OK(dbfull()->Put(wo, "foo2", "bar2"));
-          token.reset();
-
-  for (auto& t : threads) {
-    t.join();
-  }
-  ASSERT_GE(sleep_count.load(), 1);
-
-  wo.no_slowdown = true;
-  ASSERT_OK(dbfull()->Put(wo, "foo3", "bar"));
-}
-
-TEST_F(DBTest, MixedSlowdownOptionsInQueue) {
-  Options options = CurrentOptions();
-  options.env = env_;
-  options.write_buffer_size = 100000;
-  CreateAndReopenWithCF({"pikachu"}, options);
-  std::vector<port::Thread> threads;
-  std::atomic<int> thread_num(0);
-
-  std::function<void()> write_no_slowdown_func = [&]() {
-    int a = thread_num.fetch_add(1);
-    std::string key = "foo" + std::to_string(a);
-    WriteOptions wo;
-    wo.no_slowdown = true;
-    ASSERT_NOK(dbfull()->Put(wo, key, "bar"));
-  };
-  // Use a small number to ensure a large delay that is still effective
-  // when we do Put
-  // TODO(myabandeh): this is time dependent and could potentially make
-  // the test flaky
-  auto token = dbfull()->TEST_write_controler().GetDelayToken(1);
-  std::atomic<int> sleep_count(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::DelayWrite:Sleep",
-      [&](void* /*arg*/) {
-        sleep_count.fetch_add(1);
-        if (threads.empty()) {
-          for (int i = 0; i < 2; ++i) {
-            threads.emplace_back(write_no_slowdown_func);
-          }
-          // Sleep for 2s to allow the threads to insert themselves into the
-          // write queue
-          env_->SleepForMicroseconds(3000000ULL);
-        }
-      });
-  std::atomic<int> wait_count(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::DelayWrite:Wait",
-      [&](void* /*arg*/) { wait_count.fetch_add(1); });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  WriteOptions wo;
-  wo.sync = false;
-  wo.disableWAL = false;
-  wo.no_slowdown = false;
-  dbfull()->Put(wo, "foo", "bar");
-  // 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_OK(dbfull()->Put(wo, "foo2", "bar2"));
-          token.reset();
-
-  for (auto& t : threads) {
-    t.join();
-  }
-  ASSERT_EQ(sleep_count.load(), 1);
-  ASSERT_GE(wait_count.load(), 0);
-}
-
-TEST_F(DBTest, MixedSlowdownOptionsStop) {
-  Options options = CurrentOptions();
-  options.env = env_;
-  options.write_buffer_size = 100000;
-  CreateAndReopenWithCF({"pikachu"}, options);
-  std::vector<port::Thread> threads;
-  std::atomic<int> thread_num(0);
-
-  std::function<void()> write_slowdown_func = [&]() {
-    int a = thread_num.fetch_add(1);
-    std::string key = "foo" + std::to_string(a);
-    WriteOptions wo;
-    wo.no_slowdown = false;
-    ASSERT_OK(dbfull()->Put(wo, key, "bar"));
-  };
-  std::function<void()> write_no_slowdown_func = [&]() {
-    int a = thread_num.fetch_add(1);
-    std::string key = "foo" + std::to_string(a);
-    WriteOptions wo;
-    wo.no_slowdown = true;
-    ASSERT_NOK(dbfull()->Put(wo, key, "bar"));
-  };
-  std::function<void()> wakeup_writer = [&]() {
-    dbfull()->mutex_.Lock();
-    dbfull()->bg_cv_.SignalAll();
-    dbfull()->mutex_.Unlock();
-  };
-  // Use a small number to ensure a large delay that is still effective
-  // when we do Put
-  // TODO(myabandeh): this is time dependent and could potentially make
-  // the test flaky
-  auto token = dbfull()->TEST_write_controler().GetStopToken();
-  std::atomic<int> wait_count(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "DBImpl::DelayWrite:Wait",
-      [&](void* /*arg*/) {
-        wait_count.fetch_add(1);
-        if (threads.empty()) {
-          for (int i = 0; i < 2; ++i) {
-            threads.emplace_back(write_slowdown_func);
-          }
-          for (int i = 0; i < 2; ++i) {
-            threads.emplace_back(write_no_slowdown_func);
-          }
-          // Sleep for 2s to allow the threads to insert themselves into the
-          // write queue
-          env_->SleepForMicroseconds(3000000ULL);
-        }
-        token.reset();
-        threads.emplace_back(wakeup_writer);
-      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  WriteOptions wo;
-  wo.sync = false;
-  wo.disableWAL = false;
-  wo.no_slowdown = false;
-  dbfull()->Put(wo, "foo", "bar");
-  // 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_OK(dbfull()->Put(wo, "foo2", "bar2"));
-          token.reset();
-
-  for (auto& t : threads) {
-    t.join();
-  }
-  ASSERT_GE(wait_count.load(), 1);
-
-  wo.no_slowdown = true;
-  ASSERT_OK(dbfull()->Put(wo, "foo3", "bar"));
-}
 #ifndef ROCKSDB_LITE

 TEST_F(DBTest, LevelLimitReopen) {
@@ -2313,9 +2123,6 @@ INSTANTIATE_TEST_CASE_P(
 #endif  // ROCKSDB_LITE

 // Group commit test:
-#if !defined(TRAVIS) && !defined(OS_WIN)
-// Disable this test temporarily on Travis and appveyor as it fails
-// intermittently. Github issue: #4151
 namespace {

 static const int kGCNumThreads = 4;
@@ -2351,9 +2158,7 @@ TEST_F(DBTest, GroupCommitTest) {

    rocksdb::SyncPoint::GetInstance()->LoadDependency(
        {{"WriteThread::JoinBatchGroup:BeganWaiting",
-          "DBImpl::WriteImpl:BeforeLeaderEnters"},
-          {"WriteThread::AwaitState:BlockingWaiting",
-          "WriteThread::EnterAsBatchGroupLeader:End"}});
+          "DBImpl::WriteImpl:BeforeLeaderEnters"}});
    rocksdb::SyncPoint::GetInstance()->EnableProcessing();

    // Start threads
@@ -2390,7 +2195,6 @@ TEST_F(DBTest, GroupCommitTest) {
    ASSERT_GT(hist_data.average, 0.0);
  } while (ChangeOptions(kSkipNoSeekToLast));
 }
-#endif  // TRAVIS

 namespace {
 typedef std::map<std::string, std::string> KVMap;
@@ -2661,11 +2465,6 @@ class ModelDB : public DB {
    Status ret;
    return ret;
  }
-  virtual Status Flush(
-      const rocksdb::FlushOptions& /*options*/,
-      const std::vector<ColumnFamilyHandle*>& /*column_families*/) override {
-    return Status::OK();
-  }

  virtual Status SyncWAL() override { return Status::OK(); }

@@ -2689,8 +2488,7 @@ class ModelDB : public DB {
  }

  virtual Status GetUpdatesSince(
-      rocksdb::SequenceNumber,
-      std::unique_ptr<rocksdb::TransactionLogIterator>*,
+      rocksdb::SequenceNumber, unique_ptr<rocksdb::TransactionLogIterator>*,
      const TransactionLogIterator::ReadOptions& /*read_options*/ =
          TransactionLogIterator::ReadOptions()) override {
    return Status::NotSupported("Not supported in Model DB");
@@ -3512,14 +3310,8 @@ TEST_F(DBTest, SanitizeNumThreads) {
                     (i < 4) ? Env::Priority::LOW : Env::Priority::HIGH);
    }

-    // Wait until 10s for they are scheduled.
-    for (int i = 0; i < 10000; i++) {
-      if (options.env->GetThreadPoolQueueLen(Env::Priority::LOW) <= 1 &&
-          options.env->GetThreadPoolQueueLen(Env::Priority::HIGH) <= 2) {
-        break;
-      }
-      env_->SleepForMicroseconds(1000);
-    }
+    // Wait 100 milliseconds for they are scheduled.
+    env_->SleepForMicroseconds(100000);

    // pool size 3, total task 4. Queue size should be 1.
    ASSERT_EQ(1U, options.env->GetThreadPoolQueueLen(Env::Priority::LOW));
@@ -4535,7 +4327,7 @@ TEST_F(DBTest, DynamicCompactionOptions) {
  // Clean up memtable and L0. Block compaction threads. If continue to write
  // and flush memtables. We should see put stop after 8 memtable flushes
  // since level0_stop_writes_trigger = 8
-  dbfull()->TEST_FlushMemTable(true, true);
+  dbfull()->TEST_FlushMemTable(true);
  dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr);
  // Block compaction
  test::SleepingBackgroundTask sleeping_task_low;
@@ -4548,7 +4340,7 @@ TEST_F(DBTest, DynamicCompactionOptions) {
  WriteOptions wo;
  while (count < 64) {
    ASSERT_OK(Put(Key(count), RandomString(&rnd, 1024), wo));
-    dbfull()->TEST_FlushMemTable(true, true);
+    dbfull()->TEST_FlushMemTable(true);
    count++;
    if (dbfull()->TEST_write_controler().IsStopped()) {
      sleeping_task_low.WakeUp();
@@ -4576,7 +4368,7 @@ TEST_F(DBTest, DynamicCompactionOptions) {
  count = 0;
  while (count < 64) {
    ASSERT_OK(Put(Key(count), RandomString(&rnd, 1024), wo));
-    dbfull()->TEST_FlushMemTable(true, true);
+    dbfull()->TEST_FlushMemTable(true);
    count++;
    if (dbfull()->TEST_write_controler().IsStopped()) {
      sleeping_task_low.WakeUp();
@@ -5655,18 +5447,41 @@ TEST_F(DBTest, HardLimit) {
 #if !defined(ROCKSDB_LITE) && !defined(ROCKSDB_DISABLE_STALL_NOTIFICATION)
 class WriteStallListener : public EventListener {
 public:
-  WriteStallListener() : condition_(WriteStallCondition::kNormal) {}
+  WriteStallListener()
+      : cond_(&mutex_),
+        condition_(WriteStallCondition::kNormal),
+        expected_(WriteStallCondition::kNormal),
+        expected_set_(false) {}
  void OnStallConditionsChanged(const WriteStallInfo& info) override {
    MutexLock l(&mutex_);
    condition_ = info.condition.cur;
+    if (expected_set_ && condition_ == expected_) {
+      cond_.Signal();
+      expected_set_ = false;
+    }
  }
  bool CheckCondition(WriteStallCondition expected) {
    MutexLock l(&mutex_);
-    return expected == condition_;
+    if (expected != condition_) {
+      expected_ = expected;
+      expected_set_ = true;
+      while (expected != condition_) {
+        // We bail out on timeout 500 milliseconds
+        const uint64_t timeout_us = 500000;
+        if (cond_.TimedWait(timeout_us)) {
+          expected_set_ = false;
+          return false;
+        }
+      }
+    }
+    return true;
  }
 private:
  port::Mutex   mutex_;
+  port::CondVar cond_;
  WriteStallCondition condition_;
+  WriteStallCondition expected_;
+  bool                expected_set_;
 };

 TEST_F(DBTest, SoftLimit) {
@@ -5687,48 +5502,13 @@ TEST_F(DBTest, SoftLimit) {
  WriteStallListener* listener = new WriteStallListener();
  options.listeners.emplace_back(listener);

-  // FlushMemtable with opt.wait=true does not wait for
-  // `OnStallConditionsChanged` being called. The event listener is triggered
-  // on `JobContext::Clean`, which happens after flush result is installed.
-  // We use sync point to create a custom WaitForFlush that waits for
-  // context cleanup.
-  port::Mutex flush_mutex;
-  port::CondVar flush_cv(&flush_mutex);
-  bool flush_finished = false;
-  auto InstallFlushCallback = [&]() {
-    {
-      MutexLock l(&flush_mutex);
-      flush_finished = false;
-    }
-    SyncPoint::GetInstance()->SetCallBack(
-        "DBImpl::BackgroundCallFlush:ContextCleanedUp", [&](void*) {
-          {
-            MutexLock l(&flush_mutex);
-            flush_finished = true;
-          }
-          flush_cv.SignalAll();
-        });
-  };
-  auto WaitForFlush = [&]() {
-    {
-      MutexLock l(&flush_mutex);
-      while (!flush_finished) {
-        flush_cv.Wait();
-      }
-    }
-    SyncPoint::GetInstance()->ClearCallBack(
-        "DBImpl::BackgroundCallFlush:ContextCleanedUp");
-  };
-
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
  Reopen(options);

  // Generating 360KB in Level 3
  for (int i = 0; i < 72; i++) {
    Put(Key(i), std::string(5000, 'x'));
    if (i % 10 == 0) {
-      dbfull()->TEST_FlushMemTable(true, true);
+      Flush();
    }
  }
  dbfull()->TEST_WaitForCompact();
@@ -5738,7 +5518,7 @@ TEST_F(DBTest, SoftLimit) {
  for (int i = 0; i < 72; i++) {
    Put(Key(i), std::string(5000, 'x'));
    if (i % 10 == 0) {
-      dbfull()->TEST_FlushMemTable(true, true);
+      Flush();
    }
  }
  dbfull()->TEST_WaitForCompact();
@@ -5757,9 +5537,7 @@ TEST_F(DBTest, SoftLimit) {
    Put(Key(i), std::string(5000, 'x'));
    Put(Key(100 - i), std::string(5000, 'x'));
    // Flush the file. File size is around 30KB.
-    InstallFlushCallback();
-    dbfull()->TEST_FlushMemTable(true, true);
-    WaitForFlush();
+    Flush();
  }
  ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
  ASSERT_TRUE(listener->CheckCondition(WriteStallCondition::kDelayed));
@@ -5784,6 +5562,8 @@ TEST_F(DBTest, SoftLimit) {
                       &sleeping_task_low, Env::Priority::LOW);
      });

+  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
+
  env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
                 Env::Priority::LOW);
  sleeping_task_low.WaitUntilSleeping();
@@ -5792,9 +5572,7 @@ TEST_F(DBTest, SoftLimit) {
    Put(Key(10 + i), std::string(5000, 'x'));
    Put(Key(90 - i), std::string(5000, 'x'));
    // Flush the file. File size is around 30KB.
-    InstallFlushCallback();
-    dbfull()->TEST_FlushMemTable(true, true);
-    WaitForFlush();
+    Flush();
  }

  // Wake up sleep task to enable compaction to run and waits
@@ -5815,9 +5593,7 @@ TEST_F(DBTest, SoftLimit) {
    Put(Key(20 + i), std::string(5000, 'x'));
    Put(Key(80 - i), std::string(5000, 'x'));
    // Flush the file. File size is around 30KB.
-    InstallFlushCallback();
-    dbfull()->TEST_FlushMemTable(true, true);
-    WaitForFlush();
+    Flush();
  }
  // Wake up sleep task to enable compaction to run and waits
  // for it to go to sleep state again to make sure one compaction
@@ -61,7 +61,7 @@ TEST_P(PrefixFullBloomWithReverseComparator,
    bbto.block_cache->EraseUnRefEntries();
  }

-  std::unique_ptr<Iterator> iter(db_->NewIterator(ReadOptions()));
+  unique_ptr<Iterator> iter(db_->NewIterator(ReadOptions()));
  iter->Seek("bar345");
  ASSERT_OK(iter->status());
  ASSERT_TRUE(iter->Valid());
@@ -454,22 +454,6 @@ TEST_F(DBTest2, SharedWriteBufferLimitAcrossDB) {
  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
 }

-TEST_F(DBTest2, TestWriteBufferNoLimitWithCache) {
-  Options options = CurrentOptions();
-  options.arena_block_size = 4096;
-  std::shared_ptr<Cache> cache =
-      NewLRUCache(LRUCacheOptions(10000000, 1, false, 0.0));
-  options.write_buffer_size = 50000;  // this is never hit
-  // Use a write buffer total size so that the soft limit is about
-  // 105000.
-  options.write_buffer_manager.reset(new WriteBufferManager(0, cache));
-  Reopen(options);
-
-  ASSERT_OK(Put("foo", "bar"));
-  // One dummy entry is 1MB.
-  ASSERT_GT(cache->GetUsage(), 500000);
-}
-
 namespace {
  void ValidateKeyExistence(DB* db, const std::vector<Slice>& keys_must_exist,
    const std::vector<Slice>& keys_must_not_exist) {
@@ -1244,14 +1228,7 @@ TEST_F(DBTest2, CompressionOptions) {

 class CompactionStallTestListener : public EventListener {
 public:
-  CompactionStallTestListener() : compacting_files_cnt_(0), compacted_files_cnt_(0) {}
-
-  void OnCompactionBegin(DB* /*db*/, const CompactionJobInfo& ci) override {
-    ASSERT_EQ(ci.cf_name, "default");
-    ASSERT_EQ(ci.base_input_level, 0);
-    ASSERT_EQ(ci.compaction_reason, CompactionReason::kLevelL0FilesNum);
-    compacting_files_cnt_ += ci.input_files.size();
-  }
+  CompactionStallTestListener() : compacted_files_cnt_(0) {}

  void OnCompactionCompleted(DB* /*db*/, const CompactionJobInfo& ci) override {
    ASSERT_EQ(ci.cf_name, "default");
@@ -1259,8 +1236,6 @@ class CompactionStallTestListener : public EventListener {
    ASSERT_EQ(ci.compaction_reason, CompactionReason::kLevelL0FilesNum);
    compacted_files_cnt_ += ci.input_files.size();
  }
-
-  std::atomic<size_t> compacting_files_cnt_;
  std::atomic<size_t> compacted_files_cnt_;
 };

@@ -1269,8 +1244,6 @@ TEST_F(DBTest2, CompactionStall) {
      {{"DBImpl::BGWorkCompaction", "DBTest2::CompactionStall:0"},
       {"DBImpl::BGWorkCompaction", "DBTest2::CompactionStall:1"},
       {"DBTest2::CompactionStall:2",
-        "DBImpl::NotifyOnCompactionBegin::UnlockMutex"},
-       {"DBTest2::CompactionStall:3",
        "DBImpl::NotifyOnCompactionCompleted::UnlockMutex"}});
  rocksdb::SyncPoint::GetInstance()->EnableProcessing();

@@ -1312,18 +1285,14 @@ TEST_F(DBTest2, CompactionStall) {
  // Wait for another compaction to be triggered
  TEST_SYNC_POINT("DBTest2::CompactionStall:1");

-  // Hold NotifyOnCompactionBegin in the unlock mutex section
-  TEST_SYNC_POINT("DBTest2::CompactionStall:2");
-
  // Hold NotifyOnCompactionCompleted in the unlock mutex section
-  TEST_SYNC_POINT("DBTest2::CompactionStall:3");
+  TEST_SYNC_POINT("DBTest2::CompactionStall:2");

  dbfull()->TEST_WaitForCompact();
  ASSERT_LT(NumTableFilesAtLevel(0),
            options.level0_file_num_compaction_trigger);
  ASSERT_GT(listener->compacted_files_cnt_.load(),
            10 - options.level0_file_num_compaction_trigger);
-  ASSERT_EQ(listener->compacting_files_cnt_.load(), listener->compacted_files_cnt_.load());

  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
 }
@@ -2352,9 +2321,9 @@ TEST_F(DBTest2, RateLimitedCompactionReads) {
        options.rate_limiter->GetTotalBytesThrough(Env::IO_LOW);
    // Include the explicit prefetch of the footer in direct I/O case.
    size_t direct_io_extra = use_direct_io ? 512 * 1024 : 0;
-    ASSERT_GE(
-        rate_limited_bytes,
-        static_cast<size_t>(kNumKeysPerFile * kBytesPerKey * kNumL0Files));
+    ASSERT_GE(rate_limited_bytes,
+              static_cast<size_t>(kNumKeysPerFile * kBytesPerKey * kNumL0Files +
+                                  direct_io_extra));
    ASSERT_LT(
        rate_limited_bytes,
        static_cast<size_t>(2 * kNumKeysPerFile * kBytesPerKey * kNumL0Files +
@@ -2531,231 +2500,6 @@ TEST_F(DBTest2, LiveFilesOmitObsoleteFiles) {
  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
 }

-TEST_F(DBTest2, TestNumPread) {
-  Options options = CurrentOptions();
-  // disable block cache
-  BlockBasedTableOptions table_options;
-  table_options.no_block_cache = true;
-  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
-  Reopen(options);
-  env_->count_random_reads_ = true;
-
-  env_->random_file_open_counter_.store(0);
-  ASSERT_OK(Put("bar", "foo"));
-  ASSERT_OK(Put("foo", "bar"));
-  ASSERT_OK(Flush());
-  // After flush, we'll open the file and read footer, meta block,
-  // property block and index block.
-  ASSERT_EQ(4, env_->random_read_counter_.Read());
-  ASSERT_EQ(1, env_->random_file_open_counter_.load());
-
-  // One pread per a normal data block read
-  env_->random_file_open_counter_.store(0);
-  env_->random_read_counter_.Reset();
-  ASSERT_EQ("bar", Get("foo"));
-  ASSERT_EQ(1, env_->random_read_counter_.Read());
-  // All files are already opened.
-  ASSERT_EQ(0, env_->random_file_open_counter_.load());
-
-  env_->random_file_open_counter_.store(0);
-  env_->random_read_counter_.Reset();
-  ASSERT_OK(Put("bar2", "foo2"));
-  ASSERT_OK(Put("foo2", "bar2"));
-  ASSERT_OK(Flush());
-  // After flush, we'll open the file and read footer, meta block,
-  // property block and index block.
-  ASSERT_EQ(4, env_->random_read_counter_.Read());
-  ASSERT_EQ(1, env_->random_file_open_counter_.load());
-
-  // Compaction needs two input blocks, which requires 2 preads, and
-  // generate a new SST file which needs 4 preads (footer, meta block,
-  // property block and index block). In total 6.
-  env_->random_file_open_counter_.store(0);
-  env_->random_read_counter_.Reset();
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
-  ASSERT_EQ(6, env_->random_read_counter_.Read());
-  // All compactin input files should have already been opened.
-  ASSERT_EQ(1, env_->random_file_open_counter_.load());
-
-  // One pread per a normal data block read
-  env_->random_file_open_counter_.store(0);
-  env_->random_read_counter_.Reset();
-  ASSERT_EQ("foo2", Get("bar2"));
-  ASSERT_EQ(1, env_->random_read_counter_.Read());
-  // SST files are already opened.
-  ASSERT_EQ(0, env_->random_file_open_counter_.load());
-}
-
-TEST_F(DBTest2, TraceAndReplay) {
-  Options options = CurrentOptions();
-  options.merge_operator = MergeOperators::CreatePutOperator();
-  ReadOptions ro;
-  WriteOptions wo;
-  TraceOptions trace_opts;
-  EnvOptions env_opts;
-  CreateAndReopenWithCF({"pikachu"}, options);
-  Random rnd(301);
-  Iterator* single_iter = nullptr;
-
-  std::string trace_filename = dbname_ + "/rocksdb.trace";
-  std::unique_ptr<TraceWriter> trace_writer;
-  ASSERT_OK(NewFileTraceWriter(env_, env_opts, trace_filename, &trace_writer));
-  ASSERT_OK(db_->StartTrace(trace_opts, std::move(trace_writer)));
-
-  ASSERT_OK(Put(0, "a", "1"));
-  ASSERT_OK(Merge(0, "b", "2"));
-  ASSERT_OK(Delete(0, "c"));
-  ASSERT_OK(SingleDelete(0, "d"));
-  ASSERT_OK(db_->DeleteRange(wo, dbfull()->DefaultColumnFamily(), "e", "f"));
-
-  WriteBatch batch;
-  ASSERT_OK(batch.Put("f", "11"));
-  ASSERT_OK(batch.Merge("g", "12"));
-  ASSERT_OK(batch.Delete("h"));
-  ASSERT_OK(batch.SingleDelete("i"));
-  ASSERT_OK(batch.DeleteRange("j", "k"));
-  ASSERT_OK(db_->Write(wo, &batch));
-
-  single_iter = db_->NewIterator(ro);
-  single_iter->Seek("f");
-  single_iter->SeekForPrev("g");
-  delete single_iter;
-
-  ASSERT_EQ("1", Get(0, "a"));
-  ASSERT_EQ("12", Get(0, "g"));
-
-  ASSERT_OK(Put(1, "foo", "bar"));
-  ASSERT_OK(Put(1, "rocksdb", "rocks"));
-  ASSERT_EQ("NOT_FOUND", Get(1, "leveldb"));
-
-  ASSERT_OK(db_->EndTrace());
-  // These should not get into the trace file as it is after EndTrace.
-  Put("hello", "world");
-  Merge("foo", "bar");
-
-  // Open another db, replay, and verify the data
-  std::string value;
-  std::string dbname2 = test::TmpDir(env_) + "/db_replay";
-  ASSERT_OK(DestroyDB(dbname2, options));
-
-  // Using a different name than db2, to pacify infer's use-after-lifetime
-  // warnings (http://fbinfer.com).
-  DB* db2_init = nullptr;
-  options.create_if_missing = true;
-  ASSERT_OK(DB::Open(options, dbname2, &db2_init));
-  ColumnFamilyHandle* cf;
-  ASSERT_OK(
-      db2_init->CreateColumnFamily(ColumnFamilyOptions(), "pikachu", &cf));
-  delete cf;
-  delete db2_init;
-
-  DB* db2 = nullptr;
-  std::vector<ColumnFamilyDescriptor> column_families;
-  ColumnFamilyOptions cf_options;
-  cf_options.merge_operator = MergeOperators::CreatePutOperator();
-  column_families.push_back(ColumnFamilyDescriptor("default", cf_options));
-  column_families.push_back(
-      ColumnFamilyDescriptor("pikachu", ColumnFamilyOptions()));
-  std::vector<ColumnFamilyHandle*> handles;
-  ASSERT_OK(DB::Open(DBOptions(), dbname2, column_families, &handles, &db2));
-
-  env_->SleepForMicroseconds(100);
-  // Verify that the keys don't already exist
-  ASSERT_TRUE(db2->Get(ro, handles[0], "a", &value).IsNotFound());
-  ASSERT_TRUE(db2->Get(ro, handles[0], "g", &value).IsNotFound());
-
-  std::unique_ptr<TraceReader> trace_reader;
-  ASSERT_OK(NewFileTraceReader(env_, env_opts, trace_filename, &trace_reader));
-  Replayer replayer(db2, handles_, std::move(trace_reader));
-  ASSERT_OK(replayer.Replay());
-
-  ASSERT_OK(db2->Get(ro, handles[0], "a", &value));
-  ASSERT_EQ("1", value);
-  ASSERT_OK(db2->Get(ro, handles[0], "g", &value));
-  ASSERT_EQ("12", value);
-  ASSERT_TRUE(db2->Get(ro, handles[0], "hello", &value).IsNotFound());
-  ASSERT_TRUE(db2->Get(ro, handles[0], "world", &value).IsNotFound());
-
-  ASSERT_OK(db2->Get(ro, handles[1], "foo", &value));
-  ASSERT_EQ("bar", value);
-  ASSERT_OK(db2->Get(ro, handles[1], "rocksdb", &value));
-  ASSERT_EQ("rocks", value);
-
-  for (auto handle : handles) {
-    delete handle;
-  }
-  delete db2;
-  ASSERT_OK(DestroyDB(dbname2, options));
-}
-
-TEST_F(DBTest2, TraceWithLimit) {
-  Options options = CurrentOptions();
-  options.merge_operator = MergeOperators::CreatePutOperator();
-  ReadOptions ro;
-  WriteOptions wo;
-  TraceOptions trace_opts;
-  EnvOptions env_opts;
-  CreateAndReopenWithCF({"pikachu"}, options);
-  Random rnd(301);
-
-  // test the max trace file size options
-  trace_opts.max_trace_file_size = 5;
-  std::string trace_filename = dbname_ + "/rocksdb.trace1";
-  std::unique_ptr<TraceWriter> trace_writer;
-  ASSERT_OK(NewFileTraceWriter(env_, env_opts, trace_filename, &trace_writer));
-  ASSERT_OK(db_->StartTrace(trace_opts, std::move(trace_writer)));
-  ASSERT_OK(Put(0, "a", "1"));
-  ASSERT_OK(Put(0, "b", "1"));
-  ASSERT_OK(Put(0, "c", "1"));
-  ASSERT_OK(db_->EndTrace());
-
-  std::string dbname2 = test::TmpDir(env_) + "/db_replay2";
-  std::string value;
-  ASSERT_OK(DestroyDB(dbname2, options));
-
-  // Using a different name than db2, to pacify infer's use-after-lifetime
-  // warnings (http://fbinfer.com).
-  DB* db2_init = nullptr;
-  options.create_if_missing = true;
-  ASSERT_OK(DB::Open(options, dbname2, &db2_init));
-  ColumnFamilyHandle* cf;
-  ASSERT_OK(
-      db2_init->CreateColumnFamily(ColumnFamilyOptions(), "pikachu", &cf));
-  delete cf;
-  delete db2_init;
-
-  DB* db2 = nullptr;
-  std::vector<ColumnFamilyDescriptor> column_families;
-  ColumnFamilyOptions cf_options;
-  cf_options.merge_operator = MergeOperators::CreatePutOperator();
-  column_families.push_back(ColumnFamilyDescriptor("default", cf_options));
-  column_families.push_back(
-      ColumnFamilyDescriptor("pikachu", ColumnFamilyOptions()));
-  std::vector<ColumnFamilyHandle*> handles;
-  ASSERT_OK(DB::Open(DBOptions(), dbname2, column_families, &handles, &db2));
-
-  env_->SleepForMicroseconds(100);
-  // Verify that the keys don't already exist
-  ASSERT_TRUE(db2->Get(ro, handles[0], "a", &value).IsNotFound());
-  ASSERT_TRUE(db2->Get(ro, handles[0], "b", &value).IsNotFound());
-  ASSERT_TRUE(db2->Get(ro, handles[0], "c", &value).IsNotFound());
-
-  std::unique_ptr<TraceReader> trace_reader;
-  ASSERT_OK(NewFileTraceReader(env_, env_opts, trace_filename, &trace_reader));
-  Replayer replayer(db2, handles_, std::move(trace_reader));
-  ASSERT_OK(replayer.Replay());
-
-  ASSERT_TRUE(db2->Get(ro, handles[0], "a", &value).IsNotFound());
-  ASSERT_TRUE(db2->Get(ro, handles[0], "b", &value).IsNotFound());
-  ASSERT_TRUE(db2->Get(ro, handles[0], "c", &value).IsNotFound());
-
-  for (auto handle : handles) {
-    delete handle;
-  }
-  delete db2;
-  ASSERT_OK(DestroyDB(dbname2, options));
-}
-
 #endif  // ROCKSDB_LITE

 TEST_F(DBTest2, PinnableSliceAndMmapReads) {
@@ -2803,291 +2547,6 @@ TEST_F(DBTest2, PinnableSliceAndMmapReads) {
 #endif
 }

-TEST_F(DBTest2, DISABLED_IteratorPinnedMemory) {
-  Options options = CurrentOptions();
-  options.create_if_missing = true;
-  options.statistics = rocksdb::CreateDBStatistics();
-  BlockBasedTableOptions bbto;
-  bbto.no_block_cache = false;
-  bbto.cache_index_and_filter_blocks = false;
-  bbto.block_cache = NewLRUCache(100000);
-  bbto.block_size = 400;  // small block size
-  options.table_factory.reset(new BlockBasedTableFactory(bbto));
-  Reopen(options);
-
-  Random rnd(301);
-  std::string v = RandomString(&rnd, 400);
-
-  // Since v is the size of a block, each key should take a block
-  // of 400+ bytes.
-  Put("1", v);
-  Put("3", v);
-  Put("5", v);
-  Put("7", v);
-  ASSERT_OK(Flush());
-
-  ASSERT_EQ(0, bbto.block_cache->GetPinnedUsage());
-
-  // Verify that iterators don't pin more than one data block in block cache
-  // at each time.
-  {
-    std::unique_ptr<Iterator> iter(db_->NewIterator(ReadOptions()));
-    iter->SeekToFirst();
-
-    for (int i = 0; i < 4; i++) {
-      ASSERT_TRUE(iter->Valid());
-      // Block cache should contain exactly one block.
-      ASSERT_GT(bbto.block_cache->GetPinnedUsage(), 0);
-      ASSERT_LT(bbto.block_cache->GetPinnedUsage(), 800);
-      iter->Next();
-    }
-    ASSERT_FALSE(iter->Valid());
-
-    iter->Seek("4");
-    ASSERT_TRUE(iter->Valid());
-
-    ASSERT_GT(bbto.block_cache->GetPinnedUsage(), 0);
-    ASSERT_LT(bbto.block_cache->GetPinnedUsage(), 800);
-
-    iter->Seek("3");
-    ASSERT_TRUE(iter->Valid());
-
-    ASSERT_GT(bbto.block_cache->GetPinnedUsage(), 0);
-    ASSERT_LT(bbto.block_cache->GetPinnedUsage(), 800);
-  }
-  ASSERT_EQ(0, bbto.block_cache->GetPinnedUsage());
-
-  // Test compaction case
-  Put("2", v);
-  Put("5", v);
-  Put("6", v);
-  Put("8", v);
-  ASSERT_OK(Flush());
-
-  // Clear existing data in block cache
-  bbto.block_cache->SetCapacity(0);
-  bbto.block_cache->SetCapacity(100000);
-
-  // Verify compaction input iterators don't hold more than one data blocks at
-  // one time.
-  std::atomic<bool> finished(false);
-  std::atomic<int> block_newed(0);
-  std::atomic<int> block_destroyed(0);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "Block::Block:0", [&](void* /*arg*/) {
-        if (finished) {
-          return;
-        }
-        // Two iterators. At most 2 outstanding blocks.
-        EXPECT_GE(block_newed.load(), block_destroyed.load());
-        EXPECT_LE(block_newed.load(), block_destroyed.load() + 1);
-        block_newed.fetch_add(1);
-      });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "Block::~Block", [&](void* /*arg*/) {
-        if (finished) {
-          return;
-        }
-        // Two iterators. At most 2 outstanding blocks.
-        EXPECT_GE(block_newed.load(), block_destroyed.load() + 1);
-        EXPECT_LE(block_newed.load(), block_destroyed.load() + 2);
-        block_destroyed.fetch_add(1);
-      });
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "CompactionJob::Run:BeforeVerify",
-      [&](void* /*arg*/) { finished = true; });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
-
-  // Two input files. Each of them has 4 data blocks.
-  ASSERT_EQ(8, block_newed.load());
-  ASSERT_EQ(8, block_destroyed.load());
-
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-}
-
-TEST_F(DBTest2, TestBBTTailPrefetch) {
-  std::atomic<bool> called(false);
-  size_t expected_lower_bound = 512 * 1024;
-  size_t expected_higher_bound = 512 * 1024;
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "BlockBasedTable::Open::TailPrefetchLen", [&](void* arg) {
-        size_t* prefetch_size = static_cast<size_t*>(arg);
-        EXPECT_LE(expected_lower_bound, *prefetch_size);
-        EXPECT_GE(expected_higher_bound, *prefetch_size);
-        called = true;
-      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  Put("1", "1");
-  Put("9", "1");
-  Flush();
-
-  expected_lower_bound = 0;
-  expected_higher_bound = 8 * 1024;
-
-  Put("1", "1");
-  Put("9", "1");
-  Flush();
-
-  Put("1", "1");
-  Put("9", "1");
-  Flush();
-
-  // Full compaction to make sure there is no L0 file after the open.
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
-
-  ASSERT_TRUE(called.load());
-  called = false;
-
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
-
-  std::atomic<bool> first_call(true);
-  rocksdb::SyncPoint::GetInstance()->SetCallBack(
-      "BlockBasedTable::Open::TailPrefetchLen", [&](void* arg) {
-        size_t* prefetch_size = static_cast<size_t*>(arg);
-        if (first_call) {
-          EXPECT_EQ(4 * 1024, *prefetch_size);
-          first_call = false;
-        } else {
-          EXPECT_GE(4 * 1024, *prefetch_size);
-        }
-        called = true;
-      });
-  rocksdb::SyncPoint::GetInstance()->EnableProcessing();
-
-  Options options = CurrentOptions();
-  options.max_file_opening_threads = 1;  // one thread
-  BlockBasedTableOptions table_options;
-  table_options.cache_index_and_filter_blocks = true;
-  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
-  options.max_open_files = -1;
-  Reopen(options);
-
-  Put("1", "1");
-  Put("9", "1");
-  Flush();
-
-  Put("1", "1");
-  Put("9", "1");
-  Flush();
-
-  ASSERT_TRUE(called.load());
-  called = false;
-
-  // Parallel loading SST files
-  options.max_file_opening_threads = 16;
-  Reopen(options);
-
-  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
-
-  ASSERT_TRUE(called.load());
-
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
-}
-
-TEST_F(DBTest2, TestGetColumnFamilyHandleUnlocked) {
-  // Setup sync point dependency to reproduce the race condition of
-  // DBImpl::GetColumnFamilyHandleUnlocked
-  rocksdb::SyncPoint::GetInstance()->LoadDependency(
-      { {"TestGetColumnFamilyHandleUnlocked::GetColumnFamilyHandleUnlocked1",
-         "TestGetColumnFamilyHandleUnlocked::PreGetColumnFamilyHandleUnlocked2"},
-        {"TestGetColumnFamilyHandleUnlocked::GetColumnFamilyHandleUnlocked2",
-         "TestGetColumnFamilyHandleUnlocked::ReadColumnFamilyHandle1"},
-      });
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  CreateColumnFamilies({"test1", "test2"}, Options());
-  ASSERT_EQ(handles_.size(), 2);
-
-  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
-  port::Thread user_thread1([&]() {
-    auto cfh = dbi->GetColumnFamilyHandleUnlocked(handles_[0]->GetID());
-    ASSERT_EQ(cfh->GetID(), handles_[0]->GetID());
-    TEST_SYNC_POINT("TestGetColumnFamilyHandleUnlocked::GetColumnFamilyHandleUnlocked1");
-    TEST_SYNC_POINT("TestGetColumnFamilyHandleUnlocked::ReadColumnFamilyHandle1");
-    ASSERT_EQ(cfh->GetID(), handles_[0]->GetID());
-  });
-
-  port::Thread user_thread2([&]() {
-    TEST_SYNC_POINT("TestGetColumnFamilyHandleUnlocked::PreGetColumnFamilyHandleUnlocked2");
-    auto cfh = dbi->GetColumnFamilyHandleUnlocked(handles_[1]->GetID());
-    ASSERT_EQ(cfh->GetID(), handles_[1]->GetID());
-    TEST_SYNC_POINT("TestGetColumnFamilyHandleUnlocked::GetColumnFamilyHandleUnlocked2");
-    ASSERT_EQ(cfh->GetID(), handles_[1]->GetID());
-  });
-
-  user_thread1.join();
-  user_thread2.join();
-
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
-}
-
-#ifndef ROCKSDB_LITE
-TEST_F(DBTest2, TestCompactFiles) {
-  // Setup sync point dependency to reproduce the race condition of
-  // DBImpl::GetColumnFamilyHandleUnlocked
-  rocksdb::SyncPoint::GetInstance()->LoadDependency({
-      {"TestCompactFiles::IngestExternalFile1",
-       "TestCompactFiles::IngestExternalFile2"},
-  });
-  SyncPoint::GetInstance()->EnableProcessing();
-
-  Options options;
-  options.num_levels = 2;
-  options.disable_auto_compactions = true;
-  Reopen(options);
-  auto* handle = db_->DefaultColumnFamily();
-  ASSERT_EQ(db_->NumberLevels(handle), 2);
-
-  rocksdb::SstFileWriter sst_file_writer{rocksdb::EnvOptions(), options};
-  std::string external_file1 = dbname_ + "/test_compact_files1.sst_t";
-  std::string external_file2 = dbname_ + "/test_compact_files2.sst_t";
-  std::string external_file3 = dbname_ + "/test_compact_files3.sst_t";
-
-  ASSERT_OK(sst_file_writer.Open(external_file1));
-  ASSERT_OK(sst_file_writer.Put("1", "1"));
-  ASSERT_OK(sst_file_writer.Put("2", "2"));
-  ASSERT_OK(sst_file_writer.Finish());
-
-  ASSERT_OK(sst_file_writer.Open(external_file2));
-  ASSERT_OK(sst_file_writer.Put("3", "3"));
-  ASSERT_OK(sst_file_writer.Put("4", "4"));
-  ASSERT_OK(sst_file_writer.Finish());
-
-  ASSERT_OK(sst_file_writer.Open(external_file3));
-  ASSERT_OK(sst_file_writer.Put("5", "5"));
-  ASSERT_OK(sst_file_writer.Put("6", "6"));
-  ASSERT_OK(sst_file_writer.Finish());
-
-  ASSERT_OK(db_->IngestExternalFile(handle, {external_file1, external_file3},
-                                    IngestExternalFileOptions()));
-  ASSERT_EQ(NumTableFilesAtLevel(1, 0), 2);
-  std::vector<std::string> files;
-  GetSstFiles(env_, dbname_, &files);
-  ASSERT_EQ(files.size(), 2);
-
-  port::Thread user_thread1(
-      [&]() { db_->CompactFiles(CompactionOptions(), handle, files, 1); });
-
-  port::Thread user_thread2([&]() {
-    ASSERT_OK(db_->IngestExternalFile(handle, {external_file2},
-                                      IngestExternalFileOptions()));
-    TEST_SYNC_POINT("TestCompactFiles::IngestExternalFile1");
-  });
-
-  user_thread1.join();
-  user_thread2.join();
-
-  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
-  rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
-}
-#endif  // ROCKSDB_LITE
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -259,47 +259,6 @@ bool DBTestBase::ChangeFilterOptions() {
  return true;
 }

-// Switch between different DB options for file ingestion tests.
-bool DBTestBase::ChangeOptionsForFileIngestionTest() {
-  if (option_config_ == kDefault) {
-    option_config_ = kUniversalCompaction;
-    Destroy(last_options_);
-    auto options = CurrentOptions();
-    options.create_if_missing = true;
-    TryReopen(options);
-    return true;
-  } else if (option_config_ == kUniversalCompaction) {
-    option_config_ = kUniversalCompactionMultiLevel;
-    Destroy(last_options_);
-    auto options = CurrentOptions();
-    options.create_if_missing = true;
-    TryReopen(options);
-    return true;
-  } else if (option_config_ == kUniversalCompactionMultiLevel) {
-    option_config_ = kLevelSubcompactions;
-    Destroy(last_options_);
-    auto options = CurrentOptions();
-    assert(options.max_subcompactions > 1);
-    TryReopen(options);
-    return true;
-  } else if (option_config_ == kLevelSubcompactions) {
-    option_config_ = kUniversalSubcompactions;
-    Destroy(last_options_);
-    auto options = CurrentOptions();
-    assert(options.max_subcompactions > 1);
-    TryReopen(options);
-    return true;
-  } else if (option_config_ == kUniversalSubcompactions) {
-    option_config_ = kDirectIO;
-    Destroy(last_options_);
-    auto options = CurrentOptions();
-    TryReopen(options);
-    return true;
-  } else {
-    return false;
-  }
-}
-
 // Return the current option configuration.
 Options DBTestBase::CurrentOptions(
    const anon::OptionsOverride& options_override) const {
@@ -471,10 +430,6 @@ Options DBTestBase::GetOptions(
      table_options.checksum = kxxHash;
      break;
    }
-    case kxxHash64Checksum: {
-      table_options.checksum = kxxHash64;
-      break;
-    }
    case kFIFOCompaction: {
      options.compaction_style = kCompactionStyleFIFO;
      break;
@@ -494,16 +449,15 @@ Options DBTestBase::GetOptions(
      options.prefix_extractor.reset(NewNoopTransform());
      break;
    }
-    case kBlockBasedTableWithPartitionedIndexFormat4: {
-      table_options.format_version = 4;
-      // Format 4 changes the binary index format. Since partitioned index is a
+    case kBlockBasedTableWithPartitionedIndexFormat3: {
+      table_options.format_version = 3;
+      // Format 3 changes the binary index format. Since partitioned index is a
      // super-set of simple indexes, we are also using kTwoLevelIndexSearch to
      // test this format.
      table_options.index_type = BlockBasedTableOptions::kTwoLevelIndexSearch;
-      // The top-level index in partition filters are also affected by format 4.
+      // The top-level index in partition filters are also affected by format 3.
      table_options.filter_policy.reset(NewBloomFilterPolicy(10, false));
      table_options.partition_filters = true;
-      table_options.index_block_restart_interval = 8;
      break;
    }
    case kBlockBasedTableWithIndexRestartInterval: {
@@ -653,19 +607,30 @@ Status DBTestBase::ReadOnlyReopen(const Options& options) {
 Status DBTestBase::TryReopen(const Options& options) {
  Close();
  last_options_.table_factory.reset();
-  // Note: operator= is an unsafe approach here since it destructs
-  // std::shared_ptr in the same order of their creation, in contrast to
-  // destructors which destructs them in the opposite order of creation. One
-  // particular problme is that the cache destructor might invoke callback
-  // functions that use Option members such as statistics. To work around this
-  // problem, we manually call destructor of table_facotry which eventually
-  // clears the block cache.
+  // Note: operator= is an unsafe approach here since it destructs shared_ptr in
+  // the same order of their creation, in contrast to destructors which
+  // destructs them in the opposite order of creation. One particular problme is
+  // that the cache destructor might invoke callback functions that use Option
+  // members such as statistics. To work around this problem, we manually call
+  // destructor of table_facotry which eventually clears the block cache.
  last_options_ = options;
  return DB::Open(options, dbname_, &db_);
 }

 bool DBTestBase::IsDirectIOSupported() {
-  return test::IsDirectIOSupported(env_, dbname_);
+  EnvOptions env_options;
+  env_options.use_mmap_writes = false;
+  env_options.use_direct_writes = true;
+  std::string tmp = TempFileName(dbname_, 999);
+  Status s;
+  {
+    unique_ptr<WritableFile> file;
+    s = env_->NewWritableFile(tmp, &file, env_options);
+  }
+  if (s.ok()) {
+    s = env_->DeleteFile(tmp);
+  }
+  return s.ok();
 }

 bool DBTestBase::IsMemoryMappedAccessSupported() const {
@@ -680,13 +645,6 @@ Status DBTestBase::Flush(int cf) {
  }
 }

-Status DBTestBase::Flush(const std::vector<int>& cf_ids) {
-  std::vector<ColumnFamilyHandle*> cfhs;
-  std::for_each(cf_ids.begin(), cf_ids.end(),
-                [&cfhs, this](int id) { cfhs.emplace_back(handles_[id]); });
-  return db_->Flush(FlushOptions(), cfhs);
-}
-
 Status DBTestBase::Put(const Slice& k, const Slice& v, WriteOptions wo) {
  if (kMergePut == option_config_) {
    return db_->Merge(wo, k, v);
@@ -814,15 +772,13 @@ std::string DBTestBase::AllEntriesFor(const Slice& user_key, int cf) {
  Arena arena;
  auto options = CurrentOptions();
  InternalKeyComparator icmp(options.comparator);
-  ReadRangeDelAggregator range_del_agg(&icmp,
-                                       kMaxSequenceNumber /* upper_bound */);
+  RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
  ScopedArenaIterator iter;
  if (cf == 0) {
-    iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg,
-                                           kMaxSequenceNumber));
+    iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg));
  } else {
-    iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg,
-                                           kMaxSequenceNumber, handles_[cf]));
+    iter.set(
+        dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[cf]));
  }
  InternalKey target(user_key, kMaxSequenceNumber, kTypeValue);
  iter->Seek(target.Encode());
@@ -1227,17 +1183,15 @@ void DBTestBase::validateNumberOfEntries(int numValues, int cf) {
  Arena arena;
  auto options = CurrentOptions();
  InternalKeyComparator icmp(options.comparator);
-  ReadRangeDelAggregator range_del_agg(&icmp,
-                                       kMaxSequenceNumber /* upper_bound */);
+  RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
  // This should be defined after range_del_agg so that it destructs the
  // assigned iterator before it range_del_agg is already destructed.
  ScopedArenaIterator iter;
  if (cf != 0) {
-    iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg,
-                                           kMaxSequenceNumber, handles_[cf]));
+    iter.set(
+        dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[cf]));
  } else {
-    iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg,
-                                           kMaxSequenceNumber));
+    iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg));
  }
  iter->SeekToFirst();
  ASSERT_EQ(iter->status().ok(), true);
@@ -1257,9 +1211,9 @@ void DBTestBase::validateNumberOfEntries(int numValues, int cf) {
 void DBTestBase::CopyFile(const std::string& source,
                          const std::string& destination, uint64_t size) {
  const EnvOptions soptions;
-  std::unique_ptr<SequentialFile> srcfile;
+  unique_ptr<SequentialFile> srcfile;
  ASSERT_OK(env_->NewSequentialFile(source, &srcfile, soptions));
-  std::unique_ptr<WritableFile> destfile;
+  unique_ptr<WritableFile> destfile;
  ASSERT_OK(env_->NewWritableFile(destination, &destfile, soptions));

  if (size == 0) {
@@ -1437,10 +1391,8 @@ void DBTestBase::VerifyDBInternal(
    std::vector<std::pair<std::string, std::string>> true_data) {
  Arena arena;
  InternalKeyComparator icmp(last_options_.comparator);
-  ReadRangeDelAggregator range_del_agg(&icmp,
-                                       kMaxSequenceNumber /* upper_bound */);
-  auto iter =
-      dbfull()->NewInternalIterator(&arena, &range_del_agg, kMaxSequenceNumber);
+  RangeDelAggregator range_del_agg(icmp, {});
+  auto iter = dbfull()->NewInternalIterator(&arena, &range_del_agg);
  iter->SeekToFirst();
  for (auto p : true_data) {
    ASSERT_TRUE(iter->Valid());
@@ -46,7 +46,6 @@
 #include "table/scoped_arena_iterator.h"
 #include "util/compression.h"
 #include "util/filename.h"
-#include "util/mock_time_env.h"
 #include "util/mutexlock.h"

 #include "util/string_util.h"
@@ -110,6 +109,8 @@ struct OptionsOverride {
  // These will be used only if filter_policy is set
  bool partition_filters = false;
  uint64_t metadata_block_size = 1024;
+  BlockBasedTableOptions::IndexType index_type =
+      BlockBasedTableOptions::IndexType::kBinarySearch;

  // Used as a bit mask of individual enums in which to skip an XF test point
  int skip_policy = 0;
@@ -169,7 +170,7 @@ class SpecialMemTableRep : public MemTableRep {
  virtual ~SpecialMemTableRep() override {}

 private:
-  std::unique_ptr<MemTableRep> memtable_;
+  unique_ptr<MemTableRep> memtable_;
  int num_entries_flush_;
  int num_entries_;
 };
@@ -207,15 +208,15 @@ class SpecialEnv : public EnvWrapper {
 public:
  explicit SpecialEnv(Env* base);

-  Status NewWritableFile(const std::string& f, std::unique_ptr<WritableFile>* r,
+  Status NewWritableFile(const std::string& f, unique_ptr<WritableFile>* r,
                         const EnvOptions& soptions) override {
    class SSTableFile : public WritableFile {
     private:
      SpecialEnv* env_;
-      std::unique_ptr<WritableFile> base_;
+      unique_ptr<WritableFile> base_;

     public:
-      SSTableFile(SpecialEnv* env, std::unique_ptr<WritableFile>&& base)
+      SSTableFile(SpecialEnv* env, unique_ptr<WritableFile>&& base)
          : env_(env), base_(std::move(base)) {}
      Status Append(const Slice& data) override {
        if (env_->table_write_callback_) {
@@ -295,7 +296,7 @@ class SpecialEnv : public EnvWrapper {
    };
    class ManifestFile : public WritableFile {
     public:
-      ManifestFile(SpecialEnv* env, std::unique_ptr<WritableFile>&& b)
+      ManifestFile(SpecialEnv* env, unique_ptr<WritableFile>&& b)
          : env_(env), base_(std::move(b)) {}
      Status Append(const Slice& data) override {
        if (env_->manifest_write_error_.load(std::memory_order_acquire)) {
@@ -316,17 +317,14 @@ class SpecialEnv : public EnvWrapper {
        }
      }
      uint64_t GetFileSize() override { return base_->GetFileSize(); }
-      Status Allocate(uint64_t offset, uint64_t len) override {
-        return base_->Allocate(offset, len);
-      }

     private:
      SpecialEnv* env_;
-      std::unique_ptr<WritableFile> base_;
+      unique_ptr<WritableFile> base_;
    };
    class WalFile : public WritableFile {
     public:
-      WalFile(SpecialEnv* env, std::unique_ptr<WritableFile>&& b)
+      WalFile(SpecialEnv* env, unique_ptr<WritableFile>&& b)
          : env_(env), base_(std::move(b)) {
        env_->num_open_wal_file_.fetch_add(1);
      }
@@ -372,13 +370,10 @@ class SpecialEnv : public EnvWrapper {
      bool IsSyncThreadSafe() const override {
        return env_->is_wal_sync_thread_safe_.load();
      }
-      Status Allocate(uint64_t offset, uint64_t len) override {
-        return base_->Allocate(offset, len);
-      }

     private:
      SpecialEnv* env_;
-      std::unique_ptr<WritableFile> base_;
+      unique_ptr<WritableFile> base_;
    };

    if (non_writeable_rate_.load(std::memory_order_acquire) > 0) {
@@ -420,11 +415,11 @@ class SpecialEnv : public EnvWrapper {
  }

  Status NewRandomAccessFile(const std::string& f,
-                             std::unique_ptr<RandomAccessFile>* r,
+                             unique_ptr<RandomAccessFile>* r,
                             const EnvOptions& soptions) override {
    class CountingFile : public RandomAccessFile {
     public:
-      CountingFile(std::unique_ptr<RandomAccessFile>&& target,
+      CountingFile(unique_ptr<RandomAccessFile>&& target,
                   anon::AtomicCounter* counter,
                   std::atomic<size_t>* bytes_read)
          : target_(std::move(target)),
@@ -439,7 +434,7 @@ class SpecialEnv : public EnvWrapper {
      }

     private:
-      std::unique_ptr<RandomAccessFile> target_;
+      unique_ptr<RandomAccessFile> target_;
      anon::AtomicCounter* counter_;
      std::atomic<size_t>* bytes_read_;
    };
@@ -457,11 +452,11 @@ class SpecialEnv : public EnvWrapper {
  }

  virtual Status NewSequentialFile(const std::string& f,
-                                   std::unique_ptr<SequentialFile>* r,
+                                   unique_ptr<SequentialFile>* r,
                                   const EnvOptions& soptions) override {
    class CountingFile : public SequentialFile {
     public:
-      CountingFile(std::unique_ptr<SequentialFile>&& target,
+      CountingFile(unique_ptr<SequentialFile>&& target,
                   anon::AtomicCounter* counter)
          : target_(std::move(target)), counter_(counter) {}
      virtual Status Read(size_t n, Slice* result, char* scratch) override {
@@ -471,7 +466,7 @@ class SpecialEnv : public EnvWrapper {
      virtual Status Skip(uint64_t n) override { return target_->Skip(n); }

     private:
-      std::unique_ptr<SequentialFile> target_;
+      unique_ptr<SequentialFile> target_;
      anon::AtomicCounter* counter_;
    };

@@ -574,13 +569,44 @@ class SpecialEnv : public EnvWrapper {

  std::atomic<int> delete_count_;

-  std::atomic<bool> time_elapse_only_sleep_;
+  bool time_elapse_only_sleep_;

  bool no_slowdown_;

  std::atomic<bool> is_wal_sync_thread_safe_{true};

-  std::atomic<size_t> compaction_readahead_size_{};
+  std::atomic<size_t> compaction_readahead_size_;
+};
+
+class MockTimeEnv : public EnvWrapper {
+ public:
+  explicit MockTimeEnv(Env* base) : EnvWrapper(base) {}
+
+  virtual Status GetCurrentTime(int64_t* time) override {
+    assert(time != nullptr);
+    assert(current_time_ <=
+           static_cast<uint64_t>(std::numeric_limits<int64_t>::max()));
+    *time = static_cast<int64_t>(current_time_);
+    return Status::OK();
+  }
+
+  virtual uint64_t NowMicros() override {
+    assert(current_time_ <= std::numeric_limits<uint64_t>::max() / 1000000);
+    return current_time_ * 1000000;
+  }
+
+  virtual uint64_t NowNanos() override {
+    assert(current_time_ <= std::numeric_limits<uint64_t>::max() / 1000000000);
+    return current_time_ * 1000000000;
+  }
+
+  void set_current_time(uint64_t time) {
+    assert(time >= current_time_);
+    current_time_ = time;
+  }
+
+ private:
+  std::atomic<uint64_t> current_time_{0};
 };

 #ifndef ROCKSDB_LITE
@@ -672,10 +698,9 @@ class DBTestBase : public testing::Test {
    kLevelSubcompactions,
    kBlockBasedTableWithIndexRestartInterval,
    kBlockBasedTableWithPartitionedIndex,
-    kBlockBasedTableWithPartitionedIndexFormat4,
+    kBlockBasedTableWithPartitionedIndexFormat3,
    kPartitionedFilterWithNewTableReaderForCompactions,
    kUniversalSubcompactions,
-    kxxHash64Checksum,
    // This must be the last line
    kEnd,
  };
@@ -747,9 +772,6 @@ class DBTestBase : public testing::Test {
  // Jump from kDefault to kFilter to kFullFilter
  bool ChangeFilterOptions();

-  // Switch between different DB options for file ingestion tests.
-  bool ChangeOptionsForFileIngestionTest();
-
  // Return the current option configuration.
  Options CurrentOptions(const anon::OptionsOverride& options_override =
                             anon::OptionsOverride()) const;
@@ -803,8 +825,6 @@ class DBTestBase : public testing::Test {

  Status Flush(int cf = 0);

-  Status Flush(const std::vector<int>& cf_ids);
-
  Status Put(const Slice& k, const Slice& v, WriteOptions wo = WriteOptions());

  Status Put(int cf, const Slice& k, const Slice& v,
@@ -1154,7 +1154,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionCompressRatio2) {
    dbfull()->TEST_WaitForFlushMemTable();
    dbfull()->TEST_WaitForCompact();
  }
-  ASSERT_LT(TotalSize(), 120000U * 12 * 0.82 + 120000 * 2);
+  ASSERT_LT(TotalSize(), 120000U * 12 * 0.8 + 120000 * 2);
 }

 #ifndef ROCKSDB_VALGRIND_RUN
@@ -1824,7 +1824,7 @@ TEST_P(DBTestUniversalCompaction, FinalSortedRunCompactFilesConflict) {

  port::Thread compact_files_thread([&]() {
    ASSERT_OK(dbfull()->CompactFiles(CompactionOptions(), default_cfh,
-                                     {first_sst_filename}, num_levels_ - 1));
+        {first_sst_filename}, num_levels_ - 1));
  });

  TEST_SYNC_POINT(
@@ -18,23 +18,13 @@ namespace rocksdb {
 class DBWALTest : public DBTestBase {
 public:
  DBWALTest() : DBTestBase("/db_wal_test") {}
-
-#if defined(ROCKSDB_PLATFORM_POSIX)
-  uint64_t GetAllocatedFileSize(std::string file_name) {
-    struct stat sbuf;
-    int err = stat(file_name.c_str(), &sbuf);
-    assert(err == 0);
-    return sbuf.st_blocks * 512;
-  }
-#endif
 };

 // A SpecialEnv enriched to give more insight about deleted files
 class EnrichedSpecialEnv : public SpecialEnv {
 public:
  explicit EnrichedSpecialEnv(Env* base) : SpecialEnv(base) {}
-  Status NewSequentialFile(const std::string& f,
-                           std::unique_ptr<SequentialFile>* r,
+  Status NewSequentialFile(const std::string& f, unique_ptr<SequentialFile>* r,
                           const EnvOptions& soptions) override {
    InstrumentedMutexLock l(&env_mutex_);
    if (f == skipped_wal) {
@@ -89,7 +79,6 @@ class DBWALTestWithEnrichedEnv : public DBTestBase {
    enriched_env_ = new EnrichedSpecialEnv(env_->target());
    auto options = CurrentOptions();
    options.env = enriched_env_;
-    options.allow_2pc = true;
    Reopen(options);
    delete env_;
    // to be deleted by the parent class
@@ -804,12 +793,12 @@ class RecoveryTestHelper {

    *count = 0;

-    std::shared_ptr<Cache> table_cache = NewLRUCache(50, 0);
+    shared_ptr<Cache> table_cache = NewLRUCache(50, 0);
    EnvOptions env_options;
    WriteBufferManager write_buffer_manager(db_options.db_write_buffer_size);

-    std::unique_ptr<VersionSet> versions;
-    std::unique_ptr<WalManager> wal_manager;
+    unique_ptr<VersionSet> versions;
+    unique_ptr<WalManager> wal_manager;
    WriteController write_controller;

    versions.reset(new VersionSet(test->dbname_, &db_options, env_options,
@@ -823,10 +812,10 @@ class RecoveryTestHelper {
    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;
+      unique_ptr<WritableFile> file;
      ASSERT_OK(db_options.env->NewWritableFile(fname, &file, env_options));
-      std::unique_ptr<WritableFileWriter> file_writer(
-          new WritableFileWriter(std::move(file), fname, env_options));
+      unique_ptr<WritableFileWriter> file_writer(
+          new WritableFileWriter(std::move(file), env_options));
      current_log_writer.reset(
          new log::Writer(std::move(file_writer), current_log_number,
                          db_options.recycle_log_file_num > 0));
@@ -1340,99 +1329,6 @@ TEST_F(DBWALTest, RecoverFromCorruptedWALWithoutFlush) {
  }
 }

-// Tests that total log size is recovered if we set
-// avoid_flush_during_recovery=true.
-// Flush should trigger if max_total_wal_size is reached.
-TEST_F(DBWALTest, RestoreTotalLogSizeAfterRecoverWithoutFlush) {
-  class TestFlushListener : public EventListener {
-   public:
-    std::atomic<int> count{0};
-
-    TestFlushListener() = default;
-
-    void OnFlushBegin(DB* /*db*/, const FlushJobInfo& flush_job_info) override {
-      count++;
-      assert(FlushReason::kWriteBufferManager == flush_job_info.flush_reason);
-    }
-  };
-  std::shared_ptr<TestFlushListener> test_listener =
-      std::make_shared<TestFlushListener>();
-
-  constexpr size_t kKB = 1024;
-  constexpr size_t kMB = 1024 * 1024;
-  Options options = CurrentOptions();
-  options.avoid_flush_during_recovery = true;
-  options.max_total_wal_size = 1 * kMB;
-  options.listeners.push_back(test_listener);
-  // Have to open DB in multi-CF mode to trigger flush when
-  // max_total_wal_size is reached.
-  CreateAndReopenWithCF({"one"}, options);
-  // Write some keys and we will end up with one log file which is slightly
-  // smaller than 1MB.
-  std::string value_100k(100 * kKB, 'v');
-  std::string value_300k(300 * kKB, 'v');
-  ASSERT_OK(Put(0, "foo", "v1"));
-  for (int i = 0; i < 9; i++) {
-    ASSERT_OK(Put(1, "key" + ToString(i), value_100k));
-  }
-  // Get log files before reopen.
-  VectorLogPtr log_files_before;
-  ASSERT_OK(dbfull()->GetSortedWalFiles(log_files_before));
-  ASSERT_EQ(1, log_files_before.size());
-  uint64_t log_size_before = log_files_before[0]->SizeFileBytes();
-  ASSERT_GT(log_size_before, 900 * kKB);
-  ASSERT_LT(log_size_before, 1 * kMB);
-  ReopenWithColumnFamilies({"default", "one"}, options);
-  // Write one more value to make log larger than 1MB.
-  ASSERT_OK(Put(1, "bar", value_300k));
-  // Get log files again. A new log file will be opened.
-  VectorLogPtr log_files_after_reopen;
-  ASSERT_OK(dbfull()->GetSortedWalFiles(log_files_after_reopen));
-  ASSERT_EQ(2, log_files_after_reopen.size());
-  ASSERT_EQ(log_files_before[0]->LogNumber(),
-            log_files_after_reopen[0]->LogNumber());
-  ASSERT_GT(log_files_after_reopen[0]->SizeFileBytes() +
-                log_files_after_reopen[1]->SizeFileBytes(),
-            1 * kMB);
-  // Write one more key to trigger flush.
-  ASSERT_OK(Put(0, "foo", "v2"));
-  dbfull()->TEST_WaitForFlushMemTable();
-  // Flushed two column families.
-  ASSERT_EQ(2, test_listener->count.load());
-}
-
-#if defined(ROCKSDB_PLATFORM_POSIX)
-#if defined(ROCKSDB_FALLOCATE_PRESENT)
-// Tests that we will truncate the preallocated space of the last log from
-// previous.
-TEST_F(DBWALTest, TruncateLastLogAfterRecoverWithoutFlush) {
-  constexpr size_t kKB = 1024;
-  Options options = CurrentOptions();
-  options.avoid_flush_during_recovery = true;
-  DestroyAndReopen(options);
-  size_t preallocated_size =
-      dbfull()->TEST_GetWalPreallocateBlockSize(options.write_buffer_size);
-  ASSERT_OK(Put("foo", "v1"));
-  VectorLogPtr log_files_before;
-  ASSERT_OK(dbfull()->GetSortedWalFiles(log_files_before));
-  ASSERT_EQ(1, log_files_before.size());
-  auto& file_before = log_files_before[0];
-  ASSERT_LT(file_before->SizeFileBytes(), 1 * kKB);
-  // The log file has preallocated space.
-  ASSERT_GE(GetAllocatedFileSize(dbname_ + file_before->PathName()),
-            preallocated_size);
-  Reopen(options);
-  VectorLogPtr log_files_after;
-  ASSERT_OK(dbfull()->GetSortedWalFiles(log_files_after));
-  ASSERT_EQ(1, log_files_after.size());
-  ASSERT_LT(log_files_after[0]->SizeFileBytes(), 1 * kKB);
-  // The preallocated space should be truncated.
-  ASSERT_LT(GetAllocatedFileSize(dbname_ + file_before->PathName()),
-            preallocated_size);
-}
-#endif  // ROCKSDB_FALLOCATE_PRESENT
-#endif  // ROCKSDB_PLATFORM_POSIX
-
 #endif  // ROCKSDB_LITE

 TEST_F(DBWALTest, WalTermTest) {
@@ -48,8 +48,6 @@ EntryType GetEntryType(ValueType value_type) {
      return kEntryMerge;
    case kTypeRangeDeletion:
      return kEntryRangeDeletion;
-    case kTypeBlobIndex:
-      return kEntryBlobIndex;
    default:
      return kEntryOther;
  }
@@ -674,16 +674,5 @@ int InternalKeyComparator::CompareKeySeq(const Slice& akey,
  return r;
 }

-struct ParsedInternalKeyComparator {
-  explicit ParsedInternalKeyComparator(const InternalKeyComparator* c)
-      : cmp(c) {}
-
-  bool operator()(const ParsedInternalKey& a,
-                  const ParsedInternalKey& b) const {
-    return cmp->Compare(a, b) < 0;
-  }
-
-  const InternalKeyComparator* cmp;
-};

 }  // namespace rocksdb
@@ -4,9 +4,7 @@
 //  (found in the LICENSE.Apache file in the root directory).
 //
 #include "db/error_handler.h"
-#include "db/db_impl.h"
 #include "db/event_helpers.h"
-#include "util/sst_file_manager_impl.h"

 namespace rocksdb {

@@ -35,7 +33,7 @@ std::map<std::tuple<BackgroundErrorReason, Status::Code, Status::SubCode, bool>,
        // Errors during BG flush
        {std::make_tuple(BackgroundErrorReason::kFlush, Status::Code::kIOError,
                         Status::SubCode::kNoSpace, true),
-         Status::Severity::kHardError},
+         Status::Severity::kSoftError},
        {std::make_tuple(BackgroundErrorReason::kFlush, Status::Code::kIOError,
                         Status::SubCode::kNoSpace, false),
         Status::Severity::kNoError},
@@ -46,11 +44,11 @@ std::map<std::tuple<BackgroundErrorReason, Status::Code, Status::SubCode, bool>,
        {std::make_tuple(BackgroundErrorReason::kWriteCallback,
                         Status::Code::kIOError, Status::SubCode::kNoSpace,
                         true),
-         Status::Severity::kHardError},
+         Status::Severity::kFatalError},
        {std::make_tuple(BackgroundErrorReason::kWriteCallback,
                         Status::Code::kIOError, Status::SubCode::kNoSpace,
                         false),
-         Status::Severity::kHardError},
+         Status::Severity::kFatalError},
 };

 std::map<std::tuple<BackgroundErrorReason, Status::Code, bool>, Status::Severity>
@@ -120,45 +118,6 @@ std::map<std::tuple<BackgroundErrorReason, bool>, Status::Severity>
          Status::Severity::kFatalError},
 };

-void ErrorHandler::CancelErrorRecovery() {
-#ifndef ROCKSDB_LITE
-  db_mutex_->AssertHeld();
-
-  // We'll release the lock before calling sfm, so make sure no new
-  // recovery gets scheduled at that point
-  auto_recovery_ = false;
-  SstFileManagerImpl* sfm = reinterpret_cast<SstFileManagerImpl*>(
-      db_options_.sst_file_manager.get());
-  if (sfm) {
-    // This may or may not cancel a pending recovery
-    db_mutex_->Unlock();
-    bool cancelled = sfm->CancelErrorRecovery(this);
-    db_mutex_->Lock();
-    if (cancelled) {
-      recovery_in_prog_ = false;
-    }
-  }
-#endif
-}
-
-// This is the main function for looking at an error during a background
-// operation and deciding the severity, and error recovery strategy. The high
-// level algorithm is as follows -
-// 1. Classify the severity of the error based on the ErrorSeverityMap,
-//    DefaultErrorSeverityMap and DefaultReasonMap defined earlier
-// 2. Call a Status code specific override function to adjust the severity
-//    if needed. The reason for this is our ability to recover may depend on
-//    the exact options enabled in DBOptions
-// 3. Determine if auto recovery is possible. A listener notification callback
-//    is called, which can disable the auto recovery even if we decide its
-//    feasible
-// 4. For Status::NoSpace() errors, rely on SstFileManagerImpl to control
-//    the actual recovery. If no sst file manager is specified in DBOptions,
-//    a default one is allocated during DB::Open(), so there will always be
-//    one.
-// This can also get called as part of a recovery operation. In that case, we
-// also track the error separately in recovery_error_ so we can tell in the
-// end whether recovery succeeded or not
 Status ErrorHandler::SetBGError(const Status& bg_err, BackgroundErrorReason reason) {
  db_mutex_->AssertHeld();

@@ -166,12 +125,6 @@ Status ErrorHandler::SetBGError(const Status& bg_err, BackgroundErrorReason reas
    return Status::OK();
  }

-  // Check if recovery is currently in progress. If it is, we will save this
-  // error so we can check it at the end to see if recovery succeeded or not
-  if (recovery_in_prog_ && recovery_error_.ok()) {
-    recovery_error_ = bg_err;
-  }
-
  bool paranoid = db_options_.paranoid_checks;
  Status::Severity sev = Status::Severity::kFatalError;
  Status new_bg_err;
@@ -203,143 +156,15 @@ Status ErrorHandler::SetBGError(const Status& bg_err, BackgroundErrorReason reas
  }

  new_bg_err = Status(bg_err, sev);
-
-  bool auto_recovery = auto_recovery_;
-  if (new_bg_err.severity() >= Status::Severity::kFatalError && auto_recovery) {
-    auto_recovery = false;
-    ;
-  }
-
-  // Allow some error specific overrides
-  if (new_bg_err == Status::NoSpace()) {
-    new_bg_err = OverrideNoSpaceError(new_bg_err, &auto_recovery);
-  }
-
  if (!new_bg_err.ok()) {
    Status s = new_bg_err;
-    EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason, &s,
-                                          db_mutex_, &auto_recovery);
+    EventHelpers::NotifyOnBackgroundError(db_options_.listeners, reason, &s, db_mutex_);
    if (!s.ok() && (s.severity() > bg_error_.severity())) {
      bg_error_ = s;
-    } else {
-      // This error is less severe than previously encountered error. Don't
-      // take any further action
-      return bg_error_;
    }
  }

-  if (auto_recovery) {
-    recovery_in_prog_ = true;
-
-    // Kick-off error specific recovery
-    if (bg_error_ == Status::NoSpace()) {
-      RecoverFromNoSpace();
-    }
-  }
  return bg_error_;
 }

-Status ErrorHandler::OverrideNoSpaceError(Status bg_error,
-                                          bool* auto_recovery) {
-#ifndef ROCKSDB_LITE
-  if (bg_error.severity() >= Status::Severity::kFatalError) {
-    return bg_error;
-  }
-
-  if (db_options_.sst_file_manager.get() == nullptr) {
-    // We rely on SFM to poll for enough disk space and recover
-    *auto_recovery = false;
-    return bg_error;
-  }
-
-  if (db_options_.allow_2pc &&
-      (bg_error.severity() <= Status::Severity::kSoftError)) {
-    // Don't know how to recover, as the contents of the current WAL file may
-    // be inconsistent, and it may be needed for 2PC. If 2PC is not enabled,
-    // we can just flush the memtable and discard the log
-    *auto_recovery = false;
-    return Status(bg_error, Status::Severity::kFatalError);
-  }
-
-  {
-    uint64_t free_space;
-    if (db_options_.env->GetFreeSpace(db_options_.db_paths[0].path,
-                                      &free_space) == Status::NotSupported()) {
-      *auto_recovery = false;
-    }
-  }
-
-  return bg_error;
-#else
-  (void)auto_recovery;
-  return Status(bg_error, Status::Severity::kFatalError);
-#endif
-}
-
-void ErrorHandler::RecoverFromNoSpace() {
-#ifndef ROCKSDB_LITE
-  SstFileManagerImpl* sfm =
-      reinterpret_cast<SstFileManagerImpl*>(db_options_.sst_file_manager.get());
-
-  // Inform SFM of the error, so it can kick-off the recovery
-  if (sfm) {
-    sfm->StartErrorRecovery(this, bg_error_);
-  }
-#endif
-}
-
-Status ErrorHandler::ClearBGError() {
-#ifndef ROCKSDB_LITE
-  db_mutex_->AssertHeld();
-
-  // Signal that recovery succeeded
-  if (recovery_error_.ok()) {
-    Status old_bg_error = bg_error_;
-    bg_error_ = Status::OK();
-    recovery_in_prog_ = false;
-    EventHelpers::NotifyOnErrorRecoveryCompleted(db_options_.listeners,
-                                                 old_bg_error, db_mutex_);
-  }
-  return recovery_error_;
-#else
-  return bg_error_;
-#endif
-}
-
-Status ErrorHandler::RecoverFromBGError(bool is_manual) {
-#ifndef ROCKSDB_LITE
-  InstrumentedMutexLock l(db_mutex_);
-  if (is_manual) {
-    // If its a manual recovery and there's a background recovery in progress
-    // return busy status
-    if (recovery_in_prog_) {
-      return Status::Busy();
-    }
-    recovery_in_prog_ = true;
-  }
-
-  if (bg_error_.severity() == Status::Severity::kSoftError) {
-    // Simply clear the background error and return
-    recovery_error_ = Status::OK();
-    return ClearBGError();
-  }
-
-  // Reset recovery_error_. We will use this to record any errors that happen
-  // during the recovery process. While recovering, the only operations that
-  // can generate background errors should be the flush operations
-  recovery_error_ = Status::OK();
-  Status s = db_->ResumeImpl();
-  // For manual recover, shutdown, and fatal error  cases, set
-  // recovery_in_prog_ to false. For automatic background recovery, leave it
-  // as is regardless of success or failure as it will be retried
-  if (is_manual || s.IsShutdownInProgress() ||
-      bg_error_.severity() >= Status::Severity::kFatalError) {
-    recovery_in_prog_ = false;
-  }
-  return s;
-#else
-  (void)is_manual;
-  return bg_error_;
-#endif
-}
 }
@@ -11,65 +11,42 @@

 namespace rocksdb {

-class DBImpl;
-
 class ErrorHandler {
  public:
-   ErrorHandler(DBImpl* db, const ImmutableDBOptions& db_options,
-                InstrumentedMutex* db_mutex)
-       : db_(db),
-         db_options_(db_options),
-         bg_error_(Status::OK()),
-         recovery_error_(Status::OK()),
-         db_mutex_(db_mutex),
-         auto_recovery_(false),
-         recovery_in_prog_(false) {}
-   ~ErrorHandler() {}
+    ErrorHandler(const ImmutableDBOptions& db_options,
+        InstrumentedMutex* db_mutex)
+      : db_options_(db_options),
+        bg_error_(Status::OK()),
+        db_mutex_(db_mutex)
+      {}
+    ~ErrorHandler() {}

-   void EnableAutoRecovery() { auto_recovery_ = true; }
+    Status::Severity GetErrorSeverity(BackgroundErrorReason reason,
+        Status::Code code, Status::SubCode subcode);

-   Status::Severity GetErrorSeverity(BackgroundErrorReason reason,
-                                     Status::Code code,
-                                     Status::SubCode subcode);
+    Status SetBGError(const Status& bg_err, BackgroundErrorReason reason);

-   Status SetBGError(const Status& bg_err, BackgroundErrorReason reason);
+    Status GetBGError()
+    {
+      return bg_error_;
+    }

-   Status GetBGError() { return bg_error_; }
+    void ClearBGError() {
+      bg_error_ = Status::OK();
+    }

-   Status GetRecoveryError() { return recovery_error_; }
-
-   Status ClearBGError();
-
-   bool IsDBStopped() {
-     return !bg_error_.ok() &&
-            bg_error_.severity() >= Status::Severity::kHardError;
+    bool IsDBStopped() {
+      return !bg_error_.ok();
    }

    bool IsBGWorkStopped() {
-      return !bg_error_.ok() &&
-             (bg_error_.severity() >= Status::Severity::kHardError ||
-              !auto_recovery_);
+      return !bg_error_.ok();
    }

-    bool IsRecoveryInProgress() { return recovery_in_prog_; }
-
-    Status RecoverFromBGError(bool is_manual = false);
-    void CancelErrorRecovery();
-
-   private:
-    DBImpl* db_;
+  private:
    const ImmutableDBOptions& db_options_;
    Status bg_error_;
-    // A separate Status variable used to record any errors during the
-    // recovery process from hard errors
-    Status recovery_error_;
    InstrumentedMutex* db_mutex_;
-    // A flag indicating whether automatic recovery from errors is enabled
-    bool auto_recovery_;
-    bool recovery_in_prog_;
-
-    Status OverrideNoSpaceError(Status bg_error, bool* auto_recovery);
-    void RecoverFromNoSpace();
 };

 }
@@ -6,12 +6,9 @@
 // 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.
-#ifndef ROCKSDB_LITE
-
 #include "db/db_test_util.h"
 #include "port/stack_trace.h"
 #include "rocksdb/perf_context.h"
-#include "rocksdb/sst_file_manager.h"
 #include "util/fault_injection_test_env.h"
 #if !defined(ROCKSDB_LITE)
 #include "util/sync_point.h"
@@ -36,139 +33,36 @@ class DBErrorHandlingEnv : public EnvWrapper {
    bool trig_io_error;
 };

-class ErrorHandlerListener : public EventListener {
- public:
-  ErrorHandlerListener()
-      : mutex_(),
-        cv_(&mutex_),
-        no_auto_recovery_(false),
-        recovery_complete_(false),
-        file_creation_started_(false),
-        override_bg_error_(false),
-        file_count_(0),
-        fault_env_(nullptr) {}
-
-  void OnTableFileCreationStarted(
-      const TableFileCreationBriefInfo& /*ti*/) override {
-    InstrumentedMutexLock l(&mutex_);
-    file_creation_started_ = true;
-    if (file_count_ > 0) {
-      if (--file_count_ == 0) {
-        fault_env_->SetFilesystemActive(false, file_creation_error_);
-        file_creation_error_ = Status::OK();
-      }
-    }
-    cv_.SignalAll();
-  }
-
-  void OnErrorRecoveryBegin(BackgroundErrorReason /*reason*/,
-                            Status /*bg_error*/,
-                            bool* auto_recovery) override {
-    if (*auto_recovery && no_auto_recovery_) {
-      *auto_recovery = false;
-    }
-  }
-
-  void OnErrorRecoveryCompleted(Status /*old_bg_error*/) override {
-    InstrumentedMutexLock l(&mutex_);
-    recovery_complete_ = true;
-    cv_.SignalAll();
-  }
-
-  bool WaitForRecovery(uint64_t /*abs_time_us*/) {
-    InstrumentedMutexLock l(&mutex_);
-    while (!recovery_complete_) {
-      cv_.Wait(/*abs_time_us*/);
-    }
-    if (recovery_complete_) {
-      recovery_complete_ = false;
-      return true;
-    }
-    return false;
-  }
-
-  void WaitForTableFileCreationStarted(uint64_t /*abs_time_us*/) {
-    InstrumentedMutexLock l(&mutex_);
-    while (!file_creation_started_) {
-      cv_.Wait(/*abs_time_us*/);
-    }
-    file_creation_started_ = false;
-  }
-
-  void OnBackgroundError(BackgroundErrorReason /*reason*/,
-                         Status* bg_error) override {
-    if (override_bg_error_) {
-      *bg_error = bg_error_;
-      override_bg_error_ = false;
-    }
-  }
-
-  void EnableAutoRecovery(bool enable = true) { no_auto_recovery_ = !enable; }
-
-  void OverrideBGError(Status bg_err) {
-    bg_error_ = bg_err;
-    override_bg_error_ = true;
-  }
-
-  void InjectFileCreationError(FaultInjectionTestEnv* env, int file_count,
-                               Status s) {
-    fault_env_ = env;
-    file_count_ = file_count;
-    file_creation_error_ = s;
-  }
-
- private:
-  InstrumentedMutex mutex_;
-  InstrumentedCondVar cv_;
-  bool no_auto_recovery_;
-  bool recovery_complete_;
-  bool file_creation_started_;
-  bool override_bg_error_;
-  int file_count_;
-  Status file_creation_error_;
-  Status bg_error_;
-  FaultInjectionTestEnv* fault_env_;
-};
-
 TEST_F(DBErrorHandlingTest, FLushWriteError) {
  std::unique_ptr<FaultInjectionTestEnv> fault_env(
      new FaultInjectionTestEnv(Env::Default()));
-  std::shared_ptr<ErrorHandlerListener> listener(new ErrorHandlerListener());
  Options options = GetDefaultOptions();
  options.create_if_missing = true;
  options.env = fault_env.get();
-  options.listeners.emplace_back(listener);
  Status s;
-
-  listener->EnableAutoRecovery(false);
  DestroyAndReopen(options);

-  Put(Key(0), "val");
+  Put(Key(0), "va;");
  SyncPoint::GetInstance()->SetCallBack(
      "FlushJob::Start", [&](void *) {
    fault_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
  });
  SyncPoint::GetInstance()->EnableProcessing();
  s = Flush();
-  ASSERT_EQ(s.severity(), rocksdb::Status::Severity::kHardError);
-  SyncPoint::GetInstance()->DisableProcessing();
+  ASSERT_EQ(s.severity(), rocksdb::Status::Severity::kSoftError);
  fault_env->SetFilesystemActive(true);
  s = dbfull()->Resume();
  ASSERT_EQ(s, Status::OK());

-  Reopen(options);
-  ASSERT_EQ("val", Get(Key(0)));
  Destroy(options);
 }

 TEST_F(DBErrorHandlingTest, CompactionWriteError) {
  std::unique_ptr<FaultInjectionTestEnv> fault_env(
      new FaultInjectionTestEnv(Env::Default()));
-  std::shared_ptr<ErrorHandlerListener> listener(new ErrorHandlerListener());
  Options options = GetDefaultOptions();
  options.create_if_missing = true;
  options.level0_file_num_compaction_trigger = 2;
-  options.listeners.emplace_back(listener);
  options.env = fault_env.get();
  Status s;
  DestroyAndReopen(options);
@@ -178,10 +72,6 @@ TEST_F(DBErrorHandlingTest, CompactionWriteError) {
  s = Flush();
  ASSERT_EQ(s, Status::OK());

-  listener->OverrideBGError(
-      Status(Status::NoSpace(), Status::Severity::kHardError)
-      );
-  listener->EnableAutoRecovery(false);
  rocksdb::SyncPoint::GetInstance()->LoadDependency(
      {{"FlushMemTableFinished", "BackgroundCallCompaction:0"}});
  rocksdb::SyncPoint::GetInstance()->SetCallBack(
@@ -195,7 +85,7 @@ TEST_F(DBErrorHandlingTest, CompactionWriteError) {
  ASSERT_EQ(s, Status::OK());

  s = dbfull()->TEST_WaitForCompact();
-  ASSERT_EQ(s.severity(), rocksdb::Status::Severity::kHardError);
+  ASSERT_EQ(s.severity(), rocksdb::Status::Severity::kSoftError);

  fault_env->SetFilesystemActive(true);
  s = dbfull()->Resume();
@@ -239,439 +129,6 @@ TEST_F(DBErrorHandlingTest, CorruptionError) {
  Destroy(options);
 }

-TEST_F(DBErrorHandlingTest, AutoRecoverFlushError) {
-  std::unique_ptr<FaultInjectionTestEnv> fault_env(
-      new FaultInjectionTestEnv(Env::Default()));
-  std::shared_ptr<ErrorHandlerListener> listener(new ErrorHandlerListener());
-  Options options = GetDefaultOptions();
-  options.create_if_missing = true;
-  options.env = fault_env.get();
-  options.listeners.emplace_back(listener);
-  Status s;
-
-  listener->EnableAutoRecovery();
-  DestroyAndReopen(options);
-
-  Put(Key(0), "val");
-  SyncPoint::GetInstance()->SetCallBack("FlushJob::Start", [&](void*) {
-    fault_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
-  });
-  SyncPoint::GetInstance()->EnableProcessing();
-  s = Flush();
-  ASSERT_EQ(s.severity(), rocksdb::Status::Severity::kHardError);
-  SyncPoint::GetInstance()->DisableProcessing();
-  fault_env->SetFilesystemActive(true);
-  ASSERT_EQ(listener->WaitForRecovery(5000000), true);
-
-  s = Put(Key(1), "val");
-  ASSERT_EQ(s, Status::OK());
-
-  Reopen(options);
-  ASSERT_EQ("val", Get(Key(0)));
-  ASSERT_EQ("val", Get(Key(1)));
-  Destroy(options);
-}
-
-TEST_F(DBErrorHandlingTest, FailRecoverFlushError) {
-  std::unique_ptr<FaultInjectionTestEnv> fault_env(
-      new FaultInjectionTestEnv(Env::Default()));
-  std::shared_ptr<ErrorHandlerListener> listener(new ErrorHandlerListener());
-  Options options = GetDefaultOptions();
-  options.create_if_missing = true;
-  options.env = fault_env.get();
-  options.listeners.emplace_back(listener);
-  Status s;
-
-  listener->EnableAutoRecovery();
-  DestroyAndReopen(options);
-
-  Put(Key(0), "val");
-  SyncPoint::GetInstance()->SetCallBack("FlushJob::Start", [&](void*) {
-    fault_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
-  });
-  SyncPoint::GetInstance()->EnableProcessing();
-  s = Flush();
-  ASSERT_EQ(s.severity(), rocksdb::Status::Severity::kHardError);
-  // We should be able to shutdown the database while auto recovery is going
-  // on in the background
-  Close();
-  DestroyDB(dbname_, options);
-}
-
-TEST_F(DBErrorHandlingTest, WALWriteError) {
-  std::unique_ptr<FaultInjectionTestEnv> fault_env(
-      new FaultInjectionTestEnv(Env::Default()));
-  std::shared_ptr<ErrorHandlerListener> listener(new ErrorHandlerListener());
-  Options options = GetDefaultOptions();
-  options.create_if_missing = true;
-  options.writable_file_max_buffer_size = 32768;
-  options.env = fault_env.get();
-  options.listeners.emplace_back(listener);
-  Status s;
-  Random rnd(301);
-
-  listener->EnableAutoRecovery();
-  DestroyAndReopen(options);
-
-  {
-    WriteBatch batch;
-
-    for (auto i = 0; i<100; ++i) {
-      batch.Put(Key(i), RandomString(&rnd, 1024));
-    }
-
-    WriteOptions wopts;
-    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
-  };
-
-  {
-    WriteBatch batch;
-    int write_error = 0;
-
-    for (auto i = 100; i<199; ++i) {
-      batch.Put(Key(i), RandomString(&rnd, 1024));
-    }
-
-    SyncPoint::GetInstance()->SetCallBack("WritableFileWriter::Append:BeforePrepareWrite", [&](void*) {
-      write_error++;
-      if (write_error > 2) {
-        fault_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
-      }
-    });
-    SyncPoint::GetInstance()->EnableProcessing();
-    WriteOptions wopts;
-    wopts.sync = true;
-    s = dbfull()->Write(wopts, &batch);
-    ASSERT_EQ(s, s.NoSpace());
-  }
-  SyncPoint::GetInstance()->DisableProcessing();
-  fault_env->SetFilesystemActive(true);
-  ASSERT_EQ(listener->WaitForRecovery(5000000), true);
-  for (auto i=0; i<199; ++i) {
-    if (i < 100) {
-      ASSERT_NE(Get(Key(i)), "NOT_FOUND");
-    } else {
-      ASSERT_EQ(Get(Key(i)), "NOT_FOUND");
-    }
-  }
-  Reopen(options);
-  for (auto i=0; i<199; ++i) {
-    if (i < 100) {
-      ASSERT_NE(Get(Key(i)), "NOT_FOUND");
-    } else {
-      ASSERT_EQ(Get(Key(i)), "NOT_FOUND");
-    }
-  }
-  Close();
-}
-
-TEST_F(DBErrorHandlingTest, MultiCFWALWriteError) {
-  std::unique_ptr<FaultInjectionTestEnv> fault_env(
-      new FaultInjectionTestEnv(Env::Default()));
-  std::shared_ptr<ErrorHandlerListener> listener(new ErrorHandlerListener());
-  Options options = GetDefaultOptions();
-  options.create_if_missing = true;
-  options.writable_file_max_buffer_size = 32768;
-  options.env = fault_env.get();
-  options.listeners.emplace_back(listener);
-  Status s;
-  Random rnd(301);
-
-  listener->EnableAutoRecovery();
-  CreateAndReopenWithCF({"one", "two", "three"}, options);
-
-  {
-    WriteBatch batch;
-
-    for (auto i = 1; i < 4; ++i) {
-      for (auto j = 0; j < 100; ++j) {
-        batch.Put(handles_[i], Key(j), RandomString(&rnd, 1024));
-      }
-    }
-
-    WriteOptions wopts;
-    wopts.sync = true;
-    ASSERT_EQ(dbfull()->Write(wopts, &batch), Status::OK());
-  };
-
-  {
-    WriteBatch batch;
-    int write_error = 0;
-
-    // Write to one CF
-    for (auto i = 100; i < 199; ++i) {
-      batch.Put(handles_[2], Key(i), RandomString(&rnd, 1024));
-    }
-
-    SyncPoint::GetInstance()->SetCallBack(
-        "WritableFileWriter::Append:BeforePrepareWrite", [&](void*) {
-          write_error++;
-          if (write_error > 2) {
-            fault_env->SetFilesystemActive(false,
-                                           Status::NoSpace("Out of space"));
-          }
-        });
-    SyncPoint::GetInstance()->EnableProcessing();
-    WriteOptions wopts;
-    wopts.sync = true;
-    s = dbfull()->Write(wopts, &batch);
-    ASSERT_EQ(s, s.NoSpace());
-  }
-  SyncPoint::GetInstance()->DisableProcessing();
-  fault_env->SetFilesystemActive(true);
-  ASSERT_EQ(listener->WaitForRecovery(5000000), true);
-
-  for (auto i = 1; i < 4; ++i) {
-    // Every CF should have been flushed
-    ASSERT_EQ(NumTableFilesAtLevel(0, i), 1);
-  }
-
-  for (auto i = 1; i < 4; ++i) {
-    for (auto j = 0; j < 199; ++j) {
-      if (j < 100) {
-        ASSERT_NE(Get(i, Key(j)), "NOT_FOUND");
-      } else {
-        ASSERT_EQ(Get(i, Key(j)), "NOT_FOUND");
-      }
-    }
-  }
-  ReopenWithColumnFamilies({"default", "one", "two", "three"}, options);
-  for (auto i = 1; i < 4; ++i) {
-    for (auto j = 0; j < 199; ++j) {
-      if (j < 100) {
-        ASSERT_NE(Get(i, Key(j)), "NOT_FOUND");
-      } else {
-        ASSERT_EQ(Get(i, Key(j)), "NOT_FOUND");
-      }
-    }
-  }
-  Close();
-}
-
-TEST_F(DBErrorHandlingTest, MultiDBCompactionError) {
-  FaultInjectionTestEnv* def_env = new FaultInjectionTestEnv(Env::Default());
-  std::vector<std::unique_ptr<FaultInjectionTestEnv>> fault_env;
-  std::vector<Options> options;
-  std::vector<std::shared_ptr<ErrorHandlerListener>> listener;
-  std::vector<DB*> db;
-  std::shared_ptr<SstFileManager> sfm(NewSstFileManager(def_env));
-  int kNumDbInstances = 3;
-  Random rnd(301);
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    listener.emplace_back(new ErrorHandlerListener());
-    options.emplace_back(GetDefaultOptions());
-    fault_env.emplace_back(new FaultInjectionTestEnv(Env::Default()));
-    options[i].create_if_missing = true;
-    options[i].level0_file_num_compaction_trigger = 2;
-    options[i].writable_file_max_buffer_size = 32768;
-    options[i].env = fault_env[i].get();
-    options[i].listeners.emplace_back(listener[i]);
-    options[i].sst_file_manager = sfm;
-    DB* dbptr;
-    char buf[16];
-
-    listener[i]->EnableAutoRecovery();
-    // Setup for returning error for the 3rd SST, which would be level 1
-    listener[i]->InjectFileCreationError(fault_env[i].get(), 3,
-                                         Status::NoSpace("Out of space"));
-    snprintf(buf, sizeof(buf), "_%d", i);
-    DestroyDB(dbname_ + std::string(buf), options[i]);
-    ASSERT_EQ(DB::Open(options[i], dbname_ + std::string(buf), &dbptr),
-              Status::OK());
-    db.emplace_back(dbptr);
-  }
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    WriteBatch batch;
-
-    for (auto j = 0; j <= 100; ++j) {
-      batch.Put(Key(j), RandomString(&rnd, 1024));
-    }
-
-    WriteOptions wopts;
-    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
-  }
-
-  def_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    WriteBatch batch;
-
-    // Write to one CF
-    for (auto j = 100; j < 199; ++j) {
-      batch.Put(Key(j), RandomString(&rnd, 1024));
-    }
-
-    WriteOptions wopts;
-    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
-  }
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    Status s = static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true);
-    ASSERT_EQ(s.severity(), Status::Severity::kSoftError);
-    fault_env[i]->SetFilesystemActive(true);
-  }
-
-  def_env->SetFilesystemActive(true);
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    std::string prop;
-    ASSERT_EQ(listener[i]->WaitForRecovery(5000000), true);
-    EXPECT_TRUE(db[i]->GetProperty(
-        "rocksdb.num-files-at-level" + NumberToString(0), &prop));
-    EXPECT_EQ(atoi(prop.c_str()), 0);
-    EXPECT_TRUE(db[i]->GetProperty(
-        "rocksdb.num-files-at-level" + NumberToString(1), &prop));
-    EXPECT_EQ(atoi(prop.c_str()), 1);
-  }
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    char buf[16];
-    snprintf(buf, sizeof(buf), "_%d", i);
-    delete db[i];
-    fault_env[i]->SetFilesystemActive(true);
-    if (getenv("KEEP_DB")) {
-      printf("DB is still at %s%s\n", dbname_.c_str(), buf);
-    } else {
-      Status s = DestroyDB(dbname_ + std::string(buf), options[i]);
-    }
-  }
-  options.clear();
-  sfm.reset();
-  delete def_env;
-}
-
-TEST_F(DBErrorHandlingTest, MultiDBVariousErrors) {
-  FaultInjectionTestEnv* def_env = new FaultInjectionTestEnv(Env::Default());
-  std::vector<std::unique_ptr<FaultInjectionTestEnv>> fault_env;
-  std::vector<Options> options;
-  std::vector<std::shared_ptr<ErrorHandlerListener>> listener;
-  std::vector<DB*> db;
-  std::shared_ptr<SstFileManager> sfm(NewSstFileManager(def_env));
-  int kNumDbInstances = 3;
-  Random rnd(301);
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    listener.emplace_back(new ErrorHandlerListener());
-    options.emplace_back(GetDefaultOptions());
-    fault_env.emplace_back(new FaultInjectionTestEnv(Env::Default()));
-    options[i].create_if_missing = true;
-    options[i].level0_file_num_compaction_trigger = 2;
-    options[i].writable_file_max_buffer_size = 32768;
-    options[i].env = fault_env[i].get();
-    options[i].listeners.emplace_back(listener[i]);
-    options[i].sst_file_manager = sfm;
-    DB* dbptr;
-    char buf[16];
-
-    listener[i]->EnableAutoRecovery();
-    switch (i) {
-      case 0:
-        // Setup for returning error for the 3rd SST, which would be level 1
-        listener[i]->InjectFileCreationError(fault_env[i].get(), 3,
-                                             Status::NoSpace("Out of space"));
-        break;
-      case 1:
-        // Setup for returning error after the 1st SST, which would result
-        // in a hard error
-        listener[i]->InjectFileCreationError(fault_env[i].get(), 2,
-                                             Status::NoSpace("Out of space"));
-        break;
-      default:
-        break;
-    }
-    snprintf(buf, sizeof(buf), "_%d", i);
-    DestroyDB(dbname_ + std::string(buf), options[i]);
-    ASSERT_EQ(DB::Open(options[i], dbname_ + std::string(buf), &dbptr),
-              Status::OK());
-    db.emplace_back(dbptr);
-  }
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    WriteBatch batch;
-
-    for (auto j = 0; j <= 100; ++j) {
-      batch.Put(Key(j), RandomString(&rnd, 1024));
-    }
-
-    WriteOptions wopts;
-    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
-  }
-
-  def_env->SetFilesystemActive(false, Status::NoSpace("Out of space"));
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    WriteBatch batch;
-
-    // Write to one CF
-    for (auto j = 100; j < 199; ++j) {
-      batch.Put(Key(j), RandomString(&rnd, 1024));
-    }
-
-    WriteOptions wopts;
-    wopts.sync = true;
-    ASSERT_EQ(db[i]->Write(wopts, &batch), Status::OK());
-    if (i != 1) {
-      ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::OK());
-    } else {
-      ASSERT_EQ(db[i]->Flush(FlushOptions()), Status::NoSpace());
-    }
-  }
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    Status s = static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true);
-    switch (i) {
-      case 0:
-        ASSERT_EQ(s.severity(), Status::Severity::kSoftError);
-        break;
-      case 1:
-        ASSERT_EQ(s.severity(), Status::Severity::kHardError);
-        break;
-      case 2:
-        ASSERT_EQ(s, Status::OK());
-        break;
-    }
-    fault_env[i]->SetFilesystemActive(true);
-  }
-
-  def_env->SetFilesystemActive(true);
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    std::string prop;
-    if (i < 2) {
-      ASSERT_EQ(listener[i]->WaitForRecovery(5000000), true);
-    }
-    if (i == 1) {
-      ASSERT_EQ(static_cast<DBImpl*>(db[i])->TEST_WaitForCompact(true),
-                Status::OK());
-    }
-    EXPECT_TRUE(db[i]->GetProperty(
-        "rocksdb.num-files-at-level" + NumberToString(0), &prop));
-    EXPECT_EQ(atoi(prop.c_str()), 0);
-    EXPECT_TRUE(db[i]->GetProperty(
-        "rocksdb.num-files-at-level" + NumberToString(1), &prop));
-    EXPECT_EQ(atoi(prop.c_str()), 1);
-  }
-
-  for (auto i = 0; i < kNumDbInstances; ++i) {
-    char buf[16];
-    snprintf(buf, sizeof(buf), "_%d", i);
-    fault_env[i]->SetFilesystemActive(true);
-    delete db[i];
-    if (getenv("KEEP_DB")) {
-      printf("DB is still at %s%s\n", dbname_.c_str(), buf);
-    } else {
-      DestroyDB(dbname_ + std::string(buf), options[i]);
-    }
-  }
-  options.clear();
-  delete def_env;
-}
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -679,13 +136,3 @@ int main(int argc, char** argv) {
  ::testing::InitGoogleTest(&argc, argv);
  return RUN_ALL_TESTS();
 }
-
-#else
-#include <stdio.h>
-
-int main(int /*argc*/, char** /*argv*/) {
-  fprintf(stderr, "SKIPPED as Cuckoo table is not supported in ROCKSDB_LITE\n");
-  return 0;
-}
-
-#endif  // ROCKSDB_LITE
@@ -40,8 +40,8 @@ void EventHelpers::NotifyTableFileCreationStarted(

 void EventHelpers::NotifyOnBackgroundError(
    const std::vector<std::shared_ptr<EventListener>>& listeners,
-    BackgroundErrorReason reason, Status* bg_error, InstrumentedMutex* db_mutex,
-    bool* auto_recovery) {
+    BackgroundErrorReason reason, Status* bg_error,
+    InstrumentedMutex* db_mutex) {
 #ifndef ROCKSDB_LITE
  if (listeners.size() == 0U) {
    return;
@@ -51,9 +51,6 @@ void EventHelpers::NotifyOnBackgroundError(
  db_mutex->Unlock();
  for (auto& listener : listeners) {
    listener->OnBackgroundError(reason, bg_error);
-    if (*auto_recovery) {
-      listener->OnErrorRecoveryBegin(reason, *bg_error, auto_recovery);
-    }
  }
  db_mutex->Lock();
 #else
@@ -61,7 +58,6 @@ void EventHelpers::NotifyOnBackgroundError(
  (void)reason;
  (void)bg_error;
  (void)db_mutex;
-  (void)auto_recovery;
 #endif  // ROCKSDB_LITE
 }

@@ -171,25 +167,4 @@ void EventHelpers::LogAndNotifyTableFileDeletion(
 #endif  // !ROCKSDB_LITE
 }

-void EventHelpers::NotifyOnErrorRecoveryCompleted(
-    const std::vector<std::shared_ptr<EventListener>>& listeners,
-    Status old_bg_error, InstrumentedMutex* db_mutex) {
-#ifndef ROCKSDB_LITE
-  if (listeners.size() == 0U) {
-    return;
-  }
-  db_mutex->AssertHeld();
-  // release lock while notifying events
-  db_mutex->Unlock();
-  for (auto& listener : listeners) {
-    listener->OnErrorRecoveryCompleted(old_bg_error);
-  }
-  db_mutex->Lock();
-#else
-  (void)listeners;
-  (void)old_bg_error;
-  (void)db_mutex;
-#endif  // ROCKSDB_LITE
-}
-
 }  // namespace rocksdb
@@ -28,7 +28,7 @@ class EventHelpers {
  static void NotifyOnBackgroundError(
      const std::vector<std::shared_ptr<EventListener>>& listeners,
      BackgroundErrorReason reason, Status* bg_error,
-      InstrumentedMutex* db_mutex, bool* auto_recovery);
+      InstrumentedMutex* db_mutex);
  static void LogAndNotifyTableFileCreationFinished(
      EventLogger* event_logger,
      const std::vector<std::shared_ptr<EventListener>>& listeners,
@@ -41,9 +41,6 @@ class EventHelpers {
      uint64_t file_number, const std::string& file_path,
      const Status& status, const std::string& db_name,
      const std::vector<std::shared_ptr<EventListener>>& listeners);
-  static void NotifyOnErrorRecoveryCompleted(
-      const std::vector<std::shared_ptr<EventListener>>& listeners,
-      Status bg_error, InstrumentedMutex* db_mutex);

 private:
  static void LogAndNotifyTableFileCreation(
@@ -14,8 +14,7 @@
 namespace rocksdb {

 #ifndef ROCKSDB_LITE
-class ExternalSSTFileBasicTest : public DBTestBase,
-                                 public ::testing::WithParamInterface<bool> {
+class ExternalSSTFileBasicTest : public DBTestBase {
 public:
  ExternalSSTFileBasicTest() : DBTestBase("/external_sst_file_test") {
    sst_files_dir_ = dbname_ + "/sst_files/";
@@ -42,7 +41,7 @@ class ExternalSSTFileBasicTest : public DBTestBase,
      const Options options, std::vector<int> keys,
      const std::vector<ValueType>& value_types,
      std::vector<std::pair<int, int>> range_deletions, int file_id,
-      bool write_global_seqno, std::map<std::string, std::string>* true_data) {
+      std::map<std::string, std::string>* true_data) {
    assert(value_types.size() == 1 || keys.size() == value_types.size());
    std::string file_path = sst_files_dir_ + ToString(file_id);
    SstFileWriter sst_file_writer(EnvOptions(), options);
@@ -106,7 +105,6 @@ class ExternalSSTFileBasicTest : public DBTestBase,
    if (s.ok()) {
      IngestExternalFileOptions ifo;
      ifo.allow_global_seqno = true;
-      ifo.write_global_seqno = write_global_seqno;
      s = db_->IngestExternalFile({file_path}, ifo);
    }
    return s;
@@ -115,18 +113,17 @@ class ExternalSSTFileBasicTest : public DBTestBase,
  Status GenerateAndAddExternalFile(
      const Options options, std::vector<int> keys,
      const std::vector<ValueType>& value_types, int file_id,
-      bool write_global_seqno, std::map<std::string, std::string>* true_data) {
+      std::map<std::string, std::string>* true_data) {
    return GenerateAndAddExternalFile(options, keys, value_types, {}, file_id,
-                                      write_global_seqno, true_data);
+                                      true_data);
  }

  Status GenerateAndAddExternalFile(
      const Options options, std::vector<int> keys, const ValueType value_type,
-      int file_id, bool write_global_seqno,
-      std::map<std::string, std::string>* true_data) {
+      int file_id, std::map<std::string, std::string>* true_data) {
    return GenerateAndAddExternalFile(options, keys,
                                      std::vector<ValueType>(1, value_type),
-                                      file_id, write_global_seqno, true_data);
+                                      file_id, true_data);
  }

  ~ExternalSSTFileBasicTest() { test::DestroyDir(env_, sst_files_dir_); }
@@ -248,8 +245,7 @@ TEST_F(ExternalSSTFileBasicTest, NoCopy) {
  }
 }

-TEST_P(ExternalSSTFileBasicTest, IngestFileWithGlobalSeqnoPickedSeqno) {
-  bool write_global_seqno = GetParam();
+TEST_F(ExternalSSTFileBasicTest, IngestFileWithGlobalSeqnoPickedSeqno) {
  do {
    Options options = CurrentOptions();
    DestroyAndReopen(options);
@@ -259,37 +255,33 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithGlobalSeqnoPickedSeqno) {

    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2, 3, 4, 5, 6},
                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 0);

    ASSERT_OK(GenerateAndAddExternalFile(options, {10, 11, 12, 13},
                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 0);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 4, 6},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {1, 4, 6}, ValueType::kTypeValue, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 1);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {11, 15, 19},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {11, 15, 19}, ValueType::kTypeValue, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 2);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {120, 130},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {120, 130}, ValueType::kTypeValue, file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 2);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 130},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {1, 130}, ValueType::kTypeValue, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 3);

@@ -300,21 +292,18 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithGlobalSeqnoPickedSeqno) {
    }
    SequenceNumber last_seqno = dbfull()->GetLatestSequenceNumber();

-    ASSERT_OK(GenerateAndAddExternalFile(options, {60, 61, 62},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {60, 61, 62}, ValueType::kTypeValue, file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {40, 41, 42},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {40, 41, 42}, ValueType::kTypeValue, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 1);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {20, 30, 40},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {20, 30, 40}, ValueType::kTypeValue, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 2);

@@ -322,39 +311,35 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithGlobalSeqnoPickedSeqno) {

    // We will need a seqno for the file regardless if the file overwrite
    // keys in the DB or not because we have a snapshot
-    ASSERT_OK(GenerateAndAddExternalFile(options, {1000, 1002},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {1000, 1002}, ValueType::kTypeValue, file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 3);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {2000, 3002},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {2000, 3002}, ValueType::kTypeValue, file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 4);

    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 20, 40, 100, 150},
                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 5);

    db_->ReleaseSnapshot(snapshot);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {5000, 5001},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {5000, 5001}, ValueType::kTypeValue, file_id++, &true_data));
    // No snapshot anymore, no need to assign a seqno
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 5);

    size_t kcnt = 0;
    VerifyDBFromMap(true_data, &kcnt, false);
-  } while (ChangeOptionsForFileIngestionTest());
+  } while (ChangeCompactOptions());
 }

-TEST_P(ExternalSSTFileBasicTest, IngestFileWithMultipleValueType) {
-  bool write_global_seqno = GetParam();
+TEST_F(ExternalSSTFileBasicTest, IngestFileWithMultipleValueType) {
  do {
    Options options = CurrentOptions();
    options.merge_operator.reset(new TestPutOperator());
@@ -365,57 +350,52 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMultipleValueType) {

    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2, 3, 4, 5, 6},
                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 0);

    ASSERT_OK(GenerateAndAddExternalFile(options, {10, 11, 12, 13},
                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 0);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 4, 6},
-                                         ValueType::kTypeMerge, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {1, 4, 6}, ValueType::kTypeMerge, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 1);

    ASSERT_OK(GenerateAndAddExternalFile(options, {11, 15, 19},
                                         ValueType::kTypeDeletion, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 2);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {120, 130},
-                                         ValueType::kTypeMerge, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {120, 130}, ValueType::kTypeMerge, file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 2);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 130},
-                                         ValueType::kTypeDeletion, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {1, 130}, ValueType::kTypeDeletion, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 3);

-    ASSERT_OK(GenerateAndAddExternalFile(
-        options, {120}, {ValueType::kTypeValue}, {{120, 135}}, file_id++,
-        write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(options, {120},
+                                         {ValueType::kTypeValue}, {{120, 135}},
+                                         file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 4);

    ASSERT_OK(GenerateAndAddExternalFile(options, {}, {}, {{110, 120}},
-                                         file_id++, write_global_seqno,
-                                         &true_data));
+                                         file_id++, &true_data));
    // The range deletion ends on a key, but it doesn't actually delete
    // this key because the largest key in the range is exclusive. Still,
    // it counts as an overlap so a new seqno will be assigned.
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 5);

    ASSERT_OK(GenerateAndAddExternalFile(options, {}, {}, {{100, 109}},
-                                         file_id++, write_global_seqno,
-                                         &true_data));
+                                         file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 5);

@@ -426,21 +406,19 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMultipleValueType) {
    }
    SequenceNumber last_seqno = dbfull()->GetLatestSequenceNumber();

-    ASSERT_OK(GenerateAndAddExternalFile(options, {60, 61, 62},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {60, 61, 62}, ValueType::kTypeValue, file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {40, 41, 42},
-                                         ValueType::kTypeMerge, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {40, 41, 42}, ValueType::kTypeMerge, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 1);

    ASSERT_OK(GenerateAndAddExternalFile(options, {20, 30, 40},
                                         ValueType::kTypeDeletion, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 2);

@@ -448,39 +426,35 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMultipleValueType) {

    // We will need a seqno for the file regardless if the file overwrite
    // keys in the DB or not because we have a snapshot
-    ASSERT_OK(GenerateAndAddExternalFile(options, {1000, 1002},
-                                         ValueType::kTypeMerge, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {1000, 1002}, ValueType::kTypeMerge, file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 3);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {2000, 3002},
-                                         ValueType::kTypeMerge, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {2000, 3002}, ValueType::kTypeMerge, file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 4);

    ASSERT_OK(GenerateAndAddExternalFile(options, {1, 20, 40, 100, 150},
                                         ValueType::kTypeMerge, file_id++,
-                                         write_global_seqno, &true_data));
+                                         &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 5);

    db_->ReleaseSnapshot(snapshot);

-    ASSERT_OK(GenerateAndAddExternalFile(options, {5000, 5001},
-                                         ValueType::kTypeValue, file_id++,
-                                         write_global_seqno, &true_data));
+    ASSERT_OK(GenerateAndAddExternalFile(
+        options, {5000, 5001}, ValueType::kTypeValue, file_id++, &true_data));
    // No snapshot anymore, no need to assign a seqno
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 5);

    size_t kcnt = 0;
    VerifyDBFromMap(true_data, &kcnt, false);
-  } while (ChangeOptionsForFileIngestionTest());
+  } while (ChangeCompactOptions());
 }

-TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
-  bool write_global_seqno = GetParam();
+TEST_F(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
  do {
    Options options = CurrentOptions();
    options.merge_operator.reset(new TestPutOperator());
@@ -493,7 +467,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {1, 2, 3, 4, 5, 6},
        {ValueType::kTypeValue, ValueType::kTypeMerge, ValueType::kTypeValue,
         ValueType::kTypeMerge, ValueType::kTypeValue, ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 0);

@@ -501,35 +475,33 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {10, 11, 12, 13},
        {ValueType::kTypeValue, ValueType::kTypeMerge, ValueType::kTypeValue,
         ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 0);

    ASSERT_OK(GenerateAndAddExternalFile(
-        options, {1, 4, 6},
-        {ValueType::kTypeDeletion, ValueType::kTypeValue,
-         ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        options, {1, 4, 6}, {ValueType::kTypeDeletion, ValueType::kTypeValue,
+                             ValueType::kTypeMerge},
+        file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 1);

    ASSERT_OK(GenerateAndAddExternalFile(
-        options, {11, 15, 19},
-        {ValueType::kTypeDeletion, ValueType::kTypeMerge,
-         ValueType::kTypeValue},
-        file_id++, write_global_seqno, &true_data));
+        options, {11, 15, 19}, {ValueType::kTypeDeletion, ValueType::kTypeMerge,
+                                ValueType::kTypeValue},
+        file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 2);

    ASSERT_OK(GenerateAndAddExternalFile(
        options, {120, 130}, {ValueType::kTypeValue, ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 2);

    ASSERT_OK(GenerateAndAddExternalFile(
        options, {1, 130}, {ValueType::kTypeMerge, ValueType::kTypeDeletion},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 3);

@@ -537,14 +509,14 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {150, 151, 152},
        {ValueType::kTypeValue, ValueType::kTypeMerge,
         ValueType::kTypeDeletion},
-        {{150, 160}, {180, 190}}, file_id++, write_global_seqno, &true_data));
+        {{150, 160}, {180, 190}}, file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 3);

    ASSERT_OK(GenerateAndAddExternalFile(
        options, {150, 151, 152},
        {ValueType::kTypeValue, ValueType::kTypeMerge, ValueType::kTypeValue},
-        {{200, 250}}, file_id++, write_global_seqno, &true_data));
+        {{200, 250}}, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 4);

@@ -552,7 +524,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {300, 301, 302},
        {ValueType::kTypeValue, ValueType::kTypeMerge,
         ValueType::kTypeDeletion},
-        {{1, 2}, {152, 154}}, file_id++, write_global_seqno, &true_data));
+        {{1, 2}, {152, 154}}, file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), 5);

@@ -566,7 +538,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
    ASSERT_OK(GenerateAndAddExternalFile(
        options, {60, 61, 62},
        {ValueType::kTypeValue, ValueType::kTypeMerge, ValueType::kTypeValue},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File doesn't overwrite any keys, no seqno needed
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno);

@@ -574,7 +546,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {40, 41, 42},
        {ValueType::kTypeValue, ValueType::kTypeDeletion,
         ValueType::kTypeDeletion},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 1);

@@ -582,7 +554,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {20, 30, 40},
        {ValueType::kTypeDeletion, ValueType::kTypeDeletion,
         ValueType::kTypeDeletion},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // File overwrites some keys, a seqno will be assigned
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 2);

@@ -592,13 +564,13 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
    // keys in the DB or not because we have a snapshot
    ASSERT_OK(GenerateAndAddExternalFile(
        options, {1000, 1002}, {ValueType::kTypeValue, ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 3);

    ASSERT_OK(GenerateAndAddExternalFile(
        options, {2000, 3002}, {ValueType::kTypeValue, ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 4);

@@ -606,7 +578,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {
        options, {1, 20, 40, 100, 150},
        {ValueType::kTypeDeletion, ValueType::kTypeDeletion,
         ValueType::kTypeValue, ValueType::kTypeMerge, ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // A global seqno will be assigned anyway because of the snapshot
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 5);

@@ -614,13 +586,13 @@ TEST_P(ExternalSSTFileBasicTest, IngestFileWithMixedValueType) {

    ASSERT_OK(GenerateAndAddExternalFile(
        options, {5000, 5001}, {ValueType::kTypeValue, ValueType::kTypeMerge},
-        file_id++, write_global_seqno, &true_data));
+        file_id++, &true_data));
    // No snapshot anymore, no need to assign a seqno
    ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno + 5);

    size_t kcnt = 0;
    VerifyDBFromMap(true_data, &kcnt, false);
-  } while (ChangeOptionsForFileIngestionTest());
+  } while (ChangeCompactOptions());
 }

 TEST_F(ExternalSSTFileBasicTest, FadviseTrigger) {
@@ -663,7 +635,7 @@ TEST_F(ExternalSSTFileBasicTest, FadviseTrigger) {
  rocksdb::SyncPoint::GetInstance()->DisableProcessing();
 }

-TEST_P(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
+TEST_F(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
  int kNumLevels = 7;
  Options options = CurrentOptions();
  options.disable_auto_compactions = true;
@@ -690,13 +662,12 @@ TEST_P(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
  ASSERT_EQ(0, NumTableFilesAtLevel(kNumLevels - 2));
  ASSERT_EQ(1, NumTableFilesAtLevel(kNumLevels - 1));

-  bool write_global_seqno = GetParam();
  // overlaps with L0 file but not memtable, so flush is skipped and file is
  // ingested into L0
  SequenceNumber last_seqno = dbfull()->GetLatestSequenceNumber();
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {60, 90}, {ValueType::kTypeValue, ValueType::kTypeValue},
-      {{65, 70}, {70, 85}}, file_id++, write_global_seqno, &true_data));
+      {{65, 70}, {70, 85}}, file_id++, &true_data));
  ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), ++last_seqno);
  ASSERT_EQ(2, NumTableFilesAtLevel(0));
  ASSERT_EQ(0, NumTableFilesAtLevel(kNumLevels - 2));
@@ -706,7 +677,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
  // file is ingested into L5
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {10, 40}, {ValueType::kTypeValue, ValueType::kTypeValue},
-      file_id++, write_global_seqno, &true_data));
+      file_id++, &true_data));
  ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), ++last_seqno);
  ASSERT_EQ(2, NumTableFilesAtLevel(0));
  ASSERT_EQ(1, NumTableFilesAtLevel(kNumLevels - 2));
@@ -715,7 +686,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
  // overlaps with L5 file but not memtable or L0 file, so flush is skipped and
  // file is ingested into L4
  ASSERT_OK(GenerateAndAddExternalFile(options, {}, {}, {{5, 15}}, file_id++,
-                                       write_global_seqno, &true_data));
+                                       &true_data));
  ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), ++last_seqno);
  ASSERT_EQ(2, NumTableFilesAtLevel(0));
  ASSERT_EQ(1, NumTableFilesAtLevel(kNumLevels - 2));
@@ -727,7 +698,7 @@ TEST_P(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
  // count increases by two.
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {100, 140}, {ValueType::kTypeValue, ValueType::kTypeValue},
-      file_id++, write_global_seqno, &true_data));
+      file_id++, &true_data));
  ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), ++last_seqno);
  ASSERT_EQ(4, NumTableFilesAtLevel(0));
  ASSERT_EQ(1, NumTableFilesAtLevel(kNumLevels - 2));
@@ -740,16 +711,13 @@ TEST_P(ExternalSSTFileBasicTest, IngestionWithRangeDeletions) {
  // seqnum.
  ASSERT_OK(GenerateAndAddExternalFile(
      options, {151, 175}, {ValueType::kTypeValue, ValueType::kTypeValue},
-      {{160, 200}}, file_id++, write_global_seqno, &true_data));
+      {{160, 200}}, file_id++, &true_data));
  ASSERT_EQ(dbfull()->GetLatestSequenceNumber(), last_seqno);
  ASSERT_EQ(4, NumTableFilesAtLevel(0));
  ASSERT_EQ(1, NumTableFilesAtLevel(kNumLevels - 2));
  ASSERT_EQ(2, NumTableFilesAtLevel(options.num_levels - 1));
 }

-INSTANTIATE_TEST_CASE_P(ExternalSSTFileBasicTest, ExternalSSTFileBasicTest,
-                        testing::Bool());
-
 #endif  // ROCKSDB_LITE

 }  // namespace rocksdb
@@ -29,8 +29,7 @@
 namespace rocksdb {

 Status ExternalSstFileIngestionJob::Prepare(
-    const std::vector<std::string>& external_files_paths,
-    uint64_t next_file_number, SuperVersion* sv) {
+    const std::vector<std::string>& external_files_paths, SuperVersion* sv) {
  Status status;

  // Read the information of files we are ingesting
@@ -91,7 +90,7 @@ Status ExternalSstFileIngestionJob::Prepare(

  // Copy/Move external files into DB
  for (IngestedFileInfo& f : files_to_ingest_) {
-    f.fd = FileDescriptor(next_file_number++, 0, f.file_size);
+    f.fd = FileDescriptor(versions_->NewFileNumber(), 0, f.file_size);

    const std::string path_outside_db = f.external_file_path;
    const std::string path_inside_db =
@@ -225,7 +224,7 @@ void ExternalSstFileIngestionJob::UpdateStats() {
  for (IngestedFileInfo& f : files_to_ingest_) {
    InternalStats::CompactionStats stats(CompactionReason::kExternalSstIngestion, 1);
    stats.micros = total_time;
-    // If actual copy occurred for this file, then we need to count the file
+    // If actual copy occured for this file, then we need to count the file
    // size as the actual bytes written. If the file was linked, then we ignore
    // the bytes written for file metadata.
    // TODO (yanqin) maybe account for file metadata bytes for exact accuracy?
@@ -344,7 +343,7 @@ Status ExternalSstFileIngestionJob::GetIngestedFileInfo(
      file_to_ingest->global_seqno_offset = 0;
      return Status::Corruption("Was not able to find file global seqno field");
    }
-    file_to_ingest->global_seqno_offset = static_cast<size_t>(offsets_iter->second);
+    file_to_ingest->global_seqno_offset = offsets_iter->second;
  } else if (file_to_ingest->version == 1) {
    // SST file V1 should not have global seqno field
    assert(seqno_iter == uprops.end());
@@ -476,9 +475,9 @@ Status ExternalSstFileIngestionJob::AssignLevelAndSeqnoForIngestedFile(
        const SequenceNumber level_largest_seqno =
            (*max_element(level_files.begin(), level_files.end(),
                          [](FileMetaData* f1, FileMetaData* f2) {
-                            return f1->fd.largest_seqno < f2->fd.largest_seqno;
+                            return f1->largest_seqno < f2->largest_seqno;
                          }))
-                ->fd.largest_seqno;
+                ->largest_seqno;
        // should only assign seqno to current level's largest seqno when
        // the file fits
        if (level_largest_seqno != 0 &&
@@ -523,7 +522,7 @@ Status ExternalSstFileIngestionJob::CheckLevelForIngestedBehindFile(
  // at some upper level
  for (int lvl = 0; lvl < cfd_->NumberLevels() - 1; lvl++) {
    for (auto file : vstorage->LevelFiles(lvl)) {
-      if (file->fd.smallest_seqno == 0) {
+      if (file->smallest_seqno == 0) {
        return Status::InvalidArgument(
          "Can't ingest_behind file as despite allow_ingest_behind=true "
          "there are files with 0 seqno in database at upper levels!");
@@ -548,27 +547,24 @@ Status ExternalSstFileIngestionJob::AssignGlobalSeqnoForIngestedFile(
        "field");
  }

-  if (ingestion_options_.write_global_seqno) {
-    // Determine if we can write global_seqno to a given offset of file.
-    // If the file system does not support random write, then we should not.
-    // Otherwise we should.
-    std::unique_ptr<RandomRWFile> rwfile;
-    Status status = env_->NewRandomRWFile(file_to_ingest->internal_file_path,
-                                          &rwfile, env_options_);
-    if (status.ok()) {
-      std::string seqno_val;
-      PutFixed64(&seqno_val, seqno);
-      status = rwfile->Write(file_to_ingest->global_seqno_offset, seqno_val);
-      if (!status.ok()) {
-        return status;
-      }
-    } else if (!status.IsNotSupported()) {
-      return status;
-    }
+  std::unique_ptr<RandomRWFile> rwfile;
+  Status status = env_->NewRandomRWFile(file_to_ingest->internal_file_path,
+                                        &rwfile, env_options_);
+  if (!status.ok()) {
+    return status;
  }

-  file_to_ingest->assigned_seqno = seqno;
-  return Status::OK();
+  // Write the new seqno in the global sequence number field in the file
+  std::string seqno_val;
+  PutFixed64(&seqno_val, seqno);
+  status = rwfile->Write(file_to_ingest->global_seqno_offset, seqno_val);
+  if (status.ok()) {
+    status = rwfile->Fsync();
+  }
+  if (status.ok()) {
+    file_to_ingest->assigned_seqno = seqno;
+  }
+  return status;
 }

 bool ExternalSstFileIngestionJob::IngestedFileFitInLevel(
@@ -89,7 +89,7 @@ class ExternalSstFileIngestionJob {

  // Prepare the job by copying external files into the DB.
  Status Prepare(const std::vector<std::string>& external_files_paths,
-                 uint64_t next_file_number, SuperVersion* sv);
+                 SuperVersion* sv);

  // Check if we need to flush the memtable before running the ingestion job
  // This will be true if the files we are ingesting are overlapping with any
@@ -15,8 +15,7 @@

 namespace rocksdb {

-class ExternalSSTFileTest : public DBTestBase,
-                            public ::testing::WithParamInterface<bool> {
+class ExternalSSTFileTest : public DBTestBase {
 public:
  ExternalSSTFileTest() : DBTestBase("/external_sst_file_test") {
    sst_files_dir_ = dbname_ + "/sst_files/";
@@ -31,8 +30,7 @@ class ExternalSSTFileTest : public DBTestBase,
  Status GenerateAndAddExternalFile(
      const Options options,
      std::vector<std::pair<std::string, std::string>> data, int file_id = -1,
-      bool allow_global_seqno = false, bool write_global_seqno = false,
-      bool sort_data = false,
+      bool allow_global_seqno = false, bool sort_data = false,
      std::map<std::string, std::string>* true_data = nullptr,
      ColumnFamilyHandle* cfh = nullptr) {
    // Generate a file id if not provided
@@ -75,7 +73,6 @@ class ExternalSSTFileTest : public DBTestBase,
    if (s.ok()) {
      IngestExternalFileOptions ifo;
      ifo.allow_global_seqno = allow_global_seqno;
-      ifo.write_global_seqno = allow_global_seqno ? write_global_seqno : false;
      if (cfh) {
        s = db_->IngestExternalFile(cfh, {file_path}, ifo);
      } else {
@@ -152,10 +149,11 @@ class ExternalSSTFileTest : public DBTestBase,
    return s;
  }

+
+
  Status GenerateAndAddExternalFile(
      const Options options, std::vector<std::pair<int, std::string>> data,
-      int file_id = -1, bool allow_global_seqno = false,
-      bool write_global_seqno = false, bool sort_data = false,
+      int file_id = -1, bool allow_global_seqno = false, bool sort_data = false,
      std::map<std::string, std::string>* true_data = nullptr,
      ColumnFamilyHandle* cfh = nullptr) {
    std::vector<std::pair<std::string, std::string>> file_data;
@@ -163,14 +161,13 @@ class ExternalSSTFileTest : public DBTestBase,
      file_data.emplace_back(Key(entry.first), entry.second);
    }
    return GenerateAndAddExternalFile(options, file_data, file_id,
-                                      allow_global_seqno, write_global_seqno,
-                                      sort_data, true_data, cfh);
+                                      allow_global_seqno, sort_data, true_data,
+                                      cfh);
  }

  Status GenerateAndAddExternalFile(
      const Options options, std::vector<int> keys, int file_id = -1,
-      bool allow_global_seqno = false, bool write_global_seqno = false,
-      bool sort_data = false,
+      bool allow_global_seqno = false, bool sort_data = false,
      std::map<std::string, std::string>* true_data = nullptr,
      ColumnFamilyHandle* cfh = nullptr) {
    std::vector<std::pair<std::string, std::string>> file_data;
@@ -178,20 +175,18 @@ class ExternalSSTFileTest : public DBTestBase,
      file_data.emplace_back(Key(k), Key(k) + ToString(file_id));
    }
    return GenerateAndAddExternalFile(options, file_data, file_id,
-                                      allow_global_seqno, write_global_seqno,
-                                      sort_data, true_data, cfh);
+                                      allow_global_seqno, sort_data, true_data,
+                                      cfh);
  }

  Status DeprecatedAddFile(const std::vector<std::string>& files,
                           bool move_files = false,
-                           bool skip_snapshot_check = false,
-                           bool skip_write_global_seqno = false) {
+                           bool skip_snapshot_check = false) {
    IngestExternalFileOptions opts;
    opts.move_files = move_files;
    opts.snapshot_consistency = !skip_snapshot_check;
    opts.allow_global_seqno = false;
    opts.allow_blocking_flush = false;
-    opts.write_global_seqno = !skip_write_global_seqno;
    return db_->IngestExternalFile(files, opts);
  }

@@ -458,7 +453,6 @@ TEST_F(ExternalSSTFileTest, Basic) {
  } while (ChangeOptions(kSkipPlainTable | kSkipFIFOCompaction |
                         kRangeDelSkipConfigs));
 }
-
 class SstFileWriterCollector : public TablePropertiesCollector {
 public:
  explicit SstFileWriterCollector(const std::string prefix) : prefix_(prefix) {
@@ -1147,7 +1141,7 @@ TEST_F(ExternalSSTFileTest, OverlappingRanges) {
  } while (ChangeOptions(kSkipPlainTable | kSkipFIFOCompaction));
 }

-TEST_P(ExternalSSTFileTest, PickedLevel) {
+TEST_F(ExternalSSTFileTest, PickedLevel) {
  Options options = CurrentOptions();
  options.disable_auto_compactions = false;
  options.level0_file_num_compaction_trigger = 4;
@@ -1158,11 +1152,11 @@ TEST_P(ExternalSSTFileTest, PickedLevel) {

  // File 0 will go to last level (L3)
  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 10}, -1, false, false,
-                                       false, &true_data));
+                                       &true_data));
  EXPECT_EQ(FilesPerLevel(), "0,0,0,1");

  // File 1 will go to level L2 (since it overlap with file 0 in L3)
-  ASSERT_OK(GenerateAndAddExternalFile(options, {2, 9}, -1, false, false, false,
+  ASSERT_OK(GenerateAndAddExternalFile(options, {2, 9}, -1, false, false,
                                       &true_data));
  EXPECT_EQ(FilesPerLevel(), "0,0,1,1");

@@ -1192,13 +1186,13 @@ TEST_P(ExternalSSTFileTest, PickedLevel) {

  // This file overlaps with file 0 (L3), file 1 (L2) and the
  // output of compaction going to L1
-  ASSERT_OK(GenerateAndAddExternalFile(options, {4, 7}, -1, false, false, false,
+  ASSERT_OK(GenerateAndAddExternalFile(options, {4, 7}, -1, false, false,
                                       &true_data));
  EXPECT_EQ(FilesPerLevel(), "5,0,1,1");

  // This file does not overlap with any file or with the running compaction
  ASSERT_OK(GenerateAndAddExternalFile(options, {9000, 9001}, -1, false, false,
-                                       false, &true_data));
+                                       &true_data));
  EXPECT_EQ(FilesPerLevel(), "5,0,1,2");

  // Hold compaction from finishing
@@ -1317,8 +1311,7 @@ TEST_F(ExternalSSTFileTest, IngestNonExistingFile) {
  ASSERT_OK(Flush());

  ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
-  ASSERT_OK(dbfull()->TEST_WaitForCompact(true));
-  
+
  // After full compaction, there should be only 1 file.
  std::vector<std::string> files;
  env_->GetChildren(dbname_, &files);
@@ -1429,12 +1422,12 @@ TEST_F(ExternalSSTFileTest, PickedLevelDynamic) {
  // This file overlaps with the output of the compaction (going to L3)
  // so the file will be added to L0 since L3 is the base level
  ASSERT_OK(GenerateAndAddExternalFile(options, {31, 32, 33, 34}, -1, false,
-                                       false, false, &true_data));
+                                       false, &true_data));
  EXPECT_EQ(FilesPerLevel(), "5");

  // This file does not overlap with the current running compactiong
  ASSERT_OK(GenerateAndAddExternalFile(options, {9000, 9001}, -1, false, false,
-                                       false, &true_data));
+                                       &true_data));
  EXPECT_EQ(FilesPerLevel(), "5,0,0,1");

  // Hold compaction from finishing
@@ -1449,24 +1442,24 @@ TEST_F(ExternalSSTFileTest, PickedLevelDynamic) {
  Reopen(options);

  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 15, 19}, -1, false, false,
-                                       false, &true_data));
+                                       &true_data));
  ASSERT_EQ(FilesPerLevel(), "1,0,0,3");

  ASSERT_OK(GenerateAndAddExternalFile(options, {1000, 1001, 1002}, -1, false,
-                                       false, false, &true_data));
+                                       false, &true_data));
  ASSERT_EQ(FilesPerLevel(), "1,0,0,4");

  ASSERT_OK(GenerateAndAddExternalFile(options, {500, 600, 700}, -1, false,
-                                       false, false, &true_data));
+                                       false, &true_data));
  ASSERT_EQ(FilesPerLevel(), "1,0,0,5");

  // File 5 overlaps with file 2 (L3 / base level)
  ASSERT_OK(GenerateAndAddExternalFile(options, {2, 10}, -1, false, false,
-                                       false, &true_data));
+                                       &true_data));
  ASSERT_EQ(FilesPerLevel(), "2,0,0,5");

  // File 6 overlaps with file 2 (L3 / base level) and file 5 (L0)
-  ASSERT_OK(GenerateAndAddExternalFile(options, {3, 9}, -1, false, false, false,
+  ASSERT_OK(GenerateAndAddExternalFile(options, {3, 9}, -1, false, false,
                                       &true_data));
  ASSERT_EQ(FilesPerLevel(), "3,0,0,5");

@@ -1486,7 +1479,7 @@ TEST_F(ExternalSSTFileTest, PickedLevelDynamic) {

  // File 7 overlaps with file 4 (L3)
  ASSERT_OK(GenerateAndAddExternalFile(options, {650, 651, 652}, -1, false,
-                                       false, false, &true_data));
+                                       false, &true_data));
  ASSERT_EQ(FilesPerLevel(), "5,0,0,5");

  VerifyDBFromMap(true_data, &kcnt, false);
@@ -1620,13 +1613,12 @@ TEST_F(ExternalSSTFileTest, SstFileWriterNonSharedKeys) {
  ASSERT_OK(DeprecatedAddFile({file_path}));
 }

-TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoRandomized) {
+TEST_F(ExternalSSTFileTest, IngestFileWithGlobalSeqnoRandomized) {
  Options options = CurrentOptions();
  options.IncreaseParallelism(20);
  options.level0_slowdown_writes_trigger = 256;
  options.level0_stop_writes_trigger = 256;

-  bool write_global_seqno = GetParam();
  for (int iter = 0; iter < 2; iter++) {
    bool write_to_memtable = (iter == 0);
    DestroyAndReopen(options);
@@ -1651,8 +1643,7 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoRandomized) {
        }
      } else {
        ASSERT_OK(GenerateAndAddExternalFile(options, random_data, -1, true,
-                                             write_global_seqno, true,
-                                             &true_data));
+                                             true, &true_data));
      }
    }
    size_t kcnt = 0;
@@ -1662,7 +1653,7 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoRandomized) {
  }
 }

-TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
+TEST_F(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
  Options options = CurrentOptions();
  options.num_levels = 5;
  options.disable_auto_compactions = true;
@@ -1681,9 +1672,8 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
  for (int i = 0; i <= 20; i++) {
    file_data.emplace_back(Key(i), "L4");
  }
-  bool write_global_seqno = GetParam();
-  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true,
-                                       write_global_seqno, false, &true_data));
+  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true, false,
+                                       &true_data));

  // This file dont overlap with anything in the DB, will go to L4
  ASSERT_EQ("0,0,0,0,1", FilesPerLevel());
@@ -1693,8 +1683,8 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
  for (int i = 80; i <= 130; i++) {
    file_data.emplace_back(Key(i), "L0");
  }
-  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true,
-                                       write_global_seqno, false, &true_data));
+  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true, false,
+                                       &true_data));

  // This file overlap with the memtable, so it will flush it and add
  // it self to L0
@@ -1705,8 +1695,8 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
  for (int i = 30; i <= 50; i++) {
    file_data.emplace_back(Key(i), "L4");
  }
-  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true,
-                                       write_global_seqno, false, &true_data));
+  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true, false,
+                                       &true_data));

  // This file dont overlap with anything in the DB and fit in L4 as well
  ASSERT_EQ("2,0,0,0,2", FilesPerLevel());
@@ -1716,8 +1706,8 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
  for (int i = 10; i <= 40; i++) {
    file_data.emplace_back(Key(i), "L3");
  }
-  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true,
-                                       write_global_seqno, false, &true_data));
+  ASSERT_OK(GenerateAndAddExternalFile(options, file_data, -1, true, false,
+                                       &true_data));

  // This file overlap with files in L4, we will ingest it in L3
  ASSERT_EQ("2,0,0,1,2", FilesPerLevel());
@@ -1726,7 +1716,7 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoAssignedLevel) {
  VerifyDBFromMap(true_data, &kcnt, false);
 }

-TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {
+TEST_F(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {
  Options options = CurrentOptions();
  DestroyAndReopen(options);
  uint64_t entries_in_memtable;
@@ -1740,17 +1730,16 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {
                      &entries_in_memtable);
  ASSERT_GE(entries_in_memtable, 1);

-  bool write_global_seqno = GetParam();
  // No need for flush
-  ASSERT_OK(GenerateAndAddExternalFile(options, {90, 100, 110}, -1, true,
-                                       write_global_seqno, false, &true_data));
+  ASSERT_OK(GenerateAndAddExternalFile(options, {90, 100, 110}, -1, true, false,
+                                       &true_data));
  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
                      &entries_in_memtable);
  ASSERT_GE(entries_in_memtable, 1);

  // This file will flush the memtable
-  ASSERT_OK(GenerateAndAddExternalFile(options, {19, 20, 21}, -1, true,
-                                       write_global_seqno, false, &true_data));
+  ASSERT_OK(GenerateAndAddExternalFile(options, {19, 20, 21}, -1, true, false,
+                                       &true_data));
  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
                      &entries_in_memtable);
  ASSERT_EQ(entries_in_memtable, 0);
@@ -1765,14 +1754,14 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {

  // No need for flush, this file keys fit between the memtable keys
  ASSERT_OK(GenerateAndAddExternalFile(options, {202, 203, 204}, -1, true,
-                                       write_global_seqno, false, &true_data));
+                                       false, &true_data));
  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
                      &entries_in_memtable);
  ASSERT_GE(entries_in_memtable, 1);

  // This file will flush the memtable
  ASSERT_OK(GenerateAndAddExternalFile(options, {206, 207}, -1, true, false,
-                                       write_global_seqno, &true_data));
+                                       &true_data));
  db_->GetIntProperty(DB::Properties::kNumEntriesActiveMemTable,
                      &entries_in_memtable);
  ASSERT_EQ(entries_in_memtable, 0);
@@ -1781,7 +1770,7 @@ TEST_P(ExternalSSTFileTest, IngestFileWithGlobalSeqnoMemtableFlush) {
  VerifyDBFromMap(true_data, &kcnt, false);
 }

-TEST_P(ExternalSSTFileTest, L0SortingIssue) {
+TEST_F(ExternalSSTFileTest, L0SortingIssue) {
  Options options = CurrentOptions();
  options.num_levels = 2;
  DestroyAndReopen(options);
@@ -1790,13 +1779,10 @@ TEST_P(ExternalSSTFileTest, L0SortingIssue) {
  ASSERT_OK(Put(Key(1), "memtable"));
  ASSERT_OK(Put(Key(10), "memtable"));

-  bool write_global_seqno = GetParam();
  // No Flush needed, No global seqno needed, Ingest in L1
-  ASSERT_OK(GenerateAndAddExternalFile(options, {7, 8}, -1, true,
-                                       write_global_seqno, false));
+  ASSERT_OK(GenerateAndAddExternalFile(options, {7, 8}, -1, true, false));
  // No Flush needed, but need a global seqno, Ingest in L0
-  ASSERT_OK(GenerateAndAddExternalFile(options, {7, 8}, -1, true,
-                                       write_global_seqno, false));
+  ASSERT_OK(GenerateAndAddExternalFile(options, {7, 8}, -1, true, false));
  printf("%s\n", FilesPerLevel().c_str());

  // Overwrite what we added using external files
@@ -2025,17 +2011,15 @@ class TestIngestExternalFileListener : public EventListener {
  std::vector<ExternalFileIngestionInfo> ingested_files;
 };

-TEST_P(ExternalSSTFileTest, IngestionListener) {
+TEST_F(ExternalSSTFileTest, IngestionListener) {
  Options options = CurrentOptions();
  TestIngestExternalFileListener* listener =
      new TestIngestExternalFileListener();
  options.listeners.emplace_back(listener);
  CreateAndReopenWithCF({"koko", "toto"}, options);

-  bool write_global_seqno = GetParam();
  // Ingest into default cf
-  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2}, -1, true,
-                                       write_global_seqno, true, nullptr,
+  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2}, -1, true, true, nullptr,
                                       handles_[0]));
  ASSERT_EQ(listener->ingested_files.size(), 1);
  ASSERT_EQ(listener->ingested_files.back().cf_name, "default");
@@ -2046,8 +2030,7 @@ TEST_P(ExternalSSTFileTest, IngestionListener) {
            "default");

  // Ingest into cf1
-  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2}, -1, true,
-                                       write_global_seqno, true, nullptr,
+  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2}, -1, true, true, nullptr,
                                       handles_[1]));
  ASSERT_EQ(listener->ingested_files.size(), 2);
  ASSERT_EQ(listener->ingested_files.back().cf_name, "koko");
@@ -2058,8 +2041,7 @@ TEST_P(ExternalSSTFileTest, IngestionListener) {
            "koko");

  // Ingest into cf2
-  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2}, -1, true,
-                                       write_global_seqno, true, nullptr,
+  ASSERT_OK(GenerateAndAddExternalFile(options, {1, 2}, -1, true, true, nullptr,
                                       handles_[2]));
  ASSERT_EQ(listener->ingested_files.size(), 3);
  ASSERT_EQ(listener->ingested_files.back().cf_name, "toto");
@@ -2102,7 +2084,7 @@ TEST_F(ExternalSSTFileTest, SnapshotInconsistencyBug) {
  db_->ReleaseSnapshot(snap);
 }

-TEST_P(ExternalSSTFileTest, IngestBehind) {
+TEST_F(ExternalSSTFileTest, IngestBehind) {
  Options options = CurrentOptions();
  options.compaction_style = kCompactionStyleUniversal;
  options.num_levels = 3;
@@ -2126,7 +2108,6 @@ TEST_P(ExternalSSTFileTest, IngestBehind) {
  IngestExternalFileOptions ifo;
  ifo.allow_global_seqno = true;
  ifo.ingest_behind = true;
-  ifo.write_global_seqno = GetParam();

  // Can't ingest behind since allow_ingest_behind isn't set to true
  ASSERT_NOK(GenerateAndAddExternalFileIngestBehind(options, ifo,
@@ -2214,9 +2195,6 @@ TEST_F(ExternalSSTFileTest, SkipBloomFilter) {
  }
 }

-INSTANTIATE_TEST_CASE_P(ExternalSSTFileTest, ExternalSSTFileTest,
-                        testing::Bool());
-
 }  // namespace rocksdb

 int main(int argc, char** argv) {
@@ -70,7 +70,7 @@ class FaultInjectionTest
  std::unique_ptr<Env> base_env_;
  FaultInjectionTestEnv* env_;
  std::string dbname_;
-  std::shared_ptr<Cache> tiny_cache_;
+  shared_ptr<Cache> tiny_cache_;
  Options options_;
  DB* db_;
Author	SHA1	Message	Date
Vijay Nadimpalli	b8d5f909f7	Making platform 007 (gcc 7) default in build_detect_platform.sh (#5947 ) Summary: Making platform 007 (gcc 7) default in build_detect_platform.sh. Pull Request resolved: https://github.com/facebook/rocksdb/pull/5947 Differential Revision: D18038837 Pulled By: vjnadimpalli fbshipit-source-id: 9ac2ddaa93bf328a416faec028970e039886378e	2019-10-30 09:46:28 -07:00
Andrew Kryczka	777c0011bf	Add latest toolchain (gcc-8, etc.) build support for fbcode users (#4923 ) Summary: - When building with internal dependencies, specify this toolchain by setting `ROCKSDB_FBCODE_BUILD_WITH_PLATFORM007=1` - It is not enabled by default. However, it is enabled for TSAN builds in CI since there is a known problem with TSAN in gcc-5: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71090 - I did not add support for Lua since (1) we agreed to deprecate it, and (2) we only have an internal build for v5.3 with this toolchain while that has breaking changes compared to our current version (v5.2). Pull Request resolved: https://github.com/facebook/rocksdb/pull/4923 Differential Revision: D13827226 Pulled By: ajkr fbshipit-source-id: 9aa3388ed3679777cfb15ef8cbcb83c07f62f947	2019-10-30 09:46:07 -07:00
Yanqin Jin	7e1f37eb4f	Bump version and update history	2018-09-13 10:25:20 -07:00
Yanqin Jin	f50476a021	Fix Makefile target 'jtest' on PowerPC (#4357 ) Summary: Before the fix: On a PowerPC machine, run the following ``` $ make jtest ``` The command will fail due to "undefined symbol: crc32c_ppc". It was caused by 'rocksdbjava' Makefile target not including crc32c_ppc object files when generating the shared lib. The fix is simple. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4357 Differential Revision: D9779474 Pulled By: riversand963 fbshipit-source-id: 3c5ec9068c2b9c796e6500f71cd900267064fd51	2018-09-11 16:58:49 -07:00
Yanqin Jin	c97db5282c	Remove warnings caused by unused variables in jni (#4345 ) Summary: Test plan ``` $make clean jclean $make -j32 rocksdbjavastatic $make -j32 rocksdbjava ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/4345 Differential Revision: D9661256 Pulled By: riversand963 fbshipit-source-id: aed316c53b29d02fbdd3fa1063a3e832b8a66469	2018-09-05 13:46:30 -07:00
Sagar Vemuri	0d4c848325	Download bzip2 packages from Internet Archive (#4306 ) Summary: Since bzip.org is no longer maintained, download the bzip2 packages from a snapshot taken by the internet archive until we figure out a more credible source. Fixes issue: #4305 Pull Request resolved: https://github.com/facebook/rocksdb/pull/4306 Differential Revision: D9514868 Pulled By: sagar0 fbshipit-source-id: 57c6a141a62e652f94377efc7ca9916b458e68d5	2018-09-05 09:51:07 -07:00
Yanqin Jin	02028f9c19	Bump version and update history	2018-09-04 09:55:52 -07:00
jsteemann	0db501141a	fix compilation with g++ option `-Wsuggest-override` (#4272 ) Summary: Fixes compilation warnings (which are turned into compilation errors by default) when compiling with g++ option `-Wsuggest-override`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4272 Differential Revision: D9322556 Pulled By: siying fbshipit-source-id: abd57a29ec8f544bee77c0bb438f31be830b7244	2018-09-04 09:31:14 -07:00
Andrew Kryczka	6c47f8f3d9	bump version and update history for 5.15.8	2018-08-31 12:34:51 -07:00
Andrew Kryczka	c00296e9fa	Reduce empty SST creation/deletion in compaction (#4336 ) Summary: This is a followup to #4311. Checking `!RangeDelAggregator::IsEmpty()` before opening a dedicated range tombstone SST did not properly prevent empty SSTs from being generated. That's because it relies on `CollapsedRangeDelMap::Size`, which had an underflow bug when the map was empty. This PR fixes that underflow bug. Also fixed an uninitialized variable in db_stress. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4336 Differential Revision: D9600080 Pulled By: ajkr fbshipit-source-id: bc6980ca79d2cd01b825ebc9dbccd51c1a70cfc7	2018-08-31 12:31:17 -07:00
Yi Wu	39340abf23	BlobDB: Fix expired file not being evicted (#4294 ) Summary: Fix expired file not being evicted from the DB. We have a background task (previously called `CheckSeqFiles` and I rename it to `EvictExpiredFiles`) to scan and remove expired files, but it only close the files, not marking them as expired. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4294 Differential Revision: D9415984 Pulled By: yiwu-arbug fbshipit-source-id: eff7bf0331c52a7ccdb02318602bff7f64f3ef3d	2018-08-24 14:38:20 -07:00
Yi Wu	abbd5a4b93	BlobDB: Can return expiration together with Get() (#4227 ) Summary: Add API to allow fetching expiration of a key with `Get()`. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4227 Differential Revision: D9169897 Pulled By: yiwu-arbug fbshipit-source-id: 2a6f216c493dc75731ddcef1daa689b517fab31b	2018-08-24 14:38:20 -07:00
Yi Wu	4bd3bc5c4f	BlobDB: Fix VisibleToActiveSnapshot() (#4236 ) Summary: There are two issues with `VisibleToActiveSnapshot`: 1. If there are no snapshots, `oldest_snapshot` will be 0 and `VisibleToActiveSnapshot` will always return true. Since the method is used to decide whether it is safe to delete obsolete files, obsolete file won't be able to delete in this case. 2. The `auto` keyword of `auto snapshots = db_impl_->snapshots()` translate to a copy of `const SnapshotList` instead of a reference. Since copy constructor of `SnapshotList` is not defined, using the copy may yield unexpected result. Issue 2 actually hide issue 1 from being catch by tests. During test `snapshots.empty()` can return false while it should actually be empty, and `snapshots.oldest()` return an invalid address, making `oldest_snapshot` being some random large number. The issue was originally reported by BlobDB early adopter at Kuaishou. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4236 Differential Revision: D9188706 Pulled By: yiwu-arbug fbshipit-source-id: a0f2624b927cf9bf28c1bb534784fee5d106f5ea	2018-08-24 14:38:20 -07:00
Yi Wu	c1d8c1e7dc	BlobDB: Cleanup TTLExtractor interface (#4229 ) Summary: Cleanup TTLExtractor interface. The original purpose of it is to allow our users keep using existing `Write()` interface but allow it to accept TTL via `TTLExtractor`. However the interface is confusing. Will replace it with something like `WriteWithTTL(batch, ttl)` in the future. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4229 Differential Revision: D9174390 Pulled By: yiwu-arbug fbshipit-source-id: 68201703d784408b851336ab4dd9b84188245b2d	2018-08-24 14:38:20 -07:00
Andrew Kryczka	5e01c5a312	bump version and update history	2018-08-24 13:54:06 -07:00
Andrew Kryczka	316f30ce02	Reduce empty SST creation/deletion during compaction (#4311 ) Summary: I have a PR to start calling `OnTableFileCreated` for empty SSTs: #4307. However, it is a behavior change so should not go into a patch release. This PR adds back a check to make sure range deletions at least exist before starting file creation. This PR should be safe to backport to earlier versions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4311 Differential Revision: D9493734 Pulled By: ajkr fbshipit-source-id: f0d43cda4cfd904f133cfe3a6eb622f52a9ccbe8	2018-08-24 13:47:06 -07:00
Yanqin Jin	17d42cc84a	Bump version to 5.15.6 and update history	2018-08-22 09:32:19 -07:00
Andrey Zagrebin	cb2d7fc7bc	#3865 followup for fix performance degression introduced by switching order of operands (#4284 ) Summary: Followup for #4266. There is one more place in get_context.cc where MergeOperator::ShouldMerge should be called with reversed list of operands. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4284 Differential Revision: D9380008 Pulled By: sagar0 fbshipit-source-id: 70ec26e607e5b88465e1acbdcd6c6171bd76b9f2	2018-08-21 17:58:12 -07:00
Andrey Zagrebin	2c7422aa65	Summary: This PR addresses issue #3865 and implements the following approach to fix it: - adds `MergeContext::GetOperandsDirectionForward` and `MergeContext::GetOperandsDirectionBackward` to query merge operands in a specific order - `MergeContext::GetOperands` becomes a shortcut for `MergeContext::GetOperandsDirectionForward` - pass `MergeContext::GetOperandsDirectionBackward` to `MergeOperator::ShouldMerge` and document the order Pull Request resolved: https://github.com/facebook/rocksdb/pull/4266 Differential Revision: D9360750 Pulled By: sagar0 fbshipit-source-id: 20cb73ff017760b062ecdcf4382560767086e092	2018-08-21 17:57:45 -07:00
Yanqin Jin	a0a9829f7d	Add SST ingestion to ldb (#4205 ) Summary: We add two subcommands `write_extern_sst` and `ingest_extern_sst` to ldb. This PR avoids changing existing code because we hope to cherry-pick to earlier releases to support compatibility check for external SST file ingestion. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4205 Differential Revision: D9112711 Pulled By: riversand963 fbshipit-source-id: 7cae88380d4de86da8440230e87eca66755648e4	2018-08-21 17:19:17 -07:00
Yi Wu	086bae0599	Bump version to 5.15.5 and update history	2018-08-16 16:45:05 -07:00
Mikhail Antonov	4842d1ea5f	VerifyChecksum() API should preserve options Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/4275 Reviewed By: yiwu-arbug Differential Revision: D9369766 Pulled By: mikhail-antonov fbshipit-source-id: d91b64c34cc1976b324a260767fce343fa32afde	2018-08-16 16:43:30 -07:00
Yanqin Jin	e28e2fff13	Bump version and update history	2018-08-11 21:52:41 -07:00
Anand Ananthabhotla	0b6123068a	Revert changes in PR #4003 (#4263 ) Summary: Revert this change. Not generating the OnTableFileCreated() notification for a 0 byte SST on flush breaks the assumption that every OnTableFileCreationStarted() notification is followed by a corresponding OnTableFileCreated(). Pull Request resolved: https://github.com/facebook/rocksdb/pull/4263 Differential Revision: D9285623 Pulled By: anand1976 fbshipit-source-id: 808c3dcd498b4b4f4ed4be947a29a24b2296aa8d	2018-08-11 21:47:27 -07:00
Yanqin Jin	ee0cf2f473	Bump patch version number.	2018-08-10 18:58:00 -07:00
Maysam Yabandeh	3b66d4b617	Fix wrong partitioned index size recorded in properties block (#4259 ) Summary: After refactoring in https://github.com/facebook/rocksdb/pull/4158 the properties block is written after the index block. This breaks the existing logic in estimating the index size in partitioned indexes. The patch fixes that by using the accurate index block size, which is available since by the time we write the properties block, the index block is already written. The patch also fixes an issue in estimating the partition size with format_version=3 which was resulting into partitions smaller than the configured metadata_block_size. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4259 Differential Revision: D9274454 Pulled By: maysamyabandeh fbshipit-source-id: c82d045505cca3e7ed1a44ee1eaa26e4f25a4272	2018-08-10 17:04:39 -07:00
Yanqin Jin	421be4af42	Bump version and update HISTORY.md	2018-08-09 09:35:24 -07:00
Maysam Yabandeh	9810a21baa	Return correct usable_size for BlockContents (#4246 ) Summary: If jemalloc is disabled or the API is incorrectly referenced (jemalloc api on windows have a prefix je_) memory usage is incorrectly reported for all block sizes. This is because sizeof(char) is always 1. sizeof() is calculated at compile time and (char) is char. The patch uses the size of the slice to fix that. Fixes #4245 Pull Request resolved: https://github.com/facebook/rocksdb/pull/4246 Differential Revision: D9233958 Pulled By: maysamyabandeh fbshipit-source-id: 9646933b24504e2814c7379f06a31148829c6b4e	2018-08-08 20:09:54 -07:00
Andrew Kryczka	f9fc1f483a	bump version and update history	2018-08-01 15:28:24 -07:00
Andrew Kryczka	600ca9a439	Skip range deletions at seqno zero when collapsing (#4216 ) Summary: `CollapsedRangeDelMap` internally uses seqno zero as a sentinel value to denote a gap between range tombstones or the end of range tombstones. It therefore expects to never have consecutive sentinel tombstones. However, since `DeleteRange` is now supported in `SstFileWriter`, an ingested file may contain range tombstones, and that ingested file may be assigned global seqno zero. When such tombstones are added to the collapsed map, they resemble sentinel tombstones due to having seqno zero. Then, the invariant mentioned above about never having consecutive sentinel tombstones can be violated. The symptom of this violation was dereferencing the `end()` iterator (#4204). The fix in this PR is to not add range tombstones with seqno zero to the collapsed map. They're not needed anyways since they can't possibly cover anything (in case of a key and a range tombstone with the same seqno, the key is visible). Pull Request resolved: https://github.com/facebook/rocksdb/pull/4216 Differential Revision: D9121716 Pulled By: ajkr fbshipit-source-id: f5b78a70bea9527354603ea7ac8542a7e2b6a210	2018-08-01 15:25:12 -07:00
Siying Dong	976212ede4	DBImpl::IngestExternalFile() should grab mutex when releasing file number in failure case (#4189 ) Summary: `995fcf7573` has a bug: ReleaseFileNumberFromPendingOutputs() added is not protected by the DB mutex. Fix it by grabbing the lock for this operation. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4189 Differential Revision: D9015447 Pulled By: siying fbshipit-source-id: b8506e09a96c3f95a6fe32b5ca5fcdb9bee88937	2018-07-26 11:32:31 -07:00
Siying Dong	516faa0fa0	Fix bug when seeking backward against an out-of-bound iterator (#4187 ) Summary: `92ee3350e0` introduces an out-of-bound check in BlockBasedTableIterator::Valid(). However, this flag is not reset when re-seeking in backward direction. This caused the iterator to be invalide by mistake. Fix it by always resetting the out-of-bound flag in every seek. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4187 Differential Revision: D8996600 Pulled By: siying fbshipit-source-id: b6235ea614f71381e50e7904c4fb036300604ac1	2018-07-25 17:25:26 -07:00
Andrew Kryczka	9897e42813	Write properties metablock last in block-based tables (#4158 ) Summary: The properties meta-block should come at the end since we always need to read it when opening a file, unlike index/filter/other meta-blocks, which are sometimes read depending on the user's configuration. This ordering will allow us to (in a future PR) do a small readahead on the end of the file to read properties and meta-index blocks with one I/O. The bulk of this PR is a refactoring of the `BlockBasedTableBuilder::Finish` function. It was previously too large with inconsistent error handling, which made it difficult to change. So I broke it up into one function per meta-block write, and tried to make error handling consistent within those functions. Then reordering the metablocks was trivial -- just reorder the calls to these helper functions. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4158 Differential Revision: D8921705 Pulled By: ajkr fbshipit-source-id: 96c9cc3182eb1adf11af46adab79dbeba7b12fcc	2018-07-24 09:45:32 -07:00
Tomas Kolda	f132197a93	Windows JNI build fixes (#4015 ) Summary: Fixing compilation, unsatisfied link exceptions (updated list of files that needs to be linked) and warnings for Windows build. ```C++ //MSVC 2015 does not support dynamic arrays like: rocksdb::Slice key_parts[jkey_parts_len]; //I have converted to: std::vector<rocksdb::Slice> key_parts; ``` Also reusing `free_key_parts` that does the same as `free_key_value_parts` that was removed. Java elapsedTime unit test increase of sleep to 2 ms. Otherwise it was failing. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4015 Differential Revision: D8558215 Pulled By: sagar0 fbshipit-source-id: d3c34f846343f9218424da2402a2bd367bbd0aa2	2018-07-24 09:45:08 -07:00
Yanqin Jin	87f1325b2b	Fix a bug in MANIFEST group commit (#4157 ) Summary: PR #3944 introduces group commit of `VersionEdit` in MANIFEST. The implementation has a bug. When updating the log file number of each column family, we must consider only `VersionEdit`s that operate on the same column family. Otherwise, a column family may accidentally set its log file number higher than actual value, indicating that log files with smaller file number will be ignored, thus causing some updates to be lost. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4157 Differential Revision: D8916650 Pulled By: riversand963 fbshipit-source-id: 8f456cf688f17bf35ad87b38e30e899aa162f201	2018-07-24 09:44:14 -07:00
Andrew Kryczka	b79a965f7c	Smaller tail readahead when not reading index/filters (#4159 ) Summary: In all cases during `BlockBasedTable::Open`, we issue at least three read requests to the file's tail: (1) footer, (2) metaindex block, and (3) properties block. Depending on the config, we may also read other metablocks like filter and index. This PR issues smaller readahead when we expect to do only the three necessary reads mentioned above. Then, 4KB should be enough (ignoring the case where there are lots of user-defined properties). We can keep doing 512KB readahead when additional reads are expected. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4159 Differential Revision: D8924002 Pulled By: ajkr fbshipit-source-id: cfc713275de4d05ce11f18571f1d72e27ccd3356	2018-07-24 09:43:42 -07:00
Dmitri Smirnov	ad7491e11e	Return new operator for Status allocations for Windows (#4128 ) Summary: Windows requires new/delete for memory allocations to be overriden. Refactor to be less intrusive. Differential Revision: D8878047 Pulled By: siying fbshipit-source-id: 35f2b5fec2f88ea48c9be926539c6469060aab36	2018-07-19 17:30:09 -07:00
Siying Dong	c57b91f844	Cap concurrent arena's shard block size to 128KB (#4147 ) Summary: Users sometime see their memtable size far smaller than expected. They probably have hit a fragementation of shard blocks. Cap their size anyway to reduce the impact of problem. 128KB is conservative so I don't imagine it can cause any performance problem. Pull Request resolved: https://github.com/facebook/rocksdb/pull/4147 Differential Revision: D8886706 Pulled By: siying fbshipit-source-id: 8528a2a4196aa4457274522e2565fd3ff28f621e	2018-07-18 10:58:31 -07:00
Yi Wu	decf7e52b0	Fix write get stuck when pipelined write is enabled (#4143 ) Summary: Fix the issue when pipelined write is enabled, writers can get stuck indefinitely and not able to finish the write. It can show with the following example: Assume there are 4 writers W1, W2, W3, W4 (W1 is the first, W4 is the last). T1: all writers pending in WAL writer queue: WAL writer queue: W1, W2, W3, W4 memtable writer queue: empty T2. W1 finish WAL writer and move to memtable writer queue: WAL writer queue: W2, W3, W4, memtable writer queue: W1 T3. W2 and W3 finish WAL write as a batch group. W2 enter ExitAsBatchGroupLeader and move the group to memtable writer queue, but before wake up next leader. WAL writer queue: W4 memtable writer queue: W1, W2, W3 T4. W1, W2, W3 finish memtable write as a batch group. Note that W2 still in the previous ExitAsBatchGroupLeader, although W1 have done memtable write for W2. WAL writer queue: W4 memtable writer queue: empty T5. The thread corresponding to W3 create another writer W3' with the same address as W3. WAL writer queue: W4, W3' memtable writer queue: empty T6. W2 continue with ExitAsBatchGroupLeader. Because the address of W3' is the same as W3, the last writer in its group, it thinks there are no pending writers, so it reset newest_writer_ to null, emptying the queue. W4 and W3' are deleted from the queue and will never be wake up. The issue exists since pipelined write was introduced in 5.5.0. Closes #3704 Pull Request resolved: https://github.com/facebook/rocksdb/pull/4143 Differential Revision: D8871599 Pulled By: yiwu-arbug fbshipit-source-id: 3502674e51066a954a0660257e24ac588f815e2a	2018-07-18 10:18:50 -07:00
Yanqin Jin	be91970d59	Release 5.15.	2018-07-17 17:37:19 -07:00