Compare commits

..
15 Commits
Author SHA1 Message Date
sdong 5b40c0c074 Disable error as warning 2019-11-05 11:00:36 -08:00
sdong 693d04be1e Add one more #include<functional> 2019-11-05 11:00:35 -08:00
sdong 177c306b80 Add some include<functional> 2019-10-31 14:33:56 -07:00
sdong a34facd1b6 [FB Internal] Point to the latest tool chain. 2019-10-31 14:33:55 -07:00
sdong e17067c186 [fb only] revert unintended change of USE_SSE
The previuos change that use gcc-5 set USE_SSE to wrong flag by mistake. Fix it.
2017-07-17 22:21:14 -07:00
sdong 4530535923 [FB Only] use gcc-5 2017-07-17 21:38:00 -07:00
Islam AbdelRahman d788d9a9eb Release 4.3.1 2016-01-26 19:07:19 -08:00
Islam AbdelRahman 8a294d6673 Fix BlockBasedTableTest.NoopTransformSeek failure
Summary:
table_test is failing because we are creating a temp InternalComparator

14:27:28 [ RUN      ] BlockBasedTableTest.NoopTransformSeek
14:27:28 pure virtual method called
14:27:28 terminate called without an active exception
14:27:28 /bin/sh: line 7: 2346261 Aborted                 (core dumped) ./$t

Test Plan: make table_test -j64 && ./table_test --gtest_filter="BlockBasedTableTest.NoopTransformSeek"

Reviewers: igor, sdong, anthony, rven

Reviewed By: rven

Subscribers: dhruba

Differential Revision: https://reviews.facebook.net/D52671
2016-01-26 18:51:31 -08:00
Islam AbdelRahman dde3cdccde Fix issue in Iterator::Seek when using Block based filter block with prefix_extractor
Summary: Similar to D53385 we need to check InDomain before checking the filter block.

Test Plan: unit tests

Reviewers: yhchiang, rven, sdong

Reviewed By: sdong

Subscribers: dhruba

Differential Revision: https://reviews.facebook.net/D53421
2016-01-26 18:45:03 -08:00
Islam AbdelRahman 3c3020c61b Fix bug in block based tables with full filter block and prefix_extractor
Summary:
Right now when we are creating a BlockBasedTable with fill filter block
we add to the filter all the prefixes that are InDomain() based on the prefix_extractor

the problem is that when we read a key from the file, we check the filter block for the prefix whether or not it's InDomain()

Test Plan: unit tests

Reviewers: yhchiang, rven, anthony, kradhakrishnan, sdong

Reviewed By: sdong

Subscribers: dhruba

Differential Revision: https://reviews.facebook.net/D53385
2016-01-26 18:44:56 -08:00
Peter Mattis 2d8fa53ffb Fix index seeking in BlockTableReader::PrefixMayMatch.
PrefixMayMatch previously seeked in the prefix index using an internal
key with a sequence number of 0. This would cause the prefix index seek
to fall off the end if the last key in the index had a user-key greater
than or equal to the key being looked for. Falling off the end of the
index in turn results in PrefixMayMatch returning false if the index is
in memory.
2016-01-26 18:44:47 -08:00
sdong eb965e1bc3 Revert "db_bench: --soft_pending_compaction_bytes_limit should set options.soft_pending_compaction_bytes_limit"
This reverts commit c726b0fdc1.

Bad backporting
2015-12-23 09:40:37 -08:00
sdong c726b0fdc1 db_bench: --soft_pending_compaction_bytes_limit should set options.soft_pending_compaction_bytes_limit
Summary: Fix a bug that options.soft_pending_compaction_bytes_limit is not actually set with --soft_pending_compaction_bytes_limit

Test Plan: Run db_bench with this parameter and make sure the parameter is set correctly.

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D52125
2015-12-17 18:29:23 -08:00
Yueh-Hsuan Chiang e15400d1d9 Fixed the valgrind error in ColumnFamilyTest::CreateAndDropRace
Summary: Fixed the valgrind error in ColumnFamilyTest::CreateAndDropRace

Test Plan: valgrind --error-exitcode=2 --leak-check=full ./column_family_test

Reviewers: kradhakrishnan, rven, anthony, IslamAbdelRahman, sdong

Reviewed By: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D51795
2015-12-11 14:14:08 -08:00
krad 0642d4652d Revert "Support marking snapshots for write-conflict checking - Take 2"
This reverts commit e5c5f23814.
2015-12-08 17:53:20 -08:00
1028 changed files with 34222 additions and 108171 deletions

No files matched your search

+1 -1
View File
@@ -1,6 +1,6 @@
{
"project_id" : "rocksdb",
"conduit_uri" : "https://phabricator.fb.com/api/",
"conduit_uri" : "https://reviews.facebook.net/",
"copyright_holder" : "Facebook",
"load" : [
"arcanist_util"
-8
View File
@@ -32,19 +32,16 @@ build/
ldb
manifest_dump
sst_dump
column_aware_encoding_exp
util/build_version.cc
build_tools/VALGRIND_LOGS/
coverage/COVERAGE_REPORT
.gdbhistory
.gdb_history
package/
.phutil_module_cache
unity.a
tags
rocksdb_dump
rocksdb_undump
db_test2
java/out
java/target
@@ -66,8 +63,3 @@ java/javadoc
scan_build_report/
t
LOG
db_logs/
tp2/
fbcode/
fbcode
+26 -26
View File
@@ -1,41 +1,41 @@
sudo: false
language: cpp
os:
- linux
- osx
compiler:
- clang
jdk:
- oraclejdk7
addons:
apt:
sources: ['ubuntu-toolchain-r-test', 'llvm-toolchain-precise-3.6']
packages: ['clang-3.6' , 'zlib1g-dev', 'libbz2-dev', 'libsnappy-dev', 'curl', 'gflags']
env:
# Run all tests before db_block_cache_test (db_test, db_test2)
- JOB_NAME=unittests ROCKSDBTESTS_END=db_block_cache_test
# Run all tests starting from db_block_cache_test (db_block_cache_test, db_iter_test, ...)
- JOB_NAME=unittests ROCKSDBTESTS_START=db_block_cache_test
# Run java tests
- JOB_NAME=java_test
# Build ROCKSDB_LITE
- JOB_NAME=lite_build
matrix:
include:
- os: linux
compiler: clang
env: COMPILER=clang++-3.6
addons:
apt:
sources: ['ubuntu-toolchain-r-test', 'llvm-toolchain-precise-3.6']
packages: ['clang-3.6', 'clang-format-3.6', 'zlib1g-dev', 'libbz2-dev', 'libsnappy-dev', 'curl']
- os: osx
compiler: clang
install:
# Build gflags
# TODO(noetzli): Remove when gflags available through Travis
- pushd /tmp/ && curl -L https://github.com/gflags/gflags/archive/v2.1.2.tar.gz -o gflags.tar.gz && tar xfz gflags.tar.gz && cd gflags-2.1.2 && cmake . && make && popd
# Download clang-format-diff.py to check source code formatting
- pushd /tmp/ && curl -L http://llvm.org/svn/llvm-project/cfe/trunk/tools/clang-format/clang-format-diff.py -o clang-format-diff.py && chmod +x clang-format-diff.py && popd
before_script:
- if [[ "${TRAVIS_OS_NAME}" == 'linux' ]]; then CXX=clang++-3.6; fi
# Limit the maximum number of open file descriptors to 8192
- ulimit -n 8192 || true
# Add gflags to include/library paths
# TODO(noetzli): Remove when gflags available through Travis
- export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/tmp/gflags-2.1.2/lib"
- export LIBRARY_PATH="$LIBRARY_PATH:/tmp/gflags-2.1.2/lib"
- export CPLUS_INCLUDE_PATH="$CPLUS_INCLUDE_PATH:/tmp/gflags-2.1.2/include"
- if [ -n "${COMPILER}" ]; then CXX=${COMPILER}; fi
- if [[ "${TRAVIS_OS_NAME}" == 'osx' ]]; then brew install gflags snappy; fi
- ulimit -n 2000 || true
# Lousy hack to disable use and testing of fallocate, which doesn't behave quite
# as EnvPosixTest::AllocateTest expects within the Travis OpenVZ environment.
script:
- if [[ "${JOB_NAME}" == 'unittests' ]]; then OPT=-DTRAVIS V=1 make -j4 check_some; fi
- if [[ "${JOB_NAME}" == 'java_test' ]]; then OPT=-DTRAVIS V=1 make clean jclean rocksdbjava jtest; fi
- if [[ "${JOB_NAME}" == 'lite_build' ]]; then OPT="-DTRAVIS -DROCKSDB_LITE" V=1 make -j4 static_lib; fi
- if [[ "${TRAVIS_OS_NAME}" == 'linux' ]]; then OPT=-DTRAVIS CLANG_FORMAT_DIFF=/tmp/clang-format-diff.py make format || true; fi
- OPT=-DTRAVIS V=1 make -j4 check && OPT=-DTRAVIS V=1 make clean jclean rocksdbjava jtest && make clean && OPT="-DTRAVIS -DROCKSDB_LITE" V=1 make -j4 check
notifications:
email:
- leveldb@fb.com
+99 -338
View File
@@ -1,7 +1,7 @@
# This cmake build is for Windows 64-bit only.
#
# Prerequisites:
# You must have at least Visual Studio 2015 Update 3. Start the Developer Command Prompt window that is a part of Visual Studio installation.
# You must have Visual Studio 2013 Update 4 installed. Start the Developer Command Prompt window that is a part of Visual Studio installation.
# Run the build commands from within the Developer Command Prompt window to have paths to the compiler and runtime libraries set.
# You must have git.exe in your %PATH% environment variable.
#
@@ -13,7 +13,8 @@
# cd build
# 3. Run cmake to generate project files for Windows, add more options to enable required third-party libraries.
# See thirdparty.inc for more information.
# sample command: cmake -G "Visual Studio 14 Win64" -DGFLAGS=1 -DSNAPPY=1 -DJEMALLOC=1 -DJNI=1 ..
# sample command: cmake -G "Visual Studio 12 Win64" -DGFLAGS=1 -DSNAPPY=1 -DJEMALLOC=1 ..
# OR for VS Studio 15 cmake -G "Visual Studio 14 Win64" -DGFLAGS=1 -DSNAPPY=1 -DJEMALLOC=1 ..
# 4. Then build the project in debug mode (you may want to add /m[:<N>] flag to run msbuild in <N> parallel threads
# or simply /m ot use all avail cores)
# msbuild rocksdb.sln
@@ -28,117 +29,38 @@
cmake_minimum_required(VERSION 2.6)
project(rocksdb)
if(POLICY CMP0042)
cmake_policy(SET CMP0042 NEW)
endif()
include(${CMAKE_CURRENT_SOURCE_DIR}/thirdparty.inc)
list(APPEND CMAKE_MODULE_PATH "${CMAKE_SOURCE_DIR}/cmake/modules/")
if(WIN32)
include(${CMAKE_CURRENT_SOURCE_DIR}/thirdparty.inc)
else()
option(WITH_JEMALLOC "build with JeMalloc" OFF)
if(WITH_JEMALLOC)
find_package(JeMalloc REQUIRED)
add_definitions(-DROCKSDB_JEMALLOC)
include_directories(${JEMALLOC_INCLUDE_DIR})
endif()
if(CMAKE_SYSTEM_NAME STREQUAL "FreeBSD")
# FreeBSD has jemaloc as default malloc
add_definitions(-DROCKSDB_JEMALLOC)
set(WITH_JEMALLOC ON)
endif()
option(WITH_SNAPPY "build with SNAPPY" OFF)
if(WITH_SNAPPY)
find_package(snappy REQUIRED)
add_definitions(-DSNAPPY)
include_directories(${SNAPPY_INCLUDE_DIR})
list(APPEND THIRDPARTY_LIBS ${SNAPPY_LIBRARIES})
endif()
endif()
if(WIN32)
execute_process(COMMAND powershell -Command "Get-Date -format MM_dd_yyyy" OUTPUT_VARIABLE DATE)
execute_process(COMMAND powershell -Command "Get-Date -format HH:mm:ss" OUTPUT_VARIABLE TIME)
string(REGEX REPLACE "(..)_(..)_..(..).*" "\\1/\\2/\\3" DATE "${DATE}")
string(REGEX REPLACE "(..):(.....).*" " \\1:\\2" TIME "${TIME}")
set(GIT_DATE_TIME "${DATE} ${TIME}")
else()
execute_process(COMMAND date "+%Y/%m/%d %H:%M:%S" OUTPUT_VARIABLE DATETIME)
string(REGEX REPLACE "\n" "" DATETIME ${DATETIME})
set(GIT_DATE_TIME "${DATETIME}")
endif()
execute_process(COMMAND $ENV{COMSPEC} " /C date /T" OUTPUT_VARIABLE DATE)
execute_process(COMMAND $ENV{COMSPEC} " /C time /T" OUTPUT_VARIABLE TIME)
string(REGEX REPLACE "(..)/(..)/..(..).*" "\\1/\\2/\\3" DATE ${DATE})
string(REGEX REPLACE "(..):(.....).*" " \\1:\\2" TIME ${TIME})
string(CONCAT GIT_DATE_TIME ${DATE} ${TIME})
find_package(Git)
if (GIT_FOUND AND EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/.git")
if(WIN32)
execute_process(COMMAND $ENV{COMSPEC} /C ${GIT_EXECUTABLE} -C ${CMAKE_CURRENT_SOURCE_DIR} rev-parse HEAD OUTPUT_VARIABLE GIT_SHA)
else()
execute_process(COMMAND ${GIT_EXECUTABLE} -C ${CMAKE_CURRENT_SOURCE_DIR} rev-parse HEAD OUTPUT_VARIABLE GIT_SHA)
endif()
else()
set(GIT_SHA 0)
set(GIT_SHA 0)
endif()
string(REGEX REPLACE "[^0-9a-f]+" "" GIT_SHA "${GIT_SHA}")
string(REGEX REPLACE "[^0-9a-f]+" "" GIT_SHA ${GIT_SHA})
set(BUILD_VERSION_CC ${CMAKE_BINARY_DIR}/build_version.cc)
configure_file(util/build_version.cc.in ${BUILD_VERSION_CC} @ONLY)
add_library(build_version OBJECT ${BUILD_VERSION_CC})
target_include_directories(build_version PRIVATE
${CMAKE_CURRENT_SOURCE_DIR}/util)
if(WIN32)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /Zi /nologo /EHsc /GS /Gd /GR /GF /fp:precise /Zc:wchar_t /Zc:forScope /errorReport:queue")
set(BUILD_VERSION_CC ${CMAKE_CURRENT_SOURCE_DIR}/util/build_version.cc)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /FC /d2Zi+ /W3 /wd4127 /wd4800 /wd4996 /wd4351")
else()
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -W -Wextra -Wall")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wsign-compare -Wshadow -Wno-unused-parameter -Wno-unused-variable -Woverloaded-virtual -Wnon-virtual-dtor -Wno-missing-field-initializers")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-omit-frame-pointer -momit-leaf-frame-pointer")
endif()
add_custom_command(OUTPUT ${BUILD_VERSION_CC}
COMMAND echo "#include \"build_version.h\"" > ${BUILD_VERSION_CC}
COMMAND echo "const char* rocksdb_build_git_sha = \"rocksdb_build_git_sha:${GIT_SHA}\";" >> ${BUILD_VERSION_CC}
COMMAND echo "const char* rocksdb_build_git_datetime = \"rocksdb_build_git_datetime:${GIT_DATE_TIME}\";" >> ${BUILD_VERSION_CC}
COMMAND echo const char* rocksdb_build_compile_date = __DATE__\; >> ${BUILD_VERSION_CC}
)
option(FAIL_ON_WARNINGS "Treat compile warnings as errors" ON)
if(FAIL_ON_WARNINGS)
if(MSVC)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /WX")
else() # assume GCC
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Werror")
endif()
endif()
add_custom_target(GenerateBuildVersion DEPENDS ${BUILD_VERSION_CC})
option(WITH_ASAN "build with ASAN" OFF)
if(WITH_ASAN)
add_definitions(-DROCKSDB_TSAN_RUN)
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -fsanitize=address")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsanitize=address")
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fsanitize=address")
if(WITH_JEMALLOC)
message(FATAL "ASAN does not work well with JeMalloc")
endif()
endif()
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /Zi /nologo /EHsc /GS /Gd /GR /GF /fp:precise /Zc:wchar_t /Zc:forScope /errorReport:queue")
option(WITH_TSAN "build with TSAN" OFF)
if(WITH_TSAN)
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -fsanitize=thread -pie")
add_definitions(-DROCKSDB_TSAN_RUN)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsanitize=thread -fPIC")
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fsanitize=thread -fPIC")
if(WITH_JEMALLOC)
message(FATAL "TSAN does not work well with JeMalloc")
endif()
endif()
option(WITH_UBSAN "build with UBSAN" OFF)
if(WITH_UBSAN)
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -fsanitize=undefined")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsanitize=undefined")
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fsanitize=undefined")
if(WITH_JEMALLOC)
message(FATAL "UBSAN does not work well with JeMalloc")
endif()
endif()
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /FC /d2Zi+ /W3 /WX /wd4127 /wd4244 /wd4267 /wd4800 /wd4996")
# Used to run CI build and tests so we can run faster
set(OPTIMIZE_DEBUG_DEFAULT 0) # Debug build is unoptimized by default use -DOPTDBG=1 to optimize
@@ -149,104 +71,34 @@ else()
set(OPTIMIZE_DEBUG ${OPTIMIZE_DEBUG_DEFAULT})
endif()
if(WIN32)
if((${OPTIMIZE_DEBUG} EQUAL 1))
message(STATUS "Debug optimization is enabled")
set(CMAKE_CXX_FLAGS_DEBUG "/Oxt /MDd")
else()
set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm /MDd")
endif()
set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /Oxt /Zp8 /Gm- /Gy /MD")
set(CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} /DEBUG")
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} /DEBUG")
if((${OPTIMIZE_DEBUG} EQUAL 1))
message(STATUS "Debug optimization is enabled")
set(CMAKE_CXX_FLAGS_DEBUG "/Oxt /MDd")
else()
set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} /Od /RTC1 /Gm /MDd")
endif()
if(CMAKE_COMPILER_IS_GNUCXX)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-builtin-memcmp")
endif()
set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} /Oxt /Zp8 /Gm- /Gy /MD")
option(ROCKSDB_LITE "Build RocksDBLite version" OFF)
if(ROCKSDB_LITE)
add_definitions(-DROCKSDB_LITE)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-exceptions")
endif()
set(CMAKE_SHARED_LINKER_FLAGS "${CMAKE_SHARED_LINKER_FLAGS} /DEBUG")
set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} /DEBUG")
if(CMAKE_SYSTEM_NAME MATCHES "Cygwin")
add_definitions(-fno-builtin-memcmp -DCYGWIN)
elseif(CMAKE_SYSTEM_NAME MATCHES "Darwin")
add_definitions(-DOS_MACOSX)
if(CMAKE_SYSTEM_PROCESSOR MATCHES arm)
add_definitions(-DIOS_CROSS_COMPILE -DROCKSDB_LITE)
# no debug info for IOS, that will make our library big
add_definitions(-DNDEBUG)
endif()
elseif(CMAKE_SYSTEM_NAME MATCHES "Linux")
add_definitions(-DOS_LINUX)
elseif(CMAKE_SYSTEM_NAME MATCHES "SunOS")
add_definitions(-DOS_SOLARIS)
elseif(CMAKE_SYSTEM_NAME MATCHES "FreeBSD")
add_definitions(-DOS_FREEBSD)
elseif(CMAKE_SYSTEM_NAME MATCHES "NetBSD")
add_definitions(-DOS_NETBSD)
elseif(CMAKE_SYSTEM_NAME MATCHES "OpenBSD")
add_definitions(-DOS_OPENBSD)
elseif(CMAKE_SYSTEM_NAME MATCHES "DragonFly")
add_definitions(-DOS_DRAGONFLYBSD)
elseif(CMAKE_SYSTEM_NAME MATCHES "Android")
add_definitions(-DOS_ANDROID)
elseif(CMAKE_SYSTEM_NAME MATCHES "Windows")
add_definitions(-DWIN32 -DOS_WIN -D_MBCS -DWIN64 -DNOMINMAX)
endif()
if(NOT WIN32)
add_definitions(-DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX)
endif()
option(WITH_FALLOCATE "build with fallocate" ON)
if(WITH_FALLOCATE)
include(CheckCSourceCompiles)
CHECK_C_SOURCE_COMPILES("
#include <fcntl.h>
#include <linux/falloc.h>
int main() {
int fd = open(\"/dev/null\", 0);
fallocate(fd, FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE, 0, 1024);
}
" HAVE_FALLOCATE)
if(HAVE_FALLOCATE)
add_definitions(-DROCKSDB_FALLOCATE_PRESENT)
endif()
endif()
include(CheckFunctionExists)
CHECK_FUNCTION_EXISTS(malloc_usable_size HAVE_MALLOC_USABLE_SIZE)
if(HAVE_MALLOC_USABLE_SIZE)
add_definitions(-DROCKSDB_MALLOC_USABLE_SIZE)
endif()
add_definitions(-DWIN32 -DOS_WIN -D_MBCS -DWIN64)
include_directories(${PROJECT_SOURCE_DIR})
include_directories(${PROJECT_SOURCE_DIR}/include)
include_directories(SYSTEM ${PROJECT_SOURCE_DIR}/third-party/gtest-1.7.0/fused-src)
find_package(Threads REQUIRED)
if(WIN32)
set(SYSTEM_LIBS ${SYSTEM_LIBS} Shlwapi.lib Rpcrt4.lib)
else()
set(SYSTEM_LIBS ${CMAKE_THREAD_LIBS_INIT})
endif()
include_directories(${PROJECT_SOURCE_DIR}/third-party/gtest-1.7.0/fused-src)
set(ROCKSDB_LIBS rocksdblib${ARTIFACT_SUFFIX})
set(THIRDPARTY_LIBS ${THIRDPARTY_LIBS} gtest)
set(SYSTEM_LIBS Shlwapi.lib Rpcrt4.lib)
set(LIBS ${ROCKSDB_LIBS} ${THIRDPARTY_LIBS} ${SYSTEM_LIBS})
add_subdirectory(third-party/gtest-1.7.0/fused-src/gtest)
add_subdirectory(tools)
# Main library source code
set(SOURCES
db/auto_roll_logger.cc
db/builder.cc
db/c.cc
db/column_family.cc
@@ -262,10 +114,8 @@ set(SOURCES
db/db_impl_debug.cc
db/db_impl_experimental.cc
db/db_impl_readonly.cc
db/db_info_dumper.cc
db/db_iter.cc
db/event_helpers.cc
db/external_sst_file_ingestion_job.cc
db/experimental.cc
db/filename.cc
db/file_indexer.cc
@@ -281,8 +131,8 @@ set(SOURCES
db/memtable_list.cc
db/merge_helper.cc
db/merge_operator.cc
db/range_del_aggregator.cc
db/repair.cc
db/slice.cc
db/snapshot_impl.cc
db/table_cache.cc
db/table_properties_collector.cc
@@ -295,13 +145,13 @@ set(SOURCES
db/write_batch_base.cc
db/write_controller.cc
db/write_thread.cc
db/xfunc_test_points.cc
memtable/hash_cuckoo_rep.cc
memtable/hash_linklist_rep.cc
memtable/hash_skiplist_rep.cc
memtable/skiplistrep.cc
memtable/vectorrep.cc
port/stack_trace.cc
port/win/env_win.cc
port/win/port_win.cc
port/win/win_logger.cc
table/adaptive_table_factory.cc
table/block.cc
table/block_based_filter_block.cc
@@ -309,6 +159,7 @@ set(SOURCES
table/block_based_table_factory.cc
table/block_based_table_reader.cc
table/block_builder.cc
table/block_hash_index.cc
table/block_prefix_index.cc
table/bloom_block.cc
table/cuckoo_table_builder.cc
@@ -322,50 +173,48 @@ set(SOURCES
table/merger.cc
table/sst_file_writer.cc
table/meta_blocks.cc
table/mock_table.cc
table/plain_table_builder.cc
table/plain_table_factory.cc
table/plain_table_index.cc
table/plain_table_key_coding.cc
table/plain_table_reader.cc
table/persistent_cache_helper.cc
table/table_properties.cc
table/two_level_iterator.cc
tools/sst_dump_tool.cc
tools/db_bench_tool.cc
tools/dump/db_dump_tool.cc
util/arena.cc
util/auto_roll_logger.cc
util/bloom.cc
util/cf_options.cc
util/clock_cache.cc
util/build_version.cc
util/cache.cc
util/coding.cc
util/compaction_job_stats_impl.cc
util/comparator.cc
util/concurrent_arena.cc
util/crc32c.cc
util/db_options.cc
util/delete_scheduler.cc
util/db_info_dumper.cc
util/delete_scheduler_impl.cc
util/dynamic_bloom.cc
util/env.cc
util/env_chroot.cc
util/env_hdfs.cc
util/event_logger.cc
util/file_util.cc
util/file_reader_writer.cc
util/sst_file_manager_impl.cc
util/filter_policy.cc
util/hash.cc
util/histogram.cc
util/histogram_windowing.cc
util/instrumented_mutex.cc
util/iostats_context.cc
util/lru_cache.cc
tools/ldb_cmd.cc
tools/ldb_tool.cc
util/logging.cc
util/log_buffer.cc
util/memenv.cc
util/mock_env.cc
util/murmurhash.cc
util/mutable_cf_options.cc
util/options.cc
util/options_builder.cc
util/options_helper.cc
util/options_parser.cc
util/options_sanity_check.cc
@@ -373,51 +222,38 @@ set(SOURCES
util/perf_level.cc
util/random.cc
util/rate_limiter.cc
util/sharded_cache.cc
util/skiplistrep.cc
util/slice.cc
util/statistics.cc
util/status.cc
util/status_message.cc
util/string_util.cc
util/sync_point.cc
util/testharness.cc
util/testutil.cc
util/thread_local.cc
util/threadpool_imp.cc
util/thread_status_impl.cc
util/thread_status_updater.cc
util/thread_status_util.cc
util/thread_status_util_debug.cc
util/transaction_test_util.cc
util/vectorrep.cc
util/xfunc.cc
util/xxhash.cc
utilities/backupable/backupable_db.cc
utilities/blob_db/blob_db.cc
utilities/checkpoint/checkpoint.cc
utilities/compaction_filters/remove_emptyvalue_compactionfilter.cc
utilities/date_tiered/date_tiered_db_impl.cc
utilities/document/document_db.cc
utilities/document/json_document.cc
utilities/document/json_document_builder.cc
utilities/env_mirror.cc
utilities/env_registry.cc
utilities/flashcache/flashcache.cc
utilities/geodb/geodb_impl.cc
utilities/leveldb_options/leveldb_options.cc
utilities/lua/rocks_lua_compaction_filter.cc
utilities/memory/memory_util.cc
utilities/merge_operators/string_append/stringappend.cc
utilities/merge_operators/string_append/stringappend2.cc
utilities/merge_operators/put.cc
utilities/merge_operators/max.cc
utilities/merge_operators/uint64add.cc
utilities/option_change_migration/option_change_migration.cc
utilities/options/options_util.cc
utilities/persistent_cache/block_cache_tier.cc
utilities/persistent_cache/block_cache_tier_file.cc
utilities/persistent_cache/block_cache_tier_metadata.cc
utilities/persistent_cache/persistent_cache_tier.cc
utilities/persistent_cache/volatile_tier_impl.cc
utilities/redis/redis_lists.cc
utilities/simulator_cache/sim_cache.cc
utilities/spatialdb/spatial_db.cc
utilities/table_properties_collectors/compact_on_deletion_collector.cc
utilities/transactions/optimistic_transaction_impl.cc
@@ -431,93 +267,59 @@ set(SOURCES
utilities/ttl/db_ttl_impl.cc
utilities/write_batch_with_index/write_batch_with_index.cc
utilities/write_batch_with_index/write_batch_with_index_internal.cc
utilities/col_buf_encoder.cc
utilities/col_buf_decoder.cc
utilities/column_aware_encoding_util.cc
$<TARGET_OBJECTS:build_version>)
)
if(WIN32)
list(APPEND SOURCES
port/win/io_win.cc
port/win/env_win.cc
port/win/env_default.cc
port/win/port_win.cc
port/win/win_logger.cc
port/win/xpress_win.cc)
else()
list(APPEND SOURCES
port/port_posix.cc
util/env_posix.cc
util/io_posix.cc)
endif()
# For test util library that is build only in DEBUG mode
# and linked to tests. Add test only code that is not #ifdefed for Release here.
set(TESTUTIL_SOURCE
db/db_test_util.cc
util/thread_status_updater_debug.cc
)
option(WITH_LIBRADOS "Build with librados" OFF)
if(WITH_LIBRADOS)
list(APPEND SOURCES
utilities/env_librados.cc)
list(APPEND THIRDPARTY_LIBS rados)
endif()
add_library(rocksdblib${ARTIFACT_SUFFIX} STATIC ${SOURCES})
target_link_libraries(rocksdblib${ARTIFACT_SUFFIX}
${THIRDPARTY_LIBS} ${SYSTEM_LIBS})
if(WIN32)
set_target_properties(rocksdblib${ARTIFACT_SUFFIX} PROPERTIES
COMPILE_FLAGS "/Fd${CMAKE_CFG_INTDIR}/rocksdblib${ARTIFACT_SUFFIX}.pdb")
endif()
add_library(rocksdblib${ARTIFACT_SUFFIX} ${SOURCES})
set_target_properties(rocksdblib${ARTIFACT_SUFFIX} PROPERTIES COMPILE_FLAGS "/Fd${CMAKE_CFG_INTDIR}/rocksdblib${ARTIFACT_SUFFIX}.pdb")
add_dependencies(rocksdblib${ARTIFACT_SUFFIX} GenerateBuildVersion)
add_library(rocksdb${ARTIFACT_SUFFIX} SHARED ${SOURCES})
target_link_libraries(rocksdb${ARTIFACT_SUFFIX}
${THIRDPARTY_LIBS} ${SYSTEM_LIBS})
set_target_properties(rocksdb${ARTIFACT_SUFFIX} PROPERTIES COMPILE_FLAGS "-DROCKSDB_DLL -DROCKSDB_LIBRARY_EXPORTS /Fd${CMAKE_CFG_INTDIR}/rocksdb${ARTIFACT_SUFFIX}.pdb")
add_dependencies(rocksdb${ARTIFACT_SUFFIX} GenerateBuildVersion)
target_link_libraries(rocksdb${ARTIFACT_SUFFIX} ${LIBS})
if(WIN32)
set_target_properties(rocksdb${ARTIFACT_SUFFIX} PROPERTIES
COMPILE_FLAGS "-DROCKSDB_DLL -DROCKSDB_LIBRARY_EXPORTS /Fd${CMAKE_CFG_INTDIR}/rocksdb${ARTIFACT_SUFFIX}.pdb")
endif()
set(APPS
db/db_bench.cc
db/memtablerep_bench.cc
table/table_reader_bench.cc
tools/db_stress.cc
tools/write_stress.cc
tools/db_repl_stress.cc
tools/sst_dump.cc
tools/dump/rocksdb_dump.cc
tools/dump/rocksdb_undump.cc
util/cache_bench.cc
)
option(WITH_JNI "build with JNI" OFF)
if(WITH_JNI OR JNI)
message(STATUS "JNI library is enabled")
add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/java)
else()
message(STATUS "JNI library is disabled")
endif()
set(C_TESTS db/c_test.c)
set(TESTS
db/auto_roll_logger_test.cc
db/column_family_test.cc
db/compact_files_test.cc
db/compaction_iterator_test.cc
db/compaction_job_test.cc
db/compaction_job_stats_test.cc
db/compaction_picker_test.cc
db/comparator_db_test.cc
db/corruption_test.cc
db/cuckoo_table_db_test.cc
db/db_iter_test.cc
db/db_test.cc
db/db_compaction_filter_test.cc
db/db_compaction_test.cc
db/db_dynamic_level_test.cc
db/db_inplace_update_test.cc
db/db_iter_test.cc
db/db_log_iter_test.cc
db/db_memtable_test.cc
db/db_merge_operator_test.cc
db/db_options_test.cc
db/db_properties_test.cc
db/db_table_properties_test.cc
db/db_tailing_iter_test.cc
db/db_test.cc
db/db_test2.cc
db/db_block_cache_test.cc
db/db_bloom_filter_test.cc
db/db_flush_test.cc
db/db_iterator_test.cc
db/db_range_del_test.cc
db/db_sst_test.cc
db/external_sst_file_test.cc
db/db_universal_compaction_test.cc
db/db_wal_test.cc
db/db_tailing_iter_test.cc
db/db_table_properties_test.cc
db/dbformat_test.cc
db/deletefile_test.cc
db/fault_injection_test.cc
@@ -535,7 +337,6 @@ set(TESTS
db/perf_context_test.cc
db/plain_table_db_test.cc
db/prefix_test.cc
db/repair_test.cc
db/skiplist_test.cc
db/table_properties_collector_test.cc
db/version_builder_test.cc
@@ -545,56 +346,47 @@ set(TESTS
db/write_batch_test.cc
db/write_callback_test.cc
db/write_controller_test.cc
db/db_io_failure_test.cc
table/block_based_filter_block_test.cc
table/block_hash_index_test.cc
table/block_test.cc
table/cuckoo_table_builder_test.cc
table/cuckoo_table_reader_test.cc
table/full_filter_block_test.cc
table/merger_test.cc
table/table_test.cc
tools/db_sanity_test.cc
tools/ldb_cmd_test.cc
tools/reduce_levels_test.cc
tools/sst_dump_test.cc
util/arena_test.cc
util/autovector_test.cc
util/auto_roll_logger_test.cc
util/bloom_test.cc
util/cache_test.cc
util/coding_test.cc
util/crc32c_test.cc
util/delete_scheduler_test.cc
util/dynamic_bloom_test.cc
util/env_basic_test.cc
util/env_test.cc
util/event_logger_test.cc
util/filelock_test.cc
util/file_reader_writer_test.cc
util/heap_test.cc
util/histogram_test.cc
util/iostats_context_test.cc
util/lru_cache_test.cc
util/memenv_test.cc
util/mock_env_test.cc
util/options_settable_test.cc
util/options_test.cc
util/rate_limiter_test.cc
util/slice_transform_test.cc
util/statistics_test.cc
util/thread_list_test.cc
util/thread_local_test.cc
utilities/backupable/backupable_db_test.cc
utilities/blob_db/blob_db_test.cc
utilities/checkpoint/checkpoint_test.cc
utilities/date_tiered/date_tiered_test.cc
utilities/document/document_db_test.cc
utilities/document/json_document_test.cc
utilities/env_registry_test.cc
utilities/geodb/geodb_test.cc
utilities/memory/memory_test.cc
utilities/merge_operators/string_append/stringappend_test.cc
utilities/option_change_migration/option_change_migration_test.cc
utilities/options/options_util_test.cc
utilities/persistent_cache/hash_table_test.cc
utilities/persistent_cache/persistent_cache_test.cc
utilities/redis/redis_lists_test.cc
utilities/spatialdb/spatial_db_test.cc
utilities/table_properties_collectors/compact_on_deletion_collector_test.cc
@@ -602,45 +394,21 @@ set(TESTS
utilities/transactions/transaction_test.cc
utilities/ttl/ttl_test.cc
utilities/write_batch_with_index/write_batch_with_index_test.cc
utilities/column_aware_encoding_test.cc
utilities/lua/rocks_lua_test.cc
)
if(WITH_LIBRADOS)
list(APPEND TESTS utilities/env_librados_test.cc)
endif()
set(BENCHMARKS
tools/db_bench.cc
table/table_reader_bench.cc
util/cache_bench.cc
db/memtablerep_bench.cc
utilities/column_aware_encoding_exp.cc
utilities/persistent_cache/hash_table_bench.cc)
add_library(testharness OBJECT util/testharness.cc)
foreach(sourcefile ${BENCHMARKS})
get_filename_component(exename ${sourcefile} NAME_WE)
add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile}
$<TARGET_OBJECTS:testharness>)
target_link_libraries(${exename}${ARTIFACT_SUFFIX} ${LIBS} gtest)
endforeach(sourcefile ${BENCHMARKS})
set(EXES ${APPS})
foreach(sourcefile ${EXES})
string(REPLACE ".cc" "" exename ${sourcefile})
string(REGEX REPLACE "^((.+)/)+" "" exename ${exename})
add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile})
target_link_libraries(${exename}${ARTIFACT_SUFFIX} ${LIBS})
endforeach(sourcefile ${EXES})
# For test util library that is build only in DEBUG mode
# and linked to tests. Add test only code that is not #ifdefed for Release here.
set(TESTUTIL_SOURCE
db/db_test_util.cc
table/mock_table.cc
util/mock_env.cc
util/fault_injection_test_env.cc
util/thread_status_updater_debug.cc
)
# test utilities are only build in debug
enable_testing()
add_custom_target(check COMMAND ${CMAKE_CTEST_COMMAND})
set(TESTUTILLIB testutillib${ARTIFACT_SUFFIX})
add_library(${TESTUTILLIB} STATIC ${TESTUTIL_SOURCE})
if(WIN32)
set_target_properties(${TESTUTILLIB} PROPERTIES COMPILE_FLAGS "/Fd${CMAKE_CFG_INTDIR}/testutillib${ARTIFACT_SUFFIX}.pdb")
endif()
set_target_properties(${TESTUTILLIB} PROPERTIES COMPILE_FLAGS "/Fd${CMAKE_CFG_INTDIR}/testutillib${ARTIFACT_SUFFIX}.pdb")
set_target_properties(${TESTUTILLIB}
PROPERTIES EXCLUDE_FROM_DEFAULT_BUILD_RELEASE 1
EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
@@ -651,23 +419,18 @@ set_target_properties(${TESTUTILLIB}
set(TEST_EXES ${TESTS})
foreach(sourcefile ${TEST_EXES})
get_filename_component(exename ${sourcefile} NAME_WE)
add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile}
$<TARGET_OBJECTS:testharness>)
string(REPLACE ".cc" "" exename ${sourcefile})
string(REGEX REPLACE "^((.+)/)+" "" exename ${exename})
add_executable(${exename}${ARTIFACT_SUFFIX} ${sourcefile})
set_target_properties(${exename}${ARTIFACT_SUFFIX}
PROPERTIES EXCLUDE_FROM_DEFAULT_BUILD_RELEASE 1
EXCLUDE_FROM_DEFAULT_BUILD_MINRELEASE 1
EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
)
target_link_libraries(${exename}${ARTIFACT_SUFFIX} testutillib${ARTIFACT_SUFFIX} ${LIBS} gtest)
if(NOT "${exename}" MATCHES "db_sanity_test")
add_test(NAME ${exename} COMMAND ${exename}${ARTIFACT_SUFFIX})
add_dependencies(check ${exename}${ARTIFACT_SUFFIX})
endif()
target_link_libraries(${exename}${ARTIFACT_SUFFIX} ${LIBS} testutillib${ARTIFACT_SUFFIX})
endforeach(sourcefile ${TEST_EXES})
# C executables must link to a shared object
set(C_TESTS db/c_test.c)
set(C_TEST_EXES ${C_TESTS})
foreach(sourcefile ${C_TEST_EXES})
@@ -680,6 +443,4 @@ foreach(sourcefile ${C_TEST_EXES})
EXCLUDE_FROM_DEFAULT_BUILD_RELWITHDEBINFO 1
)
target_link_libraries(${exename}${ARTIFACT_SUFFIX} rocksdb${ARTIFACT_SUFFIX} testutillib${ARTIFACT_SUFFIX})
add_test(NAME ${exename} COMMAND ${exename}${ARTIFACT_SUFFIX})
add_dependencies(check ${exename}${ARTIFACT_SUFFIX})
endforeach(sourcefile ${C_TEST_EXES})
-18
View File
@@ -1,18 +0,0 @@
# RocksDB default options change log
## 5.0 (11/17/2016)
* Options::allow_concurrent_memtable_write and Options::enable_write_thread_adaptive_yield are now true by default
* Options.level0_stop_writes_trigger default value changes from 24 to 32.
## 4.8.0 (5/2/2016)
* options.max_open_files changes from 5000 to -1. It improves performance, but users need to set file descriptor limit to be large enough and watch memory usage for index and bloom filters.
* options.base_background_compactions changes from max_background_compactions to 1. When users set higher max_background_compactions but the write throughput is not high, the writes are less spiky to disks.
* options.wal_recovery_mode changes from kTolerateCorruptedTailRecords to kPointInTimeRecovery. Avoid some false positive when file system or hardware reorder the writes for file data and metadata.
## 4.7.0 (4/8/2016)
* options.write_buffer_size changes from 4MB to 64MB.
* options.target_file_size_base changes from 2MB to 64MB.
* options.max_bytes_for_level_base changes from 10MB to 256MB.
* options.soft_pending_compaction_bytes_limit changes from 0 (disabled) to 64GB.
* options.hard_pending_compaction_bytes_limit changes from 0 (disabled) to 256GB.
* table_cache_numshardbits changes from 4 to 6.
* max_file_opening_threads changes from 1 to 16.
+10 -134
View File
@@ -1,128 +1,4 @@
# Rocksdb Change Log
## Unreleased
### Bug fixes
* Fix the data corruption bug in the case that concurrent memtable write is enabled and 2PC is used.
* OptimizeForPointLookup() doesn't work with the default DB setting of allow_concurrent_memtable_write=true. Fix it.
* Fix a 2PC related bug where WAL files size grow too large.
* Fix the bug that if 2PC is enabled, checkpoints may loss some recent transactions.
* When file copying is needed when creating checkpoints or bulk loading files, fsync the file after the file copying.
## 5.0.0 (11/17/2016)
### Public API Change
* Options::max_bytes_for_level_multiplier is now a double along with all getters and setters.
* Support dynamically change `delayed_write_rate` and `max_total_wal_size` options via SetDBOptions().
* Introduce DB::DeleteRange for optimized deletion of large ranges of contiguous keys.
* Support dynamically change `delayed_write_rate` option via SetDBOptions().
* Options::allow_concurrent_memtable_write and Options::enable_write_thread_adaptive_yield are now true by default.
* Remove Tickers::SEQUENCE_NUMBER to avoid confusion if statistics object is shared among RocksDB instance. Alternatively DB::GetLatestSequenceNumber() can be used to get the same value.
* Options.level0_stop_writes_trigger default value changes from 24 to 32.
* New compaction filter API: CompactionFilter::FilterV2(). Allows to drop ranges of keys.
* Removed flashcache support.
* DB::AddFile() is deprecated and is replaced with DB::IngestExternalFile(). DB::IngestExternalFile() remove all the restrictions that existed for DB::AddFile.
### New Features
* Add avoid_flush_during_shutdown option, which speeds up DB shutdown by not flushing unpersisted data (i.e. with disableWAL = true). Unpersisted data will be lost. The options is dynamically changeable via SetDBOptions().
* Add memtable_insert_with_hint_prefix_extractor option. The option is mean to reduce CPU usage for inserting keys into memtable, if keys can be group by prefix and insert for each prefix are sequential or almost sequential. See include/rocksdb/options.h for more details.
* Add LuaCompactionFilter in utilities. This allows developers to write compaction filters in Lua. To use this feature, LUA_PATH needs to be set to the root directory of Lua.
* No longer populate "LATEST_BACKUP" file in backup directory, which formerly contained the number of the latest backup. The latest backup can be determined by finding the highest numbered file in the "meta/" subdirectory.
## 4.13.0 (10/18/2016)
### Public API Change
* DB::GetOptions() reflect dynamic changed options (i.e. through DB::SetOptions()) and return copy of options instead of reference.
* Added Statistics::getAndResetTickerCount().
### New Features
* Add DB::SetDBOptions() to dynamic change base_background_compactions and max_background_compactions.
* Added Iterator::SeekForPrev(). This new API will seek to the last key that less than or equal to the target key.
## 4.12.0 (9/12/2016)
### Public API Change
* CancelAllBackgroundWork() flushes all memtables for databases containing writes that have bypassed the WAL (writes issued with WriteOptions::disableWAL=true) before shutting down background threads.
* Merge options source_compaction_factor, max_grandparent_overlap_bytes and expanded_compaction_factor into max_compaction_bytes.
* Remove ImmutableCFOptions.
* Add a compression type ZSTD, which can work with ZSTD 0.8.0 or up. Still keep ZSTDNotFinal for compatibility reasons.
### New Features
* Introduce NewClockCache, which is based on CLOCK algorithm with better concurrent performance in some cases. It can be used to replace the default LRU-based block cache and table cache. To use it, RocksDB need to be linked with TBB lib.
* Change ticker/histogram statistics implementations to accumulate data in thread-local storage, which improves CPU performance by reducing cache coherency costs. Callers of CreateDBStatistics do not need to change anything to use this feature.
* Block cache mid-point insertion, where index and filter block are inserted into LRU block cache with higher priority. The feature can be enabled by setting BlockBasedTableOptions::cache_index_and_filter_blocks_with_high_priority to true and high_pri_pool_ratio > 0 when creating NewLRUCache.
## 4.11.0 (8/1/2016)
### Public API Change
* options.memtable_prefix_bloom_huge_page_tlb_size => memtable_huge_page_size. When it is set, RocksDB will try to allocate memory from huge page for memtable too, rather than just memtable bloom filter.
### New Features
* A tool to migrate DB after options change. See include/rocksdb/utilities/option_change_migration.h.
* Add ReadOptions.background_purge_on_iterator_cleanup. If true, we avoid file deletion when destorying iterators.
## 4.10.0 (7/5/2016)
### Public API Change
* options.memtable_prefix_bloom_bits changes to options.memtable_prefix_bloom_bits_ratio and deprecate options.memtable_prefix_bloom_probes
* enum type CompressionType and PerfLevel changes from char to unsigned char. Value of all PerfLevel shift by one.
* Deprecate options.filter_deletes.
### New Features
* Add avoid_flush_during_recovery option.
* Add a read option background_purge_on_iterator_cleanup to avoid deleting files in foreground when destroying iterators. Instead, a job is scheduled in high priority queue and would be executed in a separate background thread.
* RepairDB support for column families. RepairDB now associates data with non-default column families using information embedded in the SST/WAL files (4.7 or later). For data written by 4.6 or earlier, RepairDB associates it with the default column family.
* Add options.write_buffer_manager which allows users to control total memtable sizes across multiple DB instances.
## 4.9.0 (6/9/2016)
### Public API changes
* Add bottommost_compression option, This option can be used to set a specific compression algorithm for the bottommost level (Last level containing files in the DB).
* Introduce CompactionJobInfo::compression, This field state the compression algorithm used to generate the output files of the compaction.
* Deprecate BlockBaseTableOptions.hash_index_allow_collision=false
* Deprecate options builder (GetOptions()).
### New Features
* Introduce NewSimCache() in rocksdb/utilities/sim_cache.h. This function creates a block cache that is able to give simulation results (mainly hit rate) of simulating block behavior with a configurable cache size.
## 4.8.0 (5/2/2016)
### Public API Change
* Allow preset compression dictionary for improved compression of block-based tables. This is supported for zlib, zstd, and lz4. The compression dictionary's size is configurable via CompressionOptions::max_dict_bytes.
* Delete deprecated classes for creating backups (BackupableDB) and restoring from backups (RestoreBackupableDB). Now, BackupEngine should be used for creating backups, and BackupEngineReadOnly should be used for restorations. For more details, see https://github.com/facebook/rocksdb/wiki/How-to-backup-RocksDB%3F
* Expose estimate of per-level compression ratio via DB property: "rocksdb.compression-ratio-at-levelN".
* Added EventListener::OnTableFileCreationStarted. EventListener::OnTableFileCreated will be called on failure case. User can check creation status via TableFileCreationInfo::status.
### New Features
* Add ReadOptions::readahead_size. If non-zero, NewIterator will create a new table reader which performs reads of the given size.
## 4.7.0 (4/8/2016)
### Public API Change
* rename options compaction_measure_io_stats to report_bg_io_stats and include flush too.
* Change some default options. Now default options will optimize for server-workloads. Also enable slowdown and full stop triggers for pending compaction bytes. These changes may cause sub-optimal performance or significant increase of resource usage. To avoid these risks, users can open existing RocksDB with options extracted from RocksDB option files. See https://github.com/facebook/rocksdb/wiki/RocksDB-Options-File for how to use RocksDB option files. Or you can call Options.OldDefaults() to recover old defaults. DEFAULT_OPTIONS_HISTORY.md will track change history of default options.
## 4.6.0 (3/10/2016)
### Public API Changes
* Change default of BlockBasedTableOptions.format_version to 2. It means default DB created by 4.6 or up cannot be opened by RocksDB version 3.9 or earlier.
* Added strict_capacity_limit option to NewLRUCache. If the flag is set to true, insert to cache will fail if no enough capacity can be free. Signature of Cache::Insert() is updated accordingly.
* Tickers [NUMBER_DB_NEXT, NUMBER_DB_PREV, NUMBER_DB_NEXT_FOUND, NUMBER_DB_PREV_FOUND, ITER_BYTES_READ] are not updated immediately. The are updated when the Iterator is deleted.
* Add monotonically increasing counter (DB property "rocksdb.current-super-version-number") that increments upon any change to the LSM tree.
### New Features
* Add CompactionPri::kMinOverlappingRatio, a compaction picking mode friendly to write amplification.
* Deprecate Iterator::IsKeyPinned() and replace it with Iterator::GetProperty() with prop_name="rocksdb.iterator.is.key.pinned"
## 4.5.0 (2/5/2016)
### Public API Changes
* Add a new perf context level between kEnableCount and kEnableTime. Level 2 now does not include timers for mutexes.
* Statistics of mutex operation durations will not be measured by default. If you want to have them enabled, you need to set Statistics::stats_level_ to kAll.
* DBOptions::delete_scheduler and NewDeleteScheduler() are removed, please use DBOptions::sst_file_manager and NewSstFileManager() instead
### New Features
* ldb tool now supports operations to non-default column families.
* Add kPersistedTier to ReadTier. This option allows Get and MultiGet to read only the persited data and skip mem-tables if writes were done with disableWAL = true.
* Add DBOptions::sst_file_manager. Use NewSstFileManager() in include/rocksdb/sst_file_manager.h to create a SstFileManager that can be used to track the total size of SST files and control the SST files deletion rate.
## 4.4.0 (1/14/2016)
### Public API Changes
* Change names in CompactionPri and add a new one.
* Deprecate options.soft_rate_limit and add options.soft_pending_compaction_bytes_limit.
* If options.max_write_buffer_number > 3, writes will be slowed down when writing to the last write buffer to delay a full stop.
* Introduce CompactionJobInfo::compaction_reason, this field include the reason to trigger the compaction.
* After slow down is triggered, if estimated pending compaction bytes keep increasing, slowdown more.
* Increase default options.delayed_write_rate to 2MB/s.
* Added a new parameter --path to ldb tool. --path accepts the name of either MANIFEST, SST or a WAL file. Either --db or --path can be used when calling ldb.
## 4.3.0 (12/8/2015)
### New Features
@@ -136,8 +12,8 @@
## 4.2.0 (11/9/2015)
### New Features
* Introduce CreateLoggerFromOptions(), this function create a Logger for provided DBOptions.
* Add GetAggregatedIntProperty(), which returns the sum of the GetIntProperty of all the column families.
* Introduce CreateLoggerFromOptions(), this function create a Logger for provided DBOptions.
* Add GetAggregatedIntProperty(), which returns the sum of the GetIntProperty of all the column families.
* Add MemoryUtil in rocksdb/utilities/memory.h. It currently offers a way to get the memory usage by type from a list rocksdb instances.
### Public API Changes
@@ -211,8 +87,8 @@
* options.hard_rate_limit is deprecated.
* When options.soft_rate_limit or options.level0_slowdown_writes_trigger is triggered, the way to slow down writes is changed to: write rate to DB is limited to to options.delayed_write_rate.
* DB::GetApproximateSizes() adds a parameter to allow the estimation to include data in mem table, with default to be not to include. It is now only supported in skip list mem table.
* DB::CompactRange() now accept CompactRangeOptions instead of multiple parameters. CompactRangeOptions is defined in include/rocksdb/options.h.
* CompactRange() will now skip bottommost level compaction for level based compaction if there is no compaction filter, bottommost_level_compaction is introduced in CompactRangeOptions to control when it's possible to skip bottommost level compaction. This mean that if you want the compaction to produce a single file you need to set bottommost_level_compaction to BottommostLevelCompaction::kForce.
* DB::CompactRange() now accept CompactRangeOptions instead of multiple paramters. CompactRangeOptions is defined in include/rocksdb/options.h.
* CompactRange() will now skip bottommost level compaction for level based compaction if there is no compaction filter, bottommost_level_compaction is introduced in CompactRangeOptions to control when it's possbile to skip bottommost level compaction. This mean that if you want the compaction to produce a single file you need to set bottommost_level_compaction to BottommostLevelCompaction::kForce.
* Add Cache.GetPinnedUsage() to get the size of memory occupied by entries that are in use by the system.
* DB:Open() will fail if the compression specified in Options is not linked with the binary. If you see this failure, recompile RocksDB with compression libraries present on your system. Also, previously our default compression was snappy. This behavior is now changed. Now, the default compression is snappy only if it's available on the system. If it isn't we change the default to kNoCompression.
* We changed how we account for memory used in block cache. Previously, we only counted the sum of block sizes currently present in block cache. Now, we count the actual memory usage of the blocks. For example, a block of size 4.5KB will use 8KB memory with jemalloc. This might decrease your memory usage and possibly decrease performance. Increase block cache size if you see this happening after an upgrade.
@@ -244,7 +120,7 @@
Lower numbered levels will be placed earlier in the db_paths and higher
numbered levels will be placed later in the db_paths vector.
* Potentially big performance improvements if you're using RocksDB with lots of column families (100-1000)
* Added BlockBasedTableOptions.format_version option, which allows user to specify which version of block based table he wants. As a general guideline, newer versions have more features, but might not be readable by older versions of RocksDB.
* Added BlockBasedTableOptions.format_version option, which allows user to specify which version of block based table he wants. As a general guidline, newer versions have more features, but might not be readable by older versions of RocksDB.
* Added new block based table format (version 2), which you can enable by setting BlockBasedTableOptions.format_version = 2. This format changes how we encode size information in compressed blocks and should help with memory allocations if you're using Zlib or BZip2 compressions.
* MemEnv (env that stores data in memory) is now available in default library build. You can create it by calling NewMemEnv().
* Add SliceTransform.SameResultWhenAppended() to help users determine it is safe to apply prefix bloom/hash.
@@ -320,7 +196,7 @@
## 3.5.0 (9/3/2014)
### New Features
* Add include/utilities/write_batch_with_index.h, providing a utility class to query data out of WriteBatch when building it.
* Add include/utilities/write_batch_with_index.h, providing a utilitiy class to query data out of WriteBatch when building it.
* Move BlockBasedTable related options to BlockBasedTableOptions from Options. Change corresponding JNI interface. Options affected include:
no_block_cache, block_cache, block_cache_compressed, block_size, block_size_deviation, block_restart_interval, filter_policy, whole_key_filtering. filter_policy is changed to shared_ptr from a raw pointer.
* Remove deprecated options: disable_seek_compaction and db_stats_log_interval
@@ -334,7 +210,7 @@
* Support Multiple DB paths in universal style compactions
* Add feature of storing plain table index and bloom filter in SST file.
* CompactRange() will never output compacted files to level 0. This used to be the case when all the compaction input files were at level 0.
* Added iterate_upper_bound to define the extent upto which the forward iterator will return entries. This will prevent iterating over delete markers and overwritten entries for edge cases where you want to break out the iterator anyways. This may improve performance in case there are a large number of delete markers or overwritten entries.
* Added iterate_upper_bound to define the extent upto which the forward iterator will return entries. This will prevent iterating over delete markers and overwritten entries for edge cases where you want to break out the iterator anyways. This may improve perfomance in case there are a large number of delete markers or overwritten entries.
### Public API changes
* DBOptions.db_paths now is a vector of a DBPath structure which indicates both of path and target size
@@ -349,7 +225,7 @@
### New Features
* Added JSON API prototype.
* HashLinklist reduces performance outlier caused by skewed bucket by switching data in the bucket from linked list to skip list. Add parameter threshold_use_skiplist in NewHashLinkListRepFactory().
* RocksDB is now able to reclaim storage space more effectively during the compaction process. This is done by compensating the size of each deletion entry by the 2X average value size, which makes compaction to be triggered by deletion entries more easily.
* RocksDB is now able to reclaim storage space more effectively during the compaction process. This is done by compensating the size of each deletion entry by the 2X average value size, which makes compaction to be triggerred by deletion entries more easily.
* Add TimeOut API to write. Now WriteOptions have a variable called timeout_hint_us. With timeout_hint_us set to non-zero, any write associated with this timeout_hint_us may be aborted when it runs longer than the specified timeout_hint_us, and it is guaranteed that any write completes earlier than the specified time-out will not be aborted due to the time-out condition.
* Add a rate_limiter option, which controls total throughput of flush and compaction. The throughput is specified in bytes/sec. Flush always has precedence over compaction when available bandwidth is constrained.
@@ -364,11 +240,11 @@
2) It added some complexity to the important code-paths,
3) None of our internal customers were really using it.
Because of that, Options::disable_seek_compaction is now obsolete. It is still a parameter in Options, so it does not break the build, but it does not have any effect. We plan to completely remove it at some point, so we ask users to please remove this option from your code base.
* Add two parameters to NewHashLinkListRepFactory() for logging on too many entries in a hash bucket when flushing.
* Add two paramters to NewHashLinkListRepFactory() for logging on too many entries in a hash bucket when flushing.
* Added new option BlockBasedTableOptions::hash_index_allow_collision. When enabled, prefix hash index for block-based table will not store prefix and allow hash collision, reducing memory consumption.
### New Features
* PlainTable now supports a new key encoding: for keys of the same prefix, the prefix is only written once. It can be enabled through encoding_type parameter of NewPlainTableFactory()
* PlainTable now supports a new key encoding: for keys of the same prefix, the prefix is only written once. It can be enabled through encoding_type paramter of NewPlainTableFactory()
* Add AdaptiveTableFactory, which is used to convert from a DB of PlainTable to BlockBasedTabe, or vise versa. It can be created using NewAdaptiveTableFactory()
### Performance Improvements
+5 -5
View File
@@ -21,7 +21,7 @@ depend on gflags. You will need to have gflags installed to run `make all`. This
use binaries compiled by `make all` in production.
* By default the binary we produce is optimized for the platform you're compiling on
(-march=native or the equivalent). If you want to build a portable binary, add 'PORTABLE=1' before
(-march=native). If you want to build a portable binary, add 'PORTABLE=1' before
your make commands, like this: `PORTABLE=1 make static_lib`
## Dependencies
@@ -40,7 +40,7 @@ your make commands, like this: `PORTABLE=1 make static_lib`
## Supported platforms
* **Linux - Ubuntu**
* Upgrade your gcc to version at least 4.8 to get C++11 support.
* Upgrade your gcc to version at least 4.7 to get C++11 support.
* Install gflags. First, try: `sudo apt-get install libgflags-dev`
If this doesn't work and you're using Ubuntu, here's a nice tutorial:
(http://askubuntu.com/questions/312173/installing-gflags-12-04)
@@ -49,8 +49,8 @@ your make commands, like this: `PORTABLE=1 make static_lib`
* Install zlib. Try: `sudo apt-get install zlib1g-dev`.
* Install bzip2: `sudo apt-get install libbz2-dev`.
* **Linux - CentOS**
* Upgrade your gcc to version at least 4.8 to get C++11 support:
`yum install gcc48-c++`
* Upgrade your gcc to version at least 4.7 to get C++11 support:
`yum install gcc47-c++`
* Install gflags:
wget https://gflags.googlecode.com/files/gflags-2.0-no-svn-files.tar.gz
@@ -80,7 +80,7 @@ your make commands, like this: `PORTABLE=1 make static_lib`
* Update XCode: run `xcode-select --install` (or install it from XCode App's settting).
* Install via [homebrew](http://brew.sh/).
* If you're first time developer in MacOS, you still need to run: `xcode-select --install` in your command line.
* run `brew tap homebrew/versions; brew install gcc48 --use-llvm` to install gcc 4.8 (or higher).
* run `brew tap homebrew/versions; brew install gcc47 --use-llvm` to install gcc 4.7 (or higher).
* run `brew install rocksdb`
* **iOS**:
-12
View File
@@ -1,12 +0,0 @@
This is the list of all known third-party language bindings for RocksDB. If something is missing, please open a pull request to add it.
* Java - https://github.com/facebook/rocksdb/tree/master/java
* Python - http://pyrocksdb.readthedocs.org/en/latest/
* Perl - https://metacpan.org/pod/RocksDB
* Node.js - https://npmjs.org/package/rocksdb
* Go - https://github.com/tecbot/gorocksdb
* Ruby - http://rubygems.org/gems/rocksdb-ruby
* Haskell - https://hackage.haskell.org/package/rocksdb-haskell
* PHP - https://github.com/Photonios/rocksdb-php
* C# - https://github.com/warrenfalk/rocksdb-sharp
* Rust - https://github.com/spacejam/rust-rocksdb
+1 -1
View File
@@ -2,7 +2,7 @@ BSD License
For rocksdb software
Copyright (c) 2011-present, Facebook, Inc.
Copyright (c) 2014, Facebook, Inc.
All rights reserved.
---------------------------------------------------------------------
+114 -450
View File
File diff suppressed because it is too large. Load diff
-2
View File
@@ -1,8 +1,6 @@
## RocksDB: A Persistent Key-Value Store for Flash and RAM Storage
[![Build Status](https://travis-ci.org/facebook/rocksdb.svg?branch=master)](https://travis-ci.org/facebook/rocksdb)
[![Build status](https://ci.appveyor.com/api/projects/status/fbgfu0so3afcno78/branch/master?svg=true)](https://ci.appveyor.com/project/Facebook/rocksdb/branch/master)
RocksDB is developed and maintained by Facebook Database Engineering Team.
It is built on earlier work on LevelDB by Sanjay Ghemawat (sanjay@google.com)
+2 -39
View File
@@ -1,21 +1,15 @@
This document lists users of RocksDB and their use cases. If you are using RocksDB, please open a pull request and add yourself to the list.
## Facebook
At Facebook, we use RocksDB as storage engines in multiple data management services and a backend for many different stateful services, including:
At Facebook, we use RocksDB as a backend for many different stateful services. We're also experimenting with running RocksDB as a storage engine for two databases:
1. MyRocks -- https://github.com/MySQLOnRocksDB/mysql-5.6
2. MongoRocks -- https://github.com/mongodb-partners/mongo-rocks
3. ZippyDB -- Facebook's distributed key-value store with Paxos-style replication, built on top of RocksDB.[*] https://www.youtube.com/watch?v=DfiN7pG0D0khtt
4. Laser -- Laser is a high query throughput, low (millisecond) latency, key-value storage service built on top of RocksDB.[*]
4. Dragan -- a distributed graph query engine. https://code.facebook.com/posts/1737605303120405/dragon-a-distributed-graph-query-engine/
5. Stylus -- a low-level stream processing framework writtenin C++.[*]
[*] https://research.facebook.com/publications/realtime-data-processing-at-facebook/
## LinkedIn
Two different use cases at Linkedin are using RocksDB as a storage engine:
1. LinkedIn's follow feed for storing user's activities. Check out the blog post: https://engineering.linkedin.com/blog/2016/03/followfeed--linkedin-s-feed-made-faster-and-smarter
1. LinkedIn's follow feed for storing user's activities
2. Apache Samza, open source framework for stream processing
Learn more about those use cases in a Tech Talk by Ankit Gupta and Naveen Somasundaram: http://www.youtube.com/watch?v=plqVp_OnSzg
@@ -43,34 +37,3 @@ Check out our RocksDB Protobuf merge operator at: https://github.com/vladb38/roc
## Santanader UK/Cloudera Profession Services
Check out their blog post: http://blog.cloudera.com/blog/2015/08/inside-santanders-near-real-time-data-ingest-architecture/
## Airbnb
Airbnb is using RocksDB as a storage engine for their personalized search service. You can learn more about it here: https://www.youtube.com/watch?v=ASQ6XMtogMs
## Pinterest
Pinterest's Object Retrieval System uses RocksDB for storage: https://www.youtube.com/watch?v=MtFEVEs_2Vo
## Smyte
[Smyte](https://www.smyte.com/) uses RocksDB as the storage layer for their core key-value storage, high-performance counters and time-windowed HyperLogLog services.
## Rakuten Marketing
[Rakuten Marketing](https://marketing.rakuten.com/) uses RocksDB as the disk cache layer for the real-time bidding service in their Performance DSP.
## VWO, Wingify
[VWO's](https://vwo.com/) Smart Code checker and URL helper uses RocksDB to store all the URLs where VWO's Smart Code is installed.
## quasardb
[quasardb](https://www.quasardb.net) is a high-performance, distributed, transactional key-value database that integrates well with in-memory analytics engines such as Apache Spark.
quasardb uses a heavily tuned RocksDB as its persistence layer.
## Netflix
[Netflix](http://techblog.netflix.com/2016/05/application-data-caching-using-ssds.html) uses RocksDB on spinning disks to cache application data.
## TiKV
[TiKV](https://github.com/pingcap/tikv) is a GEO-replicated, high-performance, distributed, transactional key-value database. TiKV is powered by Rust and Raft. TiKV uses RocksDB as its persistence layer.
## Apache Flink
[Apache Flink](https://flink.apache.org/news/2016/03/08/release-1.0.0.html) uses RocksDB to store state locally on a machine.
## Dgraph
[Dgraph](https://github.com/dgraph-io/dgraph) is an open-source, scalable, distributed, low latency, high throughput Graph database .They use RocksDB to store state locally on a machine.
Vendored
-1
View File
@@ -1,4 +1,3 @@
# Vagrant file
Vagrant.configure("2") do |config|
config.vm.provider "virtualbox" do |v|
+2 -4
View File
@@ -1,9 +1,8 @@
version: 1.0.{build}
image: Visual Studio 2015
before_build:
- md %APPVEYOR_BUILD_FOLDER%\build
- cd %APPVEYOR_BUILD_FOLDER%\build
- cmake -G "Visual Studio 14 2015 Win64" -DOPTDBG=1 -DXPRESS=1 ..
- cmake -G "Visual Studio 12 Win64" -DOPTDBG=1 ..
- cd ..
build:
project: build\rocksdb.sln
@@ -11,7 +10,6 @@ build:
verbosity: minimal
test:
test_script:
- ps: build_tools\run_ci_db_test.ps1 -EnableRerun -Run db_test2 -Concurrency 8
- ps: build_tools\run_ci_db_test.ps1 -EnableRerun -Run db_test -Exclude DBTest.GroupCommitTest -Concurrency 10
- ps: build_tools\run_ci_db_test.ps1 -EnableRerun -Run db_test -Exclude DBTest.Randomized,DBTest.FileCreationRandomFailure -Concurrency 18
- ps: build_tools\run_ci_db_test.ps1 -Run env_test -Concurrency 1
+30 -63
View File
@@ -6,66 +6,33 @@
* @phutil-library-version 2
*/
if (class_exists('ArcanistWorkflow')) {
phutil_register_library_map(array(
'__library_version__' => 2,
'class' =>
array(
'ArcanistCpplintLinter' => 'cpp_linter/ArcanistCpplintLinter.php',
'BaseDirectoryScopedFormatLinter' => 'cpp_linter/BaseDirectoryScopedFormatLinter.php',
'FacebookArcanistConfiguration' => 'config/FacebookArcanistConfiguration.php',
'FacebookFbcodeLintEngine' => 'lint_engine/FacebookFbcodeLintEngine.php',
'FacebookFbcodeUnitTestEngine' => 'unit_engine/FacebookFbcodeUnitTestEngine.php',
'FacebookHowtoevenLintEngine' => 'lint_engine/FacebookHowtoevenLintEngine.php',
'FacebookHowtoevenLinter' => 'cpp_linter/FacebookHowtoevenLinter.php',
'FbcodeClangFormatLinter' => 'cpp_linter/FbcodeClangFormatLinter.php',
'FbcodeCppLinter' => 'cpp_linter/FbcodeCppLinter.php',
),
'function' =>
array(
),
'xmap' =>
array(
'ArcanistCpplintLinter' => 'ArcanistLinter',
'BaseDirectoryScopedFormatLinter' => 'ArcanistLinter',
'FacebookArcanistConfiguration' => 'ArcanistConfiguration',
'FacebookFbcodeLintEngine' => 'ArcanistLintEngine',
'FacebookFbcodeUnitTestEngine' => 'ArcanistBaseUnitTestEngine',
'FacebookHowtoevenLintEngine' => 'ArcanistLintEngine',
'FacebookHowtoevenLinter' => 'ArcanistLinter',
'FbcodeClangFormatLinter' => 'BaseDirectoryScopedFormatLinter',
'FbcodeCppLinter' => 'ArcanistLinter',
),
));
} else {
phutil_register_library_map(array(
'__library_version__' => 2,
'class' =>
array(
'ArcanistCpplintLinter' => 'cpp_linter/ArcanistCpplintLinter.php',
'BaseDirectoryScopedFormatLinter' => 'cpp_linter/BaseDirectoryScopedFormatLinter.php',
'FacebookArcanistConfiguration' => 'config/FacebookOldArcanistConfiguration.php',
'FacebookFbcodeLintEngine' => 'lint_engine/FacebookFbcodeLintEngine.php',
'FacebookFbcodeUnitTestEngine' => 'unit_engine/FacebookOldFbcodeUnitTestEngine.php',
'FacebookHowtoevenLintEngine' => 'lint_engine/FacebookHowtoevenLintEngine.php',
'FacebookHowtoevenLinter' => 'cpp_linter/FacebookHowtoevenLinter.php',
'FbcodeClangFormatLinter' => 'cpp_linter/FbcodeClangFormatLinter.php',
'FbcodeCppLinter' => 'cpp_linter/FbcodeCppLinter.php',
),
'function' =>
array(
),
'xmap' =>
array(
'ArcanistCpplintLinter' => 'ArcanistLinter',
'BaseDirectoryScopedFormatLinter' => 'ArcanistLinter',
'FacebookArcanistConfiguration' => 'ArcanistConfiguration',
'FacebookFbcodeLintEngine' => 'ArcanistLintEngine',
'FacebookFbcodeUnitTestEngine' => 'ArcanistBaseUnitTestEngine',
'FacebookHowtoevenLintEngine' => 'ArcanistLintEngine',
'FacebookHowtoevenLinter' => 'ArcanistLinter',
'FbcodeClangFormatLinter' => 'BaseDirectoryScopedFormatLinter',
'FbcodeCppLinter' => 'ArcanistLinter',
),
));
}
phutil_register_library_map(array(
'__library_version__' => 2,
'class' =>
array(
'ArcanistCpplintLinter' => 'cpp_linter/ArcanistCpplintLinter.php',
'BaseDirectoryScopedFormatLinter' => 'cpp_linter/BaseDirectoryScopedFormatLinter.php',
'FacebookArcanistConfiguration' => 'config/FacebookArcanistConfiguration.php',
'FacebookFbcodeLintEngine' => 'lint_engine/FacebookFbcodeLintEngine.php',
'FacebookFbcodeUnitTestEngine' => 'unit_engine/FacebookFbcodeUnitTestEngine.php',
'FacebookHowtoevenLintEngine' => 'lint_engine/FacebookHowtoevenLintEngine.php',
'FacebookHowtoevenLinter' => 'cpp_linter/FacebookHowtoevenLinter.php',
'FbcodeClangFormatLinter' => 'cpp_linter/FbcodeClangFormatLinter.php',
'FbcodeCppLinter' => 'cpp_linter/FbcodeCppLinter.php',
),
'function' =>
array(
),
'xmap' =>
array(
'ArcanistCpplintLinter' => 'ArcanistLinter',
'BaseDirectoryScopedFormatLinter' => 'ArcanistLinter',
'FacebookArcanistConfiguration' => 'ArcanistConfiguration',
'FacebookFbcodeLintEngine' => 'ArcanistLintEngine',
'FacebookFbcodeUnitTestEngine' => 'ArcanistBaseUnitTestEngine',
'FacebookHowtoevenLintEngine' => 'ArcanistLintEngine',
'FacebookHowtoevenLinter' => 'ArcanistLinter',
'FbcodeClangFormatLinter' => 'BaseDirectoryScopedFormatLinter',
'FbcodeCppLinter' => 'ArcanistLinter',
),
));
@@ -4,33 +4,32 @@
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
require('RocksDBCommonHelper.php');
define("DIFF_COMMAND", "diff");
class FacebookArcanistConfiguration extends ArcanistConfiguration {
public function didRunWorkflow($command,
ArcanistWorkflow $workflow,
ArcanistBaseWorkflow $workflow,
$error_code) {
// Default options don't terminate on failure, but that's what we want. In
// the current case we use assertions intentionally as "terminate on failure
// invariants".
assert_options(ASSERT_BAIL, true);
assert($workflow);
assert(strlen($command) > 0);
if ($command == DIFF_COMMAND && !$workflow->isRawDiffSource()) {
$diffID = $workflow->getDiffId();
// When submitting a diff this code path gets executed multiple times in
// a row. We only care about the case when ID for the diff is provided
// because that's what we need to apply the diff and trigger the tests.
if (strlen($diffID) > 0) {
assert(is_numeric($diffID));
startTestsInSandcastle(true /* $applyDiff */, $workflow, $diffID);
}
if ($command == 'diff' && !$workflow->isRawDiffSource()) {
$this->maybePushToJenkins($workflow);
}
}
//////////////////////////////////////////////////////////////////////
/* Send off builds to jenkins */
function maybePushToJenkins($workflow) {
$diffID = $workflow->getDiffID();
if ($diffID === null) {
return;
}
$results = $workflow->getTestResults();
if (!$results) {
return;
}
$url = "https://ci-builds.fb.com/view/rocksdb/job/rocksdb_diff_check/"
."buildWithParameters?token=AUTH&DIFF_ID=$diffID";
system("curl --noproxy '*' \"$url\" > /dev/null 2>&1");
}
}
@@ -1,36 +0,0 @@
<?php
// Copyright 2004-present Facebook. All Rights Reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
require('RocksDBCommonHelper.php');
define("DIFF_COMMAND", "diff");
class FacebookArcanistConfiguration extends ArcanistConfiguration {
public function didRunWorkflow($command,
ArcanistBaseWorkflow $workflow,
$error_code) {
// Default options don't terminate on failure, but that's what we want. In
// the current case we use assertions intentionally as "terminate on failure
// invariants".
assert_options(ASSERT_BAIL, true);
assert($workflow);
assert(strlen($command) > 0);
if ($command == DIFF_COMMAND && !$workflow->isRawDiffSource()) {
$diffID = $workflow->getDiffId();
// When submitting a diff this code path gets executed multiple times in
// a row. We only care about the case when ID for the diff is provided
// because that's what we need to apply the diff and trigger the tests.
if (strlen($diffID) > 0) {
assert(is_numeric($diffID));
startTestsInSandcastle(true /* $applyDiff */, $workflow, $diffID);
}
}
}
}
@@ -1,355 +0,0 @@
<?php
// Copyright 2004-present Facebook. All Rights Reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
// Name of the environment variables which need to be set by the entity which
// triggers continuous runs so that code at the end of the file gets executed
// and Sandcastle run starts.
define("ENV_POST_RECEIVE_HOOK", "POST_RECEIVE_HOOK");
define("ENV_HTTPS_APP_VALUE", "HTTPS_APP_VALUE");
define("ENV_HTTPS_TOKEN_VALUE", "HTTPS_TOKEN_VALUE");
define("PRIMARY_TOKEN_FILE", '/home/krad/.sandcastle');
define("CONT_RUN_ALIAS", "leveldb");
//////////////////////////////////////////////////////////////////////
/* Run tests in sandcastle */
function postURL($diffID, $url) {
assert(strlen($diffID) > 0);
assert(is_numeric($diffID));
assert(strlen($url) > 0);
$cmd = 'echo \'{"diff_id": "' . $diffID . '", '
. '"name":"click here for sandcastle tests for D' . $diffID . '", '
. '"link":"' . $url . '"}\' | '
. 'arc call-conduit '
. 'differential.updateunitresults';
shell_exec($cmd);
}
function buildUpdateTestStatusCmd($diffID, $test, $status) {
assert(strlen($diffID) > 0);
assert(is_numeric($diffID));
assert(strlen($test) > 0);
assert(strlen($status) > 0);
$cmd = 'echo \'{"diff_id": "' . $diffID . '", '
. '"name":"' . $test . '", '
. '"result":"' . $status . '"}\' | '
. 'arc call-conduit '
. 'differential.updateunitresults';
return $cmd;
}
function updateTestStatus($diffID, $test) {
assert(strlen($diffID) > 0);
assert(is_numeric($diffID));
assert(strlen($test) > 0);
shell_exec(buildUpdateTestStatusCmd($diffID, $test, "waiting"));
}
function getSteps($applyDiff, $diffID, $username, $test) {
assert(strlen($username) > 0);
assert(strlen($test) > 0);
if ($applyDiff) {
assert(strlen($diffID) > 0);
assert(is_numeric($diffID));
$arcrc_content = (PHP_OS == "Darwin" ?
exec("cat ~/.arcrc | gzip -f | base64") :
exec("cat ~/.arcrc | gzip -f | base64 -w0"));
assert(strlen($arcrc_content) > 0);
// Sandcastle machines don't have arc setup. We copy the user certificate
// and authenticate using that in Sandcastle.
$setup = array(
"name" => "Setup arcrc",
"shell" => "echo " . $arcrc_content . " | base64 --decode"
. " | gzip -d > ~/.arcrc",
"user" => "root"
);
// arc demands certain permission on its config.
// also fix the sticky bit issue in sandcastle
$fix_permission = array(
"name" => "Fix environment",
"shell" => "chmod 600 ~/.arcrc && chmod +t /dev/shm",
"user" => "root"
);
// Construct the steps in the order of execution.
$steps[] = $setup;
$steps[] = $fix_permission;
}
// fbcode is a sub-repo. We cannot patch until we add it to ignore otherwise
// Git thinks it is an uncommited change.
$fix_git_ignore = array(
"name" => "Fix git ignore",
"shell" => "echo fbcode >> .git/info/exclude",
"user" => "root"
);
$steps[] = $fix_git_ignore;
// This will be the command used to execute particular type of tests.
$cmd = "";
if ($applyDiff) {
// Patch the code (keep your fingures crossed).
$patch = array(
"name" => "Patch " . $diffID,
"shell" => "arc --arcrc-file ~/.arcrc "
. "patch --nocommit --diff " . $diffID,
"user" => "root"
);
$steps[] = $patch;
updateTestStatus($diffID, $test);
$cmd = buildUpdateTestStatusCmd($diffID, $test, "running") . "; ";
}
// Run the actual command.
$cmd = $cmd . "J=$(nproc) ./build_tools/precommit_checker.py " . $test
. "; exit_code=$?; ";
if ($applyDiff) {
$cmd = $cmd . "([[ \$exit_code -eq 0 ]] &&"
. buildUpdateTestStatusCmd($diffID, $test, "pass") . ")"
. "||" . buildUpdateTestStatusCmd($diffID, $test, "fail")
. "; ";
}
// shell command to sort the tests based on exit code and print
// the output of the log files.
$cat_sorted_logs = "
while read code log_file;
do echo \"################ cat \$log_file [exit_code : \$code] ################\";
cat \$log_file;
done < <(tail -n +2 LOG | sort -k7,7n -k4,4gr | awk '{print \$7,\$NF}')";
// Shell command to cat all log files
$cat_all_logs = "for f in `ls t/!(run-*)`; do echo \$f;cat \$f; done";
// If LOG file exist use it to cat log files sorted by exit code, otherwise
// cat everything
$logs_cmd = "if [ -f LOG ]; then {$cat_sorted_logs}; else {$cat_all_logs}; fi";
$cmd = $cmd . " cat /tmp/precommit-check.log"
. "; shopt -s extglob; {$logs_cmd}"
. "; shopt -u extglob; [[ \$exit_code -eq 0 ]]";
assert(strlen($cmd) > 0);
$run_test = array(
"name" => "Run " . $test,
"shell" => $cmd,
"user" => "root",
"parser" => "python build_tools/error_filter.py " . $test,
);
$steps[] = $run_test;
if ($applyDiff) {
// Clean up the user arc config we are using.
$cleanup = array(
"name" => "Arc cleanup",
"shell" => "rm -f ~/.arcrc",
"user" => "root"
);
$steps[] = $cleanup;
}
assert(count($steps) > 0);
return $steps;
}
function getSandcastleConfig() {
$sandcastle_config = array();
$cwd = getcwd();
$cwd_token_file = "{$cwd}/.sandcastle";
// This is a case when we're executed from a continuous run. Fetch the values
// from the environment.
if (getenv(ENV_POST_RECEIVE_HOOK)) {
$sandcastle_config[0] = getenv(ENV_HTTPS_APP_VALUE);
$sandcastle_config[1] = getenv(ENV_HTTPS_TOKEN_VALUE);
} else {
// This is a typical `[p]arc diff` case. Fetch the values from the specific
// configuration files.
for ($i = 0; $i < 50; $i++) {
if (file_exists(PRIMARY_TOKEN_FILE) ||
file_exists($cwd_token_file)) {
break;
}
// If we failed to fetch the tokens, sleep for 0.2 second and try again
usleep(200000);
}
assert(file_exists(PRIMARY_TOKEN_FILE) ||
file_exists($cwd_token_file));
// Try the primary location first, followed by a secondary.
if (file_exists(PRIMARY_TOKEN_FILE)) {
$cmd = 'cat ' . PRIMARY_TOKEN_FILE;
} else {
$cmd = 'cat ' . $cwd_token_file;
}
assert(strlen($cmd) > 0);
$sandcastle_config = explode(':', rtrim(shell_exec($cmd)));
}
// In this case be very explicit about the implications.
if (count($sandcastle_config) != 2) {
echo "Sandcastle configuration files don't contain valid information " .
"or the necessary environment variables aren't defined. Unable " .
"to validate the code changes.";
exit(1);
}
assert(strlen($sandcastle_config[0]) > 0);
assert(strlen($sandcastle_config[1]) > 0);
assert(count($sandcastle_config) > 0);
return $sandcastle_config;
}
// This function can be called either from `[p]arc diff` command or during
// the Git post-receive hook.
function startTestsInSandcastle($applyDiff, $workflow, $diffID) {
// Default options don't terminate on failure, but that's what we want. In
// the current case we use assertions intentionally as "terminate on failure
// invariants".
assert_options(ASSERT_BAIL, true);
// In case of a diff we'll send notificatios to the author. Else it'll go to
// the entire team because failures indicate that build quality has regressed.
$username = $applyDiff ? exec("whoami") : CONT_RUN_ALIAS;
assert(strlen($username) > 0);
if ($applyDiff) {
assert($workflow);
assert(strlen($diffID) > 0);
assert(is_numeric($diffID));
}
// List of tests we want to run in Sandcastle.
$tests = array("unit", "unit_non_shm", "unit_481", "clang_unit", "tsan",
"asan", "lite_test", "valgrind", "release", "release_481",
"clang_release", "punit", "clang_analyze", "code_cov",
"java_build", "no_compression", "unity", "ubsan");
$send_email_template = array(
'type' => 'email',
'triggers' => array('fail'),
'emails' => array($username . '@fb.com'),
);
// Construct a job definition for each test and add it to the master plan.
foreach ($tests as $test) {
$stepName = "RocksDB diff " . $diffID . " test " . $test;
if (!$applyDiff) {
$stepName = "RocksDB continuous integration test " . $test;
}
$arg[] = array(
"name" => $stepName,
"report" => array($send_email_template),
"steps" => getSteps($applyDiff, $diffID, $username, $test)
);
}
// We cannot submit the parallel execution master plan to Sandcastle and
// need supply the job plan as a determinator. So we construct a small job
// that will spit out the master job plan which Sandcastle will parse and
// execute. Why compress the job definitions? Otherwise we run over the max
// string size.
$cmd = "echo " . base64_encode(json_encode($arg))
. (PHP_OS == "Darwin" ?
" | gzip -f | base64" :
" | gzip -f | base64 -w0");
assert(strlen($cmd) > 0);
$arg_encoded = shell_exec($cmd);
assert(strlen($arg_encoded) > 0);
$runName = "Run diff " . $diffID . "for user " . $username;
if (!$applyDiff) {
$runName = "RocksDB continuous integration build and test run";
}
$command = array(
"name" => $runName,
"steps" => array()
);
$command["steps"][] = array(
"name" => "Generate determinator",
"shell" => "echo " . $arg_encoded . " | base64 --decode | gzip -d"
. " | base64 --decode",
"determinator" => true,
"user" => "root"
);
// Submit to Sandcastle.
$url = 'https://interngraph.intern.facebook.com/sandcastle/create';
$job = array(
'command' => 'SandcastleUniversalCommand',
'args' => $command,
'capabilities' => array(
'vcs' => 'rocksdb-int-git',
'type' => 'lego',
),
'hash' => 'origin/master',
'user' => $username,
'alias' => 'rocksdb-precommit',
'tags' => array('rocksdb'),
'description' => 'Rocksdb precommit job',
);
// Fetch the configuration necessary to submit a successful HTTPS request.
$sandcastle_config = getSandcastleConfig();
$app = $sandcastle_config[0];
$token = $sandcastle_config[1];
$cmd = 'curl -s -k -F app=' . $app . ' '
. '-F token=' . $token . ' -F job=\'' . json_encode($job)
.'\' "' . $url . '"';
$output = shell_exec($cmd);
assert(strlen($output) > 0);
// Extract Sandcastle URL from the response.
preg_match('/url": "(.+)"/', $output, $sandcastle_url);
assert(count($sandcastle_url) > 0, "Unable to submit Sandcastle request.");
assert(strlen($sandcastle_url[1]) > 0, "Unable to extract Sandcastle URL.");
if ($applyDiff) {
echo "\nSandcastle URL: " . $sandcastle_url[1] . "\n";
// Ask Phabricator to display it on the diff UI.
postURL($diffID, $sandcastle_url[1]);
} else {
echo "Continuous integration started Sandcastle tests. You can look at ";
echo "the progress at:\n" . $sandcastle_url[1] . "\n";
}
}
// Continuous run cript will set the environment variable and based on that
// we'll trigger the execution of tests in Sandcastle. In that case we don't
// need to apply any diffs and there's no associated workflow either.
if (getenv(ENV_POST_RECEIVE_HOOK)) {
startTestsInSandcastle(
false /* $applyDiff */,
NULL /* $workflow */,
NULL /* $diffID */);
}
@@ -44,7 +44,7 @@ abstract class BaseDirectoryScopedFormatLinter extends ArcanistLinter {
$futures[$path] = $this->getFormatFuture($path, $changed);
}
foreach (id(new FutureIterator($futures))->limit(8) as $p => $f) {
foreach (Futures($futures)->limit(8) as $p => $f) {
$this->rawLintOutput[$p] = $f->resolvex();
}
}
+2 -2
View File
@@ -1,5 +1,5 @@
#!/usr/bin/python
# Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
# Copyright (c) 2013, Facebook, Inc. All rights reserved.
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree. An additional grant
# of patent rights can be found in the PATENTS file in the same directory.
@@ -2714,7 +2714,7 @@ def CheckSpacing(filename, clean_lines, linenum, nesting_state, error):
# Look for < that is not surrounded by spaces. This is only
# triggered if both sides are missing spaces, even though
# technically should flag if at least one side is missing a
# technically should should flag if at least one side is missing a
# space. This is done to avoid some false positives with shifts.
match = Search(r'[^\s<]<([^\s=<].*)', reduced_line)
if (match and
@@ -42,12 +42,14 @@ class FacebookFbcodeLintEngine extends ArcanistLintEngine {
$python_linter = new ArcanistPEP8Linter();
$linters[] = $python_linter;
$cpp_linters = array();
$cpp_linters[] = $linters[] = new ArcanistCpplintLinter();
$cpp_linters[] = $linters[] = new FbcodeCppLinter();
if (!$this->getCommitHookMode()) {
$cpp_linters = array();
$cpp_linters[] = $linters[] = new ArcanistCpplintLinter();
$cpp_linters[] = $linters[] = new FbcodeCppLinter();
$clang_format_linter = new FbcodeClangFormatLinter();
$linters[] = $clang_format_linter;
$clang_format_linter = new FbcodeClangFormatLinter();
$linters[] = $clang_format_linter;
}
$spelling_linter = new ArcanistSpellingLinter();
$linters[] = $spelling_linter;
@@ -4,14 +4,18 @@
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
class FacebookFbcodeUnitTestEngine extends ArcanistUnitTestEngine {
class FacebookFbcodeUnitTestEngine extends ArcanistBaseUnitTestEngine {
public function run() {
// For a call to `arc call-conduit differential.updateunitresults` to
// succeed we need at least one entry here.
$result = new ArcanistUnitTestResult();
$result->setName("dummy_placeholder_entry");
$result->setResult(ArcanistUnitTestResult::RESULT_PASS);
return array($result);
// Here we create a new unit test "jenkins_async_test" and promise we'll
// update the results later.
// Jenkins updates the results using `arc call-conduit
// differential.updateunitresults` call. If you change the name here, also
// make sure to change the name in Jenkins script that updates the test
// result -- they have to be the same.
$result = new ArcanistUnitTestResult();
$result->setName("jenkins_async_test");
$result->setResult(ArcanistUnitTestResult::RESULT_POSTPONED);
return array($result);
}
}
@@ -1,17 +0,0 @@
<?php
// Copyright 2004-present Facebook. All Rights Reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
class FacebookFbcodeUnitTestEngine extends ArcanistBaseUnitTestEngine {
public function run() {
// For a call to `arc call-conduit differential.updateunitresults` to
// succeed we need at least one entry here.
$result = new ArcanistUnitTestResult();
$result->setName("dummy_placeholder_entry");
$result->setResult(ArcanistUnitTestResult::RESULT_PASS);
return array($result);
}
}
+22 -97
View File
@@ -19,21 +19,22 @@
#
# The PLATFORM_CCFLAGS and PLATFORM_CXXFLAGS might include the following:
#
# -DROCKSDB_PLATFORM_POSIX if posix-platform based
# -DLEVELDB_PLATFORM_POSIX if cstdatomic is present
# -DLEVELDB_PLATFORM_NOATOMIC if it is not
# -DSNAPPY if the Snappy library is present
# -DLZ4 if the LZ4 library is present
# -DZSTD if the ZSTD library is present
# -DNUMA if the NUMA library is present
# -DTBB if the TBB library is present
#
# Using gflags in rocksdb:
# Our project depends on gflags, which requires users to take some extra steps
# before they can compile the whole repository:
# 1. Install gflags. You may download it from here:
# https://gflags.github.io/gflags/ (Mac users can `brew install gflags`)
# 2. Once installed, add the include path for gflags to your CPATH env var and
# the lib path to LIBRARY_PATH. If installed with default settings, the lib
# will be /usr/local/lib and the include path will be /usr/local/include
# https://code.google.com/p/gflags/
# 2. Once install, add the include path/lib path for gflags to CPATH and
# LIBRARY_PATH respectively. If installed with default mode, the
# lib and include path will be /usr/local/lib and /usr/local/include
# Mac user can do this by having brew installed and running brew install gflags
OUTPUT=$1
if test -z "$OUTPUT"; then
@@ -164,7 +165,7 @@ case "$TARGET_OS" in
;;
OS_ANDROID_CROSSCOMPILE)
PLATFORM=OS_ANDROID
COMMON_FLAGS="$COMMON_FLAGS -fno-builtin-memcmp -D_REENTRANT -DOS_ANDROID -DROCKSDB_PLATFORM_POSIX"
COMMON_FLAGS="$COMMON_FLAGS -fno-builtin-memcmp -D_REENTRANT -DOS_ANDROID -DLEVELDB_PLATFORM_POSIX"
PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS " # All pthread features are in the Android C library
# PORT_FILES=port/android/android.cc
CROSS_COMPILE=true
@@ -183,19 +184,17 @@ if [ "$CROSS_COMPILE" = "true" -o "$FBCODE_BUILD" = "true" ]; then
# Also don't need any compilation tests if compiling on fbcode
true
else
if ! test $ROCKSDB_DISABLE_FALLOCATE; then
# Test whether fallocate is available
$CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null <<EOF
#include <fcntl.h>
#include <linux/falloc.h>
int main() {
int fd = open("/dev/null", 0);
fallocate(fd, FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE, 0, 1024);
}
# Test whether fallocate is available
$CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null <<EOF
#include <fcntl.h>
#include <linux/falloc.h>
int main() {
int fd = open("/dev/null", 0);
fallocate(fd, FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE, 0, 1024);
}
EOF
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_FALLOCATE_PRESENT"
fi
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_FALLOCATE_PRESENT"
fi
# Test whether Snappy library is installed
@@ -282,7 +281,7 @@ EOF
# Test whether numa is available
$CXX $CFLAGS -x c++ - -o /dev/null -lnuma 2>/dev/null <<EOF
#include <numa.h>
#include <numaif.h>
#inlcude <numaif.h>
int main() {}
EOF
if [ "$?" = 0 ]; then
@@ -291,27 +290,15 @@ EOF
JAVA_LDFLAGS="$JAVA_LDFLAGS -lnuma"
fi
# Test whether tbb is available
$CXX $CFLAGS $LDFLAGS -x c++ - -o /dev/null -ltbb 2>/dev/null <<EOF
#include <tbb/tbb.h>
int main() {}
EOF
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DTBB"
PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -ltbb"
JAVA_LDFLAGS="$JAVA_LDFLAGS -ltbb"
fi
# Test whether jemalloc is available
if echo 'int main() {}' | $CXX $CFLAGS -x c++ - -o /dev/null -ljemalloc \
2>/dev/null; then
2>/dev/null; then
PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -ljemalloc"
JAVA_LDFLAGS="$JAVA_LDFLAGS -ljemalloc"
JEMALLOC=1
else
# jemalloc is not available. Let's try tcmalloc
if echo 'int main() {}' | $CXX $CFLAGS -x c++ - -o /dev/null \
-ltcmalloc 2>/dev/null; then
-ltcmalloc 2>/dev/null; then
PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -ltcmalloc"
JAVA_LDFLAGS="$JAVA_LDFLAGS -ltcmalloc"
fi
@@ -328,55 +315,6 @@ EOF
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_MALLOC_USABLE_SIZE"
fi
# Test whether PTHREAD_MUTEX_ADAPTIVE_NP mutex type is available
$CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null <<EOF
#include <pthread.h>
int main() {
int x = PTHREAD_MUTEX_ADAPTIVE_NP;
return 0;
}
EOF
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_PTHREAD_ADAPTIVE_MUTEX"
fi
# Test whether backtrace is available
$CXX $CFLAGS -x c++ - -o /dev/null 2>/dev/null <<EOF
#include <execinfo.h>>
int main() {
void* frames[1];
backtrace_symbols(frames, backtrace(frames, 1));
return 0;
}
EOF
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_BACKTRACE"
else
# Test whether execinfo library is installed
$CXX $CFLAGS -lexecinfo -x c++ - -o /dev/null 2>/dev/null <<EOF
#include <execinfo.h>
int main() {
void* frames[1];
backtrace_symbols(frames, backtrace(frames, 1));
}
EOF
if [ "$?" = 0 ]; then
COMMON_FLAGS="$COMMON_FLAGS -DROCKSDB_BACKTRACE"
PLATFORM_LDFLAGS="$PLATFORM_LDFLAGS -lexecinfo"
JAVA_LDFLAGS="$JAVA_LDFLAGS -lexecinfo"
fi
fi
# Test if -pg is supported
$CXX $CFLAGS -pg -x c++ - -o /dev/null 2>/dev/null <<EOF
int main() {
return 0;
}
EOF
if [ "$?" = 0 ]; then
PROFILING_FLAGS=-pg
fi
fi
# TODO(tec): Fix -Wshorten-64-to-32 errors on FreeBSD and enable the warning.
@@ -416,15 +354,7 @@ if test "$USE_SSE"; then
# if Intel SSE instruction set is supported, set USE_SSE=1
COMMON_FLAGS="$COMMON_FLAGS -msse -msse4.2 "
elif test -z "$PORTABLE"; then
if test -n "`echo $TARGET_ARCHITECTURE | grep ^ppc64`"; then
# Tune for this POWER processor, treating '+' models as base models
POWER=`LD_SHOW_AUXV=1 /bin/true | grep AT_PLATFORM | grep -E -o power[0-9]+`
COMMON_FLAGS="$COMMON_FLAGS -mcpu=$POWER -mtune=$POWER "
elif test -n "`echo $TARGET_ARCHITECTURE | grep ^s390x`"; then
COMMON_FLAGS="$COMMON_FLAGS -march=z10 "
else
COMMON_FLAGS="$COMMON_FLAGS -march=native "
fi
COMMON_FLAGS="$COMMON_FLAGS -march=native "
fi
PLATFORM_CCFLAGS="$PLATFORM_CCFLAGS $COMMON_FLAGS"
@@ -457,8 +387,3 @@ echo "ROCKSDB_MINOR=$ROCKSDB_MINOR" >> "$OUTPUT"
echo "ROCKSDB_PATCH=$ROCKSDB_PATCH" >> "$OUTPUT"
echo "CLANG_SCAN_BUILD=$CLANG_SCAN_BUILD" >> "$OUTPUT"
echo "CLANG_ANALYZER=$CLANG_ANALYZER" >> "$OUTPUT"
echo "PROFILING_FLAGS=$PROFILING_FLAGS" >> "$OUTPUT"
if test -n "$JEMALLOC"; then
echo "JEMALLOC=1" >> "$OUTPUT"
fi
echo "LUA_PATH=$LUA_PATH" >> "$OUTPUT"
-135
View File
@@ -1,135 +0,0 @@
#!/bin/bash
#
# Copyright (c) 2016, Facebook. All rights reserved.
#
# Overall wrapper script for RocksDB continuous builds. The implementation is a
# trivial pulling scheme. We loop infinitely, check if any new changes have been
# committed, if yes then trigger a Sandcastle run, and finally go to sleep again
# for a certain interval.
#
SRC_GIT_REPO=/data/git/rocksdb
error=0
function log {
DATE=`date +%Y-%m-%d:%H:%M:%S`
echo $DATE $@
}
function log_err {
log "ERROR: $@ Error code: $error."
}
function update_repo_status {
# Update the parent first.
pushd $SRC_GIT_REPO
# This is a fatal error. Something in the environment isn't right and we will
# terminate the execution.
error=$?
if [ ! $error -eq 0 ]; then
log_err "Where is $SRC_GIT_REPO?"
exit $error
fi
HTTPS_PROXY=fwdproxy:8080 git fetch -f
error=$?
if [ ! $error -eq 0 ]; then
log_err "git fetch -f failed."
popd
return $error
fi
git update-ref refs/heads/master refs/remotes/origin/master
error=$?
if [ ! $error -eq 0 ]; then
log_err "git update-ref failed."
popd
return $error
fi
popd
# We're back in an instance-specific directory. Get the latest changes.
git pull --rebase
error=$?
if [ ! $error -eq 0 ]; then
log_err "git pull --rebase failed."
return $error
fi
}
#
# Execution starts here.
#
# Path to the determinator from the root of the RocksDB repo.
CONTRUN_DETERMINATOR=./arcanist_util/config/RocksDBCommonHelper.php
# Value of the previous commit.
PREV_COMMIT=
log "Starting to monitor for new RocksDB changes ..."
log "Running under `pwd` as `whoami`."
# Paranoia. Make sure that we're using the right branch.
git checkout master
error=$?
if [ ! $error -eq 0 ]; then
log_err "This is not good. Can't checkout master. Bye-bye!"
exit 1
fi
# We'll run forever and let the execution environment terminate us if we'll
# exceed whatever timeout is set for the job.
while true;
do
# Get the latest changes committed.
update_repo_status
error=$?
if [ $error -eq 0 ]; then
LAST_COMMIT=`git log -1 | head -1 | grep commit | awk '{ print $2; }'`
log "Last commit is '$LAST_COMMIT', previous commit is '$PREV_COMMIT'."
if [ "$PREV_COMMIT" == "$LAST_COMMIT" ]; then
log "There were no changes since the last time I checked. Going to sleep."
else
if [ ! -z "$LAST_COMMIT" ]; then
log "New code has been committed or previous commit not known. " \
"Will trigger the tests."
PREV_COMMIT=$LAST_COMMIT
log "Updated previous commit to '$PREV_COMMIT'."
#
# This is where we'll trigger the Sandcastle run. The values for
# HTTPS_APP_VALUE and HTTPS_APP_VALUE will be set in the container we're
# running in.
#
POST_RECEIVE_HOOK=1 php $CONTRUN_DETERMINATOR
error=$?
if [ $error -eq 0 ]; then
log "Sandcastle run successfully triggered."
else
log_err "Failed to trigger Sandcastle run."
fi
else
log_err "Previous commit not updated. Don't know what the last one is."
fi
fi
else
log_err "Getting latest changes failed. Will skip running tests for now."
fi
# Always sleep, even if errors happens while trying to determine the latest
# commit. This will prevent us terminating in case of transient errors.
log "Will go to sleep for 5 minutes."
sleep 5m
done
-18
View File
@@ -1,18 +0,0 @@
GCC_BASE=/mnt/gvfs/third-party2/gcc/cf7d14c625ce30bae1a4661c2319c5a283e4dd22/4.8.1/centos6-native/cc6c9dc
CLANG_BASE=/mnt/gvfs/third-party2/llvm-fb/8598c375b0e94e1448182eb3df034704144a838d/stable/centos6-native/3f16ddd
LIBGCC_BASE=/mnt/gvfs/third-party2/libgcc/d6e0a7da6faba45f5e5b1638f9edd7afc2f34e7d/4.8.1/gcc-4.8.1-glibc-2.17/8aac7fc
GLIBC_BASE=/mnt/gvfs/third-party2/glibc/d282e6e8f3d20f4e40a516834847bdc038e07973/2.17/gcc-4.8.1-glibc-2.17/99df8fc
SNAPPY_BASE=/mnt/gvfs/third-party2/snappy/8c38a4c1e52b4c2cc8a9cdc31b9c947ed7dbfcb4/1.1.3/gcc-4.8.1-glibc-2.17/c3f970a
ZLIB_BASE=/mnt/gvfs/third-party2/zlib/0882df3713c7a84f15abe368dc004581f20b39d7/1.2.8/gcc-4.8.1-glibc-2.17/c3f970a
BZIP2_BASE=/mnt/gvfs/third-party2/bzip2/740325875f6729f42d28deaa2147b0854f3a347e/1.0.6/gcc-4.8.1-glibc-2.17/c3f970a
LZ4_BASE=/mnt/gvfs/third-party2/lz4/0e790b441e2d9acd68d51e1d2e028f88c6a79ddf/r131/gcc-4.8.1-glibc-2.17/c3f970a
ZSTD_BASE=/mnt/gvfs/third-party2/zstd/9455f75ff7f4831dc9fda02a6a0f8c68922fad8f/1.0.0/gcc-4.8.1-glibc-2.17/c3f970a
GFLAGS_BASE=/mnt/gvfs/third-party2/gflags/f001a51b2854957676d07306ef3abf67186b5c8b/2.1.1/gcc-4.8.1-glibc-2.17/c3f970a
JEMALLOC_BASE=/mnt/gvfs/third-party2/jemalloc/fc8a13ca1fffa4d0765c716c5a0b49f0c107518f/master/gcc-4.8.1-glibc-2.17/8d31e51
NUMA_BASE=/mnt/gvfs/third-party2/numa/17c514c4d102a25ca15f4558be564eeed76f4b6a/2.0.8/gcc-4.8.1-glibc-2.17/c3f970a
LIBUNWIND_BASE=/mnt/gvfs/third-party2/libunwind/ad576de2a1ea560c4d3434304f0fc4e079bede42/trunk/gcc-4.8.1-glibc-2.17/675d945
TBB_BASE=/mnt/gvfs/third-party2/tbb/9d9a554877d0c5bef330fe818ab7178806dd316a/4.0_update2/gcc-4.8.1-glibc-2.17/c3f970a
KERNEL_HEADERS_BASE=/mnt/gvfs/third-party2/kernel-headers/7c111ff27e0c466235163f00f280a9d617c3d2ec/4.0.9-36_fbk5_2933_gd092e3f/gcc-4.8.1-glibc-2.17/da39a3e
BINUTILS_BASE=/mnt/gvfs/third-party2/binutils/b7fd454c4b10c6a81015d4524ed06cdeab558490/2.26/centos6-native/da39a3e
VALGRIND_BASE=/mnt/gvfs/third-party2/valgrind/d7f4d4d86674a57668e3a96f76f0e17dd0eb8765/3.8.1/gcc-4.8.1-glibc-2.17/c3f970a
LUA_BASE=/mnt/gvfs/third-party2/lua/61e4abf5813bbc39bc4f548757ccfcadde175a48/5.2.3/centos6-native/730f94e
-167
View File
@@ -1,167 +0,0 @@
# Copyright (c) 2016-present, Facebook, Inc. All rights reserved.
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree. An additional grant
# of patent rights can be found in the PATENTS file in the same directory.
'''Filter for error messages in test output:
- Receives merged stdout/stderr from test on stdin
- Finds patterns of known error messages for test name (first argument)
- Prints those error messages to stdout
'''
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from __future__ import unicode_literals
import re
import sys
class ErrorParserBase(object):
def parse_error(self, line):
'''Parses a line of test output. If it contains an error, returns a
formatted message describing the error; otherwise, returns None.
Subclasses must override this method.
'''
raise NotImplementedError
class GTestErrorParser(ErrorParserBase):
'''A parser that remembers the last test that began running so it can print
that test's name upon detecting failure.
'''
_GTEST_NAME_PATTERN = re.compile(r'\[ RUN \] (\S+)$')
# format: '<filename or "unknown file">:<line #>: Failure'
_GTEST_FAIL_PATTERN = re.compile(r'(unknown file|\S+:\d+): Failure$')
def __init__(self):
self._last_gtest_name = 'Unknown test'
def parse_error(self, line):
gtest_name_match = self._GTEST_NAME_PATTERN.match(line)
if gtest_name_match:
self._last_gtest_name = gtest_name_match.group(1)
return None
gtest_fail_match = self._GTEST_FAIL_PATTERN.match(line)
if gtest_fail_match:
return '%s failed: %s' % (
self._last_gtest_name, gtest_fail_match.group(1))
return None
class MatchErrorParser(ErrorParserBase):
'''A simple parser that returns the whole line if it matches the pattern.
'''
def __init__(self, pattern):
self._pattern = re.compile(pattern)
def parse_error(self, line):
if self._pattern.match(line):
return line
return None
class CompilerErrorParser(MatchErrorParser):
def __init__(self):
# format: '<filename>:<line #>:<column #>: error: <error msg>'
super(CompilerErrorParser, self).__init__(r'\S+:\d+:\d+: error:')
class ScanBuildErrorParser(MatchErrorParser):
def __init__(self):
super(ScanBuildErrorParser, self).__init__(
r'scan-build: \d+ bugs found.$')
class DbCrashErrorParser(MatchErrorParser):
def __init__(self):
super(DbCrashErrorParser, self).__init__(r'\*\*\*.*\^$|TEST FAILED.')
class WriteStressErrorParser(MatchErrorParser):
def __init__(self):
super(WriteStressErrorParser, self).__init__(
r'ERROR: write_stress died with exitcode=\d+')
class AsanErrorParser(MatchErrorParser):
def __init__(self):
super(AsanErrorParser, self).__init__(
r'==\d+==ERROR: AddressSanitizer:')
class UbsanErrorParser(MatchErrorParser):
def __init__(self):
# format: '<filename>:<line #>:<column #>: runtime error: <error msg>'
super(UbsanErrorParser, self).__init__(r'\S+:\d+:\d+: runtime error:')
class ValgrindErrorParser(MatchErrorParser):
def __init__(self):
# just grab the summary, valgrind doesn't clearly distinguish errors
# from other log messages.
super(ValgrindErrorParser, self).__init__(r'==\d+== ERROR SUMMARY:')
class CompatErrorParser(MatchErrorParser):
def __init__(self):
super(CompatErrorParser, self).__init__(r'==== .*[Ee]rror.* ====$')
class TsanErrorParser(MatchErrorParser):
def __init__(self):
super(TsanErrorParser, self).__init__(r'WARNING: ThreadSanitizer:')
_TEST_NAME_TO_PARSERS = {
'punit': [CompilerErrorParser, GTestErrorParser],
'unit': [CompilerErrorParser, GTestErrorParser],
'release': [CompilerErrorParser, GTestErrorParser],
'unit_481': [CompilerErrorParser, GTestErrorParser],
'release_481': [CompilerErrorParser, GTestErrorParser],
'clang_unit': [CompilerErrorParser, GTestErrorParser],
'clang_release': [CompilerErrorParser, GTestErrorParser],
'clang_analyze': [CompilerErrorParser, ScanBuildErrorParser],
'code_cov': [CompilerErrorParser, GTestErrorParser],
'unity': [CompilerErrorParser, GTestErrorParser],
'lite': [CompilerErrorParser],
'lite_test': [CompilerErrorParser, GTestErrorParser],
'stress_crash': [CompilerErrorParser, DbCrashErrorParser],
'write_stress': [CompilerErrorParser, WriteStressErrorParser],
'asan': [CompilerErrorParser, GTestErrorParser, AsanErrorParser],
'asan_crash': [CompilerErrorParser, AsanErrorParser, DbCrashErrorParser],
'ubsan': [CompilerErrorParser, GTestErrorParser, UbsanErrorParser],
'ubsan_crash': [CompilerErrorParser, UbsanErrorParser, DbCrashErrorParser],
'valgrind': [CompilerErrorParser, GTestErrorParser, ValgrindErrorParser],
'tsan': [CompilerErrorParser, GTestErrorParser, TsanErrorParser],
'format_compatible': [CompilerErrorParser, CompatErrorParser],
'run_format_compatible': [CompilerErrorParser, CompatErrorParser],
'no_compression': [CompilerErrorParser, GTestErrorParser],
'run_no_compression': [CompilerErrorParser, GTestErrorParser],
'regression': [CompilerErrorParser],
'run_regression': [CompilerErrorParser],
}
def main():
if len(sys.argv) != 2:
return 'Usage: %s <test name>' % sys.argv[0]
test_name = sys.argv[1]
if test_name not in _TEST_NAME_TO_PARSERS:
return 'Unknown test name: %s' % test_name
error_parsers = []
for parser_cls in _TEST_NAME_TO_PARSERS[test_name]:
error_parsers.append(parser_cls())
for line in sys.stdin:
line = line.strip()
for error_parser in error_parsers:
error_msg = error_parser.parse_error(line)
if error_msg is not None:
print(error_msg)
if __name__ == '__main__':
sys.exit(main())
+47 -47
View File
@@ -4,82 +4,85 @@
# fbcode settings. It uses the latest g++ compiler and also
# uses jemalloc
BASEDIR=`dirname $BASH_SOURCE`
source "$BASEDIR/dependencies_4.8.1.sh"
# location of libgcc
LIBGCC_BASE="/mnt/gvfs/third-party2/libgcc/d00277f4559e261ed0a81f30f23c0ce5564e359e/4.8.1/gcc-4.8.1-glibc-2.17/8aac7fc"
LIBGCC_INCLUDE="$LIBGCC_BASE/include"
LIBGCC_LIBS=" -L $LIBGCC_BASE/lib"
LIBGCC_LIBS=" -L $LIBGCC_BASE/libs"
# location of glibc
GLIBC_INCLUDE="$GLIBC_BASE/include"
GLIBC_LIBS=" -L $GLIBC_BASE/lib"
GLIBC_REV=0600c95b31226b5e535614c590677d87c62d8016
GLIBC_INCLUDE="/mnt/gvfs/third-party2/glibc/$GLIBC_REV/2.17/gcc-4.8.1-glibc-2.17/99df8fc/include"
GLIBC_LIBS=" -L /mnt/gvfs/third-party2/glibc/$GLIBC_REV/2.17/gcc-4.8.1-glibc-2.17/99df8fc/lib"
# location of snappy headers and libraries
SNAPPY_INCLUDE=" -I $SNAPPY_BASE/include"
SNAPPY_LIBS=" $SNAPPY_BASE/lib/libsnappy.a"
SNAPPY_REV=cbf6f1f209e5bd160bdc5d971744e039f36b1566
SNAPPY_INCLUDE=" -I /mnt/gvfs/third-party2/snappy/$SNAPPY_REV/1.1.3/gcc-4.8.1-glibc-2.17/c3f970a/include"
SNAPPY_LIBS=" /mnt/gvfs/third-party2/snappy/$SNAPPY_REV/1.1.3/gcc-4.8.1-glibc-2.17/c3f970a/lib/libsnappy.a"
# location of zlib headers and libraries
ZLIB_INCLUDE=" -I $ZLIB_BASE/include"
ZLIB_LIBS=" $ZLIB_BASE/lib/libz.a"
ZLIB_REV=6d39cb54708049f527e713ad19f2aadb9d3667e8
ZLIB_INCLUDE=" -I /mnt/gvfs/third-party2/zlib/$ZLIB_REV/1.2.8/gcc-4.8.1-glibc-2.17/c3f970a/include"
ZLIB_LIBS=" /mnt/gvfs/third-party2/zlib/$ZLIB_REV/1.2.8/gcc-4.8.1-glibc-2.17/c3f970a/lib/libz.a"
# location of bzip headers and libraries
BZIP2_INCLUDE=" -I $BZIP2_BASE/include/"
BZIP2_LIBS=" $BZIP2_BASE/lib/libbz2.a"
BZIP_REV=d6c789bfc2ec4c51a63d66df2878926b8158cde8
BZIP_INCLUDE=" -I /mnt/gvfs/third-party2/bzip2/$BZIP_REV/1.0.6/gcc-4.8.1-glibc-2.17/c3f970a/include/"
BZIP_LIBS=" /mnt/gvfs/third-party2/bzip2/$BZIP_REV/1.0.6/gcc-4.8.1-glibc-2.17/c3f970a/lib/libbz2.a"
LZ4_INCLUDE=" -I $LZ4_BASE/include"
LZ4_LIBS=" $LZ4_BASE/lib/liblz4.a"
LZ4_REV=6858fac689e0f92e584224d91bdb0e39f6c8320d
LZ4_INCLUDE=" -I /mnt/gvfs/third-party2/lz4/$LZ4_REV/r131/gcc-4.8.1-glibc-2.17/c3f970a/include"
LZ4_LIBS=" /mnt/gvfs/third-party2/lz4/$LZ4_REV/r131/gcc-4.8.1-glibc-2.17/c3f970a/lib/liblz4.a"
ZSTD_INCLUDE=" -I $ZSTD_BASE/include"
ZSTD_LIBS=" $ZSTD_BASE/lib/libzstd.a"
ZSTD_REV=810b81b4705def5243e998b54701f3c504e4009e
ZSTD_INCLUDE=" -I /mnt/gvfs/third-party2/zstd/$ZSTD_REV/0.4.2/gcc-4.8.1-glibc-2.17/c3f970a/include"
ZSTD_LIBS=" /mnt/gvfs/third-party2/zstd/$ZSTD_REV/0.4.2/gcc-4.8.1-glibc-2.17/c3f970a/lib/libzstd.a"
# location of gflags headers and libraries
GFLAGS_INCLUDE=" -I $GFLAGS_BASE/include/"
GFLAGS_LIBS=" $GFLAGS_BASE/lib/libgflags.a"
GFLAGS_REV=c7275a4ceae0aca0929e56964a31dafc53c1ee96
GFLAGS_INCLUDE=" -I /mnt/gvfs/third-party2/gflags/$GFLAGS_REV/2.1.1/gcc-4.8.1-glibc-2.17/c3f970a/include/"
GFLAGS_LIBS=" /mnt/gvfs/third-party2/gflags/$GFLAGS_REV/2.1.1/gcc-4.8.1-glibc-2.17/c3f970a/lib/libgflags.a"
# location of jemalloc
JEMALLOC_INCLUDE=" -I $JEMALLOC_BASE/include"
JEMALLOC_LIB="$JEMALLOC_BASE/lib/libjemalloc.a"
JEMALLOC_REV=c370265e58c4b6602e798df23335a1e9913dae52
JEMALLOC_INCLUDE=" -I /mnt/gvfs/third-party2/jemalloc/$JEMALLOC_REV/4.0.3/gcc-4.8.1-glibc-2.17/8d31e51/include"
JEMALLOC_LIB="/mnt/gvfs/third-party2/jemalloc/$JEMALLOC_REV/4.0.3/gcc-4.8.1-glibc-2.17/8d31e51/lib/libjemalloc.a"
# location of numa
NUMA_INCLUDE=" -I $NUMA_BASE/include/"
NUMA_LIB=" $NUMA_BASE/lib/libnuma.a"
NUMA_REV=ae54a5ed22cdabb1c6446dce4e8ffae5b4446d73
NUMA_INCLUDE=" -I /mnt/gvfs/third-party2/numa/$NUMA_REV/2.0.8/gcc-4.8.1-glibc-2.17/c3f970a/include/"
NUMA_LIB=" /mnt/gvfs/third-party2/numa/$NUMA_REV/2.0.8/gcc-4.8.1-glibc-2.17/c3f970a/lib/libnuma.a"
# location of libunwind
LIBUNWIND="$LIBUNWIND_BASE/lib/libunwind.a"
# location of tbb
TBB_INCLUDE=" -isystem $TBB_BASE/include/"
TBB_LIBS="$TBB_BASE/lib/libtbb.a"
LIBUNWIND_REV=121f1a75c4414683aea8c70b761bfaf187f7c1a3
LIBUNWIND="/mnt/gvfs/third-party2/libunwind/$LIBUNWIND_REV/trunk/gcc-4.8.1-glibc-2.17/675d945/lib/libunwind.a"
# use Intel SSE support for checksum calculations
export USE_SSE=1
BINUTILS="$BINUTILS_BASE/bin"
BINUTILS="/mnt/gvfs/third-party2/binutils/75670d0d8ef4891fd1ec2a7513ef01cd002c823b/2.25/centos6-native/da39a3e/bin"
AR="$BINUTILS/ar"
DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP2_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE $TBB_INCLUDE"
DEPS_INCLUDE="$SNAPPY_INCLUDE $ZLIB_INCLUDE $BZIP_INCLUDE $LZ4_INCLUDE $ZSTD_INCLUDE $GFLAGS_INCLUDE $NUMA_INCLUDE"
GCC_BASE="/mnt/gvfs/third-party2/gcc/c0064002d2609ab649603f769f0bd110bbe48029/4.8.1/centos6-native/cc6c9dc"
STDLIBS="-L $GCC_BASE/lib64"
if [ -z "$USE_CLANG" ]; then
# gcc
CC="$GCC_BASE/bin/gcc"
CXX="$GCC_BASE/bin/g++"
CFLAGS="-B$BINUTILS/gold -m64 -mtune=generic"
CFLAGS+=" -isystem $GLIBC_INCLUDE"
CFLAGS+=" -isystem $LIBGCC_INCLUDE"
JEMALLOC=1
else
# clang
CLANG_BIN="$CLANG_BASE/bin"
CLANG_LIB="$CLANG_BASE/lib"
CLANG_INCLUDE="$CLANG_LIB/clang/*/include"
CC="$CLANG_BIN/clang"
CXX="$CLANG_BIN/clang++"
# clang
CLANG_BASE="/mnt/gvfs/third-party2/clang/ab054e9a490a8fd4537c0b6ec56e5c91c0f81c91/3.4"
CLANG_INCLUDE="$CLANG_BASE/gcc-4.8.1-glibc-2.17/fb0f730/lib/clang/3.4/include"
CC="$CLANG_BASE/centos6-native/9cefd8a/bin/clang"
CXX="$CLANG_BASE/centos6-native/9cefd8a/bin/clang++"
KERNEL_HEADERS_INCLUDE="$KERNEL_HEADERS_BASE/include/"
KERNEL_HEADERS_INCLUDE="/mnt/gvfs/third-party2/kernel-headers/1a48835975c66d30e47770ec419758ed3b9ba010/3.10.62-62_fbk17_03959_ge29cc63/gcc-4.8.1-glibc-2.17/da39a3e/include/"
CFLAGS="-B$BINUTILS/gold -nostdinc -nostdlib"
CFLAGS+=" -isystem $LIBGCC_BASE/include/c++/4.8.1 "
@@ -94,22 +97,19 @@ fi
CFLAGS+=" $DEPS_INCLUDE"
CFLAGS+=" -DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX -DROCKSDB_FALLOCATE_PRESENT -DROCKSDB_MALLOC_USABLE_SIZE"
CFLAGS+=" -DSNAPPY -DGFLAGS=google -DZLIB -DBZIP2 -DLZ4 -DZSTD -DNUMA -DTBB"
CFLAGS+=" -DSNAPPY -DGFLAGS=google -DZLIB -DBZIP2 -DLZ4 -DZSTD -DNUMA"
CXXFLAGS+=" $CFLAGS"
EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP2_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB $TBB_LIBS"
EXEC_LDFLAGS=" $SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS $NUMA_LIB"
EXEC_LDFLAGS+=" -Wl,--dynamic-linker,/usr/local/fbcode/gcc-4.8.1-glibc-2.17/lib/ld.so"
EXEC_LDFLAGS+=" $LIBUNWIND"
EXEC_LDFLAGS+=" -Wl,-rpath=/usr/local/fbcode/gcc-4.8.1-glibc-2.17/lib"
# required by libtbb
EXEC_LDFLAGS+=" -ldl"
PLATFORM_LDFLAGS="$LIBGCC_LIBS $GLIBC_LIBS $STDLIBS -lgcc -lstdc++"
EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP2_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS"
EXEC_LDFLAGS_SHARED="$SNAPPY_LIBS $ZLIB_LIBS $BZIP_LIBS $LZ4_LIBS $ZSTD_LIBS $GFLAGS_LIBS"
VALGRIND_VER="$VALGRIND_BASE/bin/"
VALGRIND_REV=af85c56f424cd5edfc2c97588299b44ecdec96bb
VALGRIND_VER="/mnt/gvfs/third-party2/valgrind/$VALGRIND_REV/3.8.1/gcc-4.8.1-glibc-2.17/c3f970a/bin/"
LUA_PATH="$LUA_BASE"
export CC CXX AR CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE LUA_PATH
export CC CXX AR CFLAGS CXXFLAGS EXEC_LDFLAGS EXEC_LDFLAGS_SHARED VALGRIND_VER JEMALLOC_LIB JEMALLOC_INCLUDE
File diff suppressed because it is too large. Load diff
+2 -2
View File
@@ -1,5 +1,5 @@
#!/bin/bash
# Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
# Copyright (c) 2013, Facebook, Inc. All rights reserved.
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree. An additional grant
# of patent rights can be found in the PATENTS file in the same directory.
@@ -10,7 +10,7 @@ then
GIT="git"
fi
# Print out the colored progress info so that it can be brainlessly
# Print out the colored progress info so that it can be brainlessly
# distinguished by users.
function title() {
echo -e "\033[1;32m$*\033[0m"
+2 -2
View File
@@ -65,8 +65,8 @@ function main() {
if [[ -d /vagrant ]]; then
if [[ $OS = "ubuntu" ]]; then
package g++-4.8
export CXX=g++-4.8
package g++-4.7
export CXX=g++-4.7
# the deb would depend on libgflags2, but the static lib is the only thing
# installed by make install
-208
View File
@@ -1,208 +0,0 @@
#!/usr/local/fbcode/gcc-4.8.1-glibc-2.17-fb/bin/python2.7
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from __future__ import unicode_literals
import argparse
import commands
import subprocess
import sys
import re
import os
import time
#
# Simple logger
#
class Log:
def __init__(self, filename):
self.filename = filename
self.f = open(self.filename, 'w+', 0)
def caption(self, str):
line = "\n##### %s #####\n" % str
if self.f:
self.f.write("%s \n" % line)
else:
print(line)
def error(self, str):
data = "\n\n##### ERROR ##### %s" % str
if self.f:
self.f.write("%s \n" % data)
else:
print(data)
def log(self, str):
if self.f:
self.f.write("%s \n" % str)
else:
print(str)
#
# Shell Environment
#
class Env(object):
def __init__(self, logfile, tests):
self.tests = tests
self.log = Log(logfile)
def shell(self, cmd, path=os.getcwd()):
if path:
os.chdir(path)
self.log.log("==== shell session ===========================")
self.log.log("%s> %s" % (path, cmd))
status = subprocess.call("cd %s; %s" % (path, cmd), shell=True,
stdout=self.log.f, stderr=self.log.f)
self.log.log("status = %s" % status)
self.log.log("============================================== \n\n")
return status
def GetOutput(self, cmd, path=os.getcwd()):
if path:
os.chdir(path)
self.log.log("==== shell session ===========================")
self.log.log("%s> %s" % (path, cmd))
status, out = commands.getstatusoutput(cmd)
self.log.log("status = %s" % status)
self.log.log("out = %s" % out)
self.log.log("============================================== \n\n")
return status, out
#
# Pre-commit checker
#
class PreCommitChecker(Env):
def __init__(self, args):
Env.__init__(self, args.logfile, args.tests)
self.ignore_failure = args.ignore_failure
#
# Get commands for a given job from the determinator file
#
def get_commands(self, test):
status, out = self.GetOutput(
"RATIO=1 build_tools/rocksdb-lego-determinator %s" % test, ".")
return status, out
#
# Run a specific CI job
#
def run_test(self, test):
self.log.caption("Running test %s locally" % test)
# get commands for the CI job determinator
status, cmds = self.get_commands(test)
if status != 0:
self.log.error("Error getting commands for test %s" % test)
return False
# Parse the JSON to extract the commands to run
cmds = re.findall("'shell':'([^\']*)'", cmds)
if len(cmds) == 0:
self.log.log("No commands found")
return False
# Run commands
for cmd in cmds:
# Replace J=<..> with the local environment variable
if "J" in os.environ:
cmd = cmd.replace("J=1", "J=%s" % os.environ["J"])
cmd = cmd.replace("make ", "make -j%s " % os.environ["J"])
# Run the command
status = self.shell(cmd, ".")
if status != 0:
self.log.error("Error running command %s for test %s"
% (cmd, test))
return False
return True
#
# Run specified CI jobs
#
def run_tests(self):
if not self.tests:
self.log.error("Invalid args. Please provide tests")
return False
self.print_separator()
self.print_row("TEST", "RESULT")
self.print_separator()
result = True
for test in self.tests:
start_time = time.time()
self.print_test(test)
result = self.run_test(test)
elapsed_min = (time.time() - start_time) / 60
if not result:
self.log.error("Error running test %s" % test)
self.print_result("FAIL (%dm)" % elapsed_min)
if not self.ignore_failure:
return False
result = False
else:
self.print_result("PASS (%dm)" % elapsed_min)
self.print_separator()
return result
#
# Print a line
#
def print_separator(self):
print("".ljust(60, "-"))
#
# Print two colums
#
def print_row(self, c0, c1):
print("%s%s" % (c0.ljust(40), c1.ljust(20)))
def print_test(self, test):
print(test.ljust(40), end="")
sys.stdout.flush()
def print_result(self, result):
print(result.ljust(20))
#
# Main
#
parser = argparse.ArgumentParser(description='RocksDB pre-commit checker.')
# --log <logfile>
parser.add_argument('--logfile', default='/tmp/precommit-check.log',
help='Log file. Default is /tmp/precommit-check.log')
# --ignore_failure
parser.add_argument('--ignore_failure', action='store_true', default=False,
help='Stop when an error occurs')
# <test ....>
parser.add_argument('tests', nargs='+',
help='CI test(s) to run. e.g: unit punit asan tsan ubsan')
args = parser.parse_args()
checker = PreCommitChecker(args)
print("Please follow log %s" % checker.log.filename)
if not checker.run_tests():
print("Error running tests. Please check log file %s"
% checker.log.filename)
sys.exit(1)
sys.exit(0)
+3 -3
View File
@@ -243,7 +243,7 @@ make release
--bloom_bits=10 \
--num=$((NUM / 4)) \
--reads=$((NUM / 4)) \
--benchmark_write_rate_limit=$(( 110 * 1024 )) \
--writes_per_second=1000 \
--write_buffer_size=100000000 \
--cache_size=6442450944 \
--cache_numshardbits=6 \
@@ -329,7 +329,7 @@ common_in_mem_args="--db=/dev/shm/rocksdb \
--use_existing_db=1 \
--duration=600 \
--threads=32 \
--benchmark_write_rate_limit=9502720 > ${STAT_FILE}.readwhilewriting_in_ram
--writes_per_second=81920 > ${STAT_FILE}.readwhilewriting_in_ram
# Seekrandomwhilewriting
./db_bench \
@@ -342,7 +342,7 @@ common_in_mem_args="--db=/dev/shm/rocksdb \
--use_tailing_iterator=1 \
--duration=600 \
--threads=32 \
--benchmark_write_rate_limit=9502720 > ${STAT_FILE}.seekwhilewriting_in_ram
--writes_per_second=81920 > ${STAT_FILE}.seekwhilewriting_in_ram
# measure fillseq with bunch of column families
./db_bench \
+29 -232
View File
@@ -8,7 +8,7 @@
# Input Value
# -------------------------------------------------------------------------
# EMAIL Email address to report on trigger conditions
# ONCALL Email address to raise a task on failure
# ONCAL Email address to raise a task on failure
# TRIGGER Trigger conditions for email. Valid values are fail, warn, all
# SUBSCRIBER Email addresss to add as subscriber for task
#
@@ -42,9 +42,6 @@ if [ ! -z $ONCALL ]; then
},"
fi
# For now, create the tasks using only the dedicated task creation tool.
CREATE_TASK=
REPORT=
if [[ ! -z $REPORT_EMAIL || ! -z $CREATE_TASK ]]; then
REPORT="'report': [
@@ -59,51 +56,19 @@ fi
CLEANUP_ENV="
{
'name':'Cleanup environment',
'shell':'rm -rf /dev/shm/rocksdb && mkdir /dev/shm/rocksdb && (chmod +t /dev/shm || true) && make clean',
'shell':'rm -rf /dev/shm/rocksdb && mkdir /dev/shm/rocksdb && make clean',
'user':'root'
}"
# We will eventually set the RATIO to 1, but we want do this
# in steps. RATIO=$(nproc) will make it work as J=1
if [ -z $RATIO ]; then
RATIO=$(nproc)
fi
if [ -z $PARALLEL_J ]; then
PARALLEL_J="J=$(expr $(nproc) / ${RATIO})"
fi
if [ -z $PARALLEL_j ]; then
PARALLEL_j="-j$(expr $(nproc) / ${RATIO})"
fi
PARALLELISM="$PARALLEL_J $PARALLEL_j"
DEBUG="OPT=-g"
SHM="TEST_TMPDIR=/dev/shm/rocksdb"
GCC_481="ROCKSDB_FBCODE_BUILD_WITH_481=1"
ASAN="COMPILE_WITH_ASAN=1"
CLANG="USE_CLANG=1"
LITE="OPT=\"-DROCKSDB_LITE -g\""
LITE="OPT=-DROCKSDB_LITE"
TSAN="COMPILE_WITH_TSAN=1"
UBSAN="COMPILE_WITH_UBSAN=1"
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-7u10-64/; export PATH=\$PATH:\$JAVA_HOME/bin"
PARSER="'parser':'python build_tools/error_filter.py $1'"
CONTRUN_NAME="ROCKSDB_CONTRUN_NAME"
# This code is getting called under various scenarios. What we care about is to
# understand when it's called from nightly contruns because in that case we'll
# create tasks for any failures. To follow the existing pattern, we'll check
# the value of $ONCALL. If it's a diff then just call `false` to make sure
# that errors will be properly propagated to the caller.
if [ ! -z $ONCALL ]; then
TASK_CREATION_TOOL="/usr/local/bin/mysql_mtr_filter --rocksdb --oncall $ONCALL"
else
TASK_CREATION_TOOL="false"
fi
PARSER="'parser':'egrep \'Failure|^#|Abort\''"
ARTIFACTS=" 'artifacts': [
{
@@ -142,14 +107,14 @@ PARALLEL_UNIT_TEST_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build and test RocksDB debug version',
'shell':'$DEBUG make -j$(nproc) all && $SHM make check > /dev/null 2>&1 || $CONTRUN_NAME=punit_check $TASK_CREATION_TOOL || cat t/log-*',
'shell':'$DEBUG make -j$(nproc) all && $SHM make check > /dev/null 2>&1 || cat t/log-*',
'user':'root',
$PARSER
},
$CLEANUP_ENV,
{
'name':'Build and test RocksDB debug version under gcc-4.8.1',
'shell':'$GCC_481 $DEBUG make -j$(nproc) all && $SHM make check > /dev/null 2>&1 || $CONTRUN_NAME=punit_check_gcc481 $TASK_CREATION_TOOL || cat t/log-*',
'shell':'$GCC_481 $DEBUG make -j$(nproc) all && $SHM make check > /dev/null 2>&1 || cat t/log-*',
'user':'root',
$PARSER
},
@@ -169,49 +134,7 @@ UNIT_TEST_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build and test RocksDB debug version',
'shell':'$SHM $DEBUG make $PARALLELISM check || $CONTRUN_NAME=check $TASK_CREATION_TOOL',
'user':'root',
$PARSER
},
],
$REPORT
}
]"
#
# RocksDB unit test not under /dev/shm
#
UNIT_TEST_NON_SHM_COMMANDS="[
{
'name':'Rocksdb Unit Test',
'oncall':'$ONCALL',
'timeout': 86400,
'steps': [
$CLEANUP_ENV,
{
'name':'Build and test RocksDB debug version',
'timeout': 86400,
'shell':'$DEBUG make $PARALLELISM check || $CONTRUN_NAME=non_shm_check $TASK_CREATION_TOOL',
'user':'root',
$PARSER
},
],
$REPORT
}
]"
#
# RocksDB release build and unit tests
#
RELEASE_BUILD_COMMANDS="[
{
'name':'Rocksdb Release Build',
'oncall':'$ONCALL',
'steps': [
$CLEANUP_ENV,
{
'name':'Build RocksDB release',
'shell':'make $PARALLEL_j release || $CONTRUN_NAME=release $TASK_CREATION_TOOL',
'shell':'$SHM $DEBUG make J=1 check',
'user':'root',
$PARSER
},
@@ -231,27 +154,7 @@ UNIT_TEST_COMMANDS_481="[
$CLEANUP_ENV,
{
'name':'Build and test RocksDB debug version',
'shell':'$SHM $GCC_481 $DEBUG make $PARALLELISM check || $CONTRUN_NAME=unit_gcc_481_check $TASK_CREATION_TOOL',
'user':'root',
$PARSER
},
],
$REPORT
}
]"
#
# RocksDB release build and unit tests
#
RELEASE_BUILD_COMMANDS_481="[
{
'name':'Rocksdb Release on GCC 4.8.1',
'oncall':'$ONCALL',
'steps': [
$CLEANUP_ENV,
{
'name':'Build RocksDB release on GCC 4.8.1',
'shell':'$GCC_481 make $PARALLEL_j release || $CONTRUN_NAME=release_gcc481 $TASK_CREATION_TOOL',
'shell':'$SHM $GCC_481 $DEBUG make J=1 check',
'user':'root',
$PARSER
},
@@ -271,27 +174,7 @@ CLANG_UNIT_TEST_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build and test RocksDB debug',
'shell':'$CLANG $SHM $DEBUG make $PARALLELISM check || $CONTRUN_NAME=clang_check $TASK_CREATION_TOOL',
'user':'root',
$PARSER
},
],
$REPORT
}
]"
#
# RocksDB release build with CLANG
#
CLANG_RELEASE_BUILD_COMMANDS="[
{
'name':'Rocksdb CLANG Release Build',
'oncall':'$ONCALL',
'steps': [
$CLEANUP_ENV,
{
'name':'Build RocksDB release',
'shell':'$CLANG make $PARALLEL_j release|| $CONTRUN_NAME=clang_release $TASK_CREATION_TOOL',
'shell':'$CLANG $SHM $DEBUG make J=1 check',
'user':'root',
$PARSER
},
@@ -311,7 +194,7 @@ CLANG_ANALYZE_COMMANDS="[
$CLEANUP_ENV,
{
'name':'RocksDB build and analyze',
'shell':'$CLANG $SHM $DEBUG make $PARALLEL_j analyze || $CONTRUN_NAME=clang_analyze $TASK_CREATION_TOOL',
'shell':'$CLANG $SHM $DEBUG make J=1 analyze',
'user':'root',
$PARSER
},
@@ -331,7 +214,7 @@ CODE_COV_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build, test and collect code coverage info',
'shell':'$SHM $DEBUG make $PARALLELISM coverage || $CONTRUN_NAME=coverage $TASK_CREATION_TOOL',
'shell':'$SHM $DEBUG make J=1 coverage',
'user':'root',
$PARSER
},
@@ -351,7 +234,7 @@ UNITY_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build, test unity test',
'shell':'$SHM $DEBUG V=1 make J=1 unity_test || $CONTRUN_NAME=unity_test $TASK_CREATION_TOOL',
'shell':'$SHM $DEBUG V=1 make J=1 unity_test',
'user':'root',
$PARSER
},
@@ -371,7 +254,7 @@ LITE_BUILD_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build RocksDB debug version',
'shell':'$LITE make J=1 static_lib || $CONTRUN_NAME=lite_static_lib $TASK_CREATION_TOOL',
'shell':'$LITE $DEBUG make J=1 static_lib',
'user':'root',
$PARSER
},
@@ -391,7 +274,7 @@ LITE_UNIT_TEST_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build RocksDB debug version',
'shell':'$SHM $LITE make J=1 check || $CONTRUN_NAME=lite_check $TASK_CREATION_TOOL',
'shell':'$SHM $LITE $DEBUG make J=1 check',
'user':'root',
$PARSER
},
@@ -412,14 +295,14 @@ STRESS_CRASH_TEST_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build and run RocksDB debug stress tests',
'shell':'$SHM $DEBUG make J=1 db_stress || $CONTRUN_NAME=db_stress $TASK_CREATION_TOOL',
'shell':'$SHM $DEBUG make J=1 db_stress',
'user':'root',
$PARSER
},
{
'name':'Build and run RocksDB debug crash tests',
'timeout': 86400,
'shell':'$SHM $DEBUG make J=1 crash_test || $CONTRUN_NAME=crash_test $TASK_CREATION_TOOL',
'shell':'$SHM $DEBUG make J=1 crash_test',
'user':'root',
$PARSER
}
@@ -440,7 +323,7 @@ WRITE_STRESS_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Build and run RocksDB write stress tests',
'shell':'make write_stress && python tools/write_stress_runner.py --runtime_sec=3600 --db=/tmp/rocksdb_write_stress || $CONTRUN_NAME=write_stress $TASK_CREATION_TOOL',
'shell':'make write_stress && python tools/write_stress_runner.py --runtime_sec=3600 --db=/tmp/rocksdb_write_stress',
'user':'root',
$PARSER
}
@@ -462,7 +345,7 @@ ASAN_TEST_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Test RocksDB debug under ASAN',
'shell':'set -o pipefail && ($SHM $ASAN $DEBUG make $PARALLELISM asan_check || $CONTRUN_NAME=asan_check $TASK_CREATION_TOOL) |& /usr/facebook/ops/scripts/asan_symbolize.py -d',
'shell':'set -o pipefail && $SHM $ASAN $DEBUG make J=1 asan_check |& /usr/facebook/ops/scripts/asan_symbolize.py -d',
'user':'root',
$PARSER
}
@@ -484,49 +367,7 @@ ASAN_CRASH_TEST_COMMANDS="[
{
'name':'Build and run RocksDB debug asan_crash_test',
'timeout': 86400,
'shell':'$SHM $DEBUG make J=1 asan_crash_test || $CONTRUN_NAME=asan_crash_test $TASK_CREATION_TOOL',
'user':'root',
$PARSER
},
],
$REPORT
}
]"
#
# RocksDB test under undefined behavior sanitizer
#
UBSAN_TEST_COMMANDS="[
{
'name':'Rocksdb Unit Test under UBSAN',
'oncall':'$ONCALL',
'steps': [
$CLEANUP_ENV,
{
'name':'Test RocksDB debug under UBSAN',
'shell':'set -o pipefail && $SHM $UBSAN $DEBUG make $PARALLELISM ubsan_check || $CONTRUN_NAME=ubsan_check $TASK_CREATION_TOOL',
'user':'root',
$PARSER
}
],
$REPORT
}
]"
#
# RocksDB crash testing under udnefined behavior sanitizer
#
UBSAN_CRASH_TEST_COMMANDS="[
{
'name':'Rocksdb crash test under UBSAN',
'oncall':'$ONCALL',
'timeout': 86400,
'steps': [
$CLEANUP_ENV,
{
'name':'Build and run RocksDB debug ubsan_crash_test',
'timeout': 86400,
'shell':'$SHM $DEBUG make J=1 ubsan_crash_test || $CONTRUN_NAME=ubsan_crash_test $TASK_CREATION_TOOL',
'shell':'$SHM $DEBUG make J=1 asan_crash_test',
'user':'root',
$PARSER
},
@@ -542,13 +383,11 @@ VALGRIND_TEST_COMMANDS="[
{
'name':'Rocksdb Unit Test under valgrind',
'oncall':'$ONCALL',
'timeout': 86400,
'steps': [
$CLEANUP_ENV,
{
'name':'Run RocksDB debug unit tests',
'timeout': 86400,
'shell':'$SHM $DEBUG make $PARALLELISM valgrind_test || $CONTRUN_NAME=valgrind_check $TASK_CREATION_TOOL',
'shell':'$DISABLE_JEMALLOC $SHM $DEBUG make valgrind_check',
'user':'root',
$PARSER
},
@@ -564,13 +403,11 @@ TSAN_UNIT_TEST_COMMANDS="[
{
'name':'Rocksdb Unit Test under TSAN',
'oncall':'$ONCALL',
'timeout': 86400,
'steps': [
$CLEANUP_ENV,
{
'name':'Run RocksDB debug unit test',
'timeout': 86400,
'shell':'set -o pipefail && $SHM $DEBUG $TSAN make $PARALLELISM check || $CONTRUN_NAME=tsan_check $TASK_CREATION_TOOL',
'shell':'set -o pipefail && $SHM $DEBUG $TSAN make J=1 check',
'user':'root',
$PARSER
},
@@ -592,7 +429,7 @@ TSAN_CRASH_TEST_COMMANDS="[
{
'name':'Compile and run',
'timeout': 86400,
'shell':'set -o pipefail && $SHM $DEBUG $TSAN 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 make J=1 crash_test',
'user':'root',
$PARSER
},
@@ -601,6 +438,8 @@ TSAN_CRASH_TEST_COMMANDS="[
}
]"
TSAN_CRASH_TEST_COMMANDS=$DISABLE_COMMANDS
#
# RocksDB format compatible
#
@@ -623,9 +462,9 @@ run_format_compatible()
if [ -e "build_detect_platform" ]
then
sed "s/tcmalloc/nothingnothingnothing/g" build_detect_platform > $TEST_TMPDIR/temp_build_file
rm -rf build_detect_platform
rm -rf build_detect_platform
cp $TEST_TMPDIR/temp_build_file build_detect_platform
chmod +x build_detect_platform
chmod +x build_detect_platform
fi
make ldb -j32
@@ -654,7 +493,7 @@ FORMAT_COMPATIBLE_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Run RocksDB debug unit test',
'shell':'build_tools/rocksdb-lego-determinator run_format_compatible || $CONTRUN_NAME=run_format_compatible $TASK_CREATION_TOOL',
'shell':'build_tools/rocksdb-lego-determinator run_format_compatible',
'user':'root',
$PARSER
},
@@ -687,7 +526,7 @@ NO_COMPRESSION_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Run RocksDB debug unit test',
'shell':'build_tools/rocksdb-lego-determinator run_no_compression || $CONTRUN_NAME=run_no_compression $TASK_CREATION_TOOL',
'shell':'build_tools/rocksdb-lego-determinator run_no_compression',
'user':'root',
$PARSER
},
@@ -744,7 +583,7 @@ REGRESSION_COMMANDS="[
$CLEANUP_ENV,
{
'name':'Make and run script',
'shell':'build_tools/rocksdb-lego-determinator run_regression || $CONTRUN_NAME=run_regression $TASK_CREATION_TOOL',
'shell':'build_tools/rocksdb-lego-determinator run_regression',
'user':'root',
$PARSER
},
@@ -753,27 +592,6 @@ REGRESSION_COMMANDS="[
}
]"
#
# RocksDB Java build
#
JAVA_BUILD_TEST_COMMANDS="[
{
'name':'Rocksdb Java Build',
'oncall':'$ONCALL',
'steps': [
$CLEANUP_ENV,
{
'name':'Build RocksDB for Java',
'shell':'$SETUP_JAVA_ENV; $SHM make rocksdbjava || $CONTRUN_NAME=rocksdbjava $TASK_CREATION_TOOL',
'user':'root',
$PARSER
},
],
$REPORT
}
]"
case $1 in
punit)
echo $PARALLEL_UNIT_TEST_COMMANDS
@@ -781,24 +599,12 @@ case $1 in
unit)
echo $UNIT_TEST_COMMANDS
;;
unit_non_shm)
echo $UNIT_TEST_NON_SHM_COMMANDS
;;
release)
echo $RELEASE_BUILD_COMMANDS
;;
unit_481)
echo $UNIT_TEST_COMMANDS_481
;;
release_481)
echo $RELEASE_BUILD_COMMANDS_481
;;
clang_unit)
echo $CLANG_UNIT_TEST_COMMANDS
;;
clang_release)
echo $CLANG_RELEASE_BUILD_COMMANDS
;;
clang_analyze)
echo $CLANG_ANALYZE_COMMANDS
;;
@@ -826,12 +632,6 @@ case $1 in
asan_crash)
echo $ASAN_CRASH_TEST_COMMANDS
;;
ubsan)
echo $UBSAN_TEST_COMMANDS
;;
ubsan_crash)
echo $UBSAN_CRASH_TEST_COMMANDS
;;
valgrind)
echo $VALGRIND_TEST_COMMANDS
;;
@@ -859,9 +659,6 @@ case $1 in
run_regression)
run_regression
;;
java_build)
echo $JAVA_BUILD_TEST_COMMANDS
;;
*)
echo "Invalid determinator command"
;;
+12 -24
View File
@@ -1,5 +1,5 @@
# This script enables you running RocksDB tests by running
# All the tests in parallel and utilizing all the cores
# All the tests in paralell and utilizing all the cores
# For db_test the script first lists and parses the tests
# and then fires them up in parallel using async PS Job functionality
# Run the script from the enlistment
@@ -9,7 +9,7 @@ Param(
[string]$WorkFolder = "", # Direct tests to use that folder
[int]$Limit = -1, # -1 means run all otherwise limit for testing purposes
[string]$Exclude = "", # Expect a comma separated list, no spaces
[string]$Run = "db_test", # Run db_test|db_test2|tests|testname1,testname2...
[string]$Run = "db_test", # Run db_test|tests|testname1,testname2...
# Number of async tasks that would run concurrently. Recommend a number below 64.
# However, CPU utlization really depends on the storage media. Recommend ram based disk.
# a value of 1 will run everything serially
@@ -46,25 +46,15 @@ if($WorkFolder -eq "") {
$Env:TEST_TMPDIR = $WorkFolder
}
Write-Output "Root: $RootFolder, WorkFolder: $WorkFolder"
# Use JEMALLOC executables
if($Run -ceq "db_test" -or
$Run -ceq "db_test2" ) {
$file_name = $Run
if($EnableJE) {
$file_name += "_je"
}
$file_name += ".exe"
$db_test = -Join ($BinariesFolder, $file_name)
Write-Output "Binaries: $BinariesFolder db_test: $db_test"
if($EnableJE) {
$db_test = -Join ($BinariesFolder, "db_test_je.exe")
} else {
$db_test = -Join ($BinariesFolder, "db_test.exe")
}
Write-Output "Root: $RootFolder, WorkFolder: $WorkFolder"
Write-Output "Binaries: $BinariesFolder db_test: $db_test"
#Exclusions that we do not want to run
$ExcludeTests = New-Object System.Collections.Generic.HashSet[string]
@@ -163,8 +153,7 @@ function Discover-TestBinaries([string]$Pattern, $HashTable) {
$TestsToRun = [ordered]@{}
if($Run -ceq "db_test" -or
$Run -ceq "db_test2") {
if($Run -ceq "db_test") {
Normalize-DbTests -HashTable $TestsToRun
} elseif($Run -ceq "tests") {
if($EnableJE) {
@@ -236,8 +225,7 @@ function RunJobs($TestToLog, [int]$ConcurrencyVal, [bool]$AddForRerun)
Write-Host "Starting $k"
$log_path = ($TestToLog.$k)
if($Run -ceq "db_test" -or
$Run -ceq "db_test2") {
if($Run -ceq "db_test") {
$job = Start-Job -Name $k -ScriptBlock $InvokeTestCase -ArgumentList @($db_test,$k,$log_path)
} else {
[string]$Exe = -Join ($BinariesFolder, $k)
@@ -270,7 +258,7 @@ function RunJobs($TestToLog, [int]$ConcurrencyVal, [bool]$AddForRerun)
Write-Warning $message
$log_content | Write-Warning
} else {
# Scan the log. If we find PASSED and no occurrence of FAILED
# Scan the log. If we find PASSED and no occurence of FAILED
# then it is a success
[bool]$pass_found = $false
ForEach($l in $log_content) {
@@ -327,4 +315,4 @@ if(!$success) {
exit 12345
}
-131
View File
@@ -1,131 +0,0 @@
#!/bin/sh
#
# Update dependencies.sh file with the latest avaliable versions
BASEDIR=$(dirname $0)
OUTPUT=""
function log_variable()
{
echo "$1=${!1}" >> "$OUTPUT"
}
TP2_LATEST="/mnt/vol/engshare/fbcode/third-party2"
## $1 => lib name
## $2 => lib version (if not provided, will try to pick latest)
## $3 => platform (if not provided, will try to pick latest gcc)
##
## get_lib_base will set a variable named ${LIB_NAME}_BASE to the lib location
function get_lib_base()
{
local lib_name=$1
local lib_version=$2
local lib_platform=$3
local result="$TP2_LATEST/$lib_name/"
# Lib Version
if [ -z "$lib_version" ] || [ "$lib_version" = "LATEST" ]; then
# version is not provided, use latest
result=`ls -dr1v $result/*/ | head -n1`
else
result="$result/$lib_version/"
fi
# Lib Platform
if [ -z "$lib_platform" ]; then
# platform is not provided, use latest gcc
result=`ls -dr1v $result/gcc-*[^fb]/ | head -n1`
else
result="$result/$lib_platform/"
fi
result=`ls -1d $result/*/ | head -n1`
# lib_name => LIB_NAME_BASE
local __res_var=${lib_name^^}"_BASE"
__res_var=`echo $__res_var | tr - _`
# LIB_NAME_BASE=$result
eval $__res_var=`readlink -f $result`
log_variable $__res_var
}
###########################################################
# 4.9.x dependencies #
###########################################################
OUTPUT="$BASEDIR/dependencies.sh"
rm -f "$OUTPUT"
touch "$OUTPUT"
echo "Writing dependencies to $OUTPUT"
# Compilers locations
GCC_BASE=`readlink -f $TP2_LATEST/gcc/4.9.x/centos6-native/*/`
CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/stable/centos6-native/*/`
log_variable GCC_BASE
log_variable CLANG_BASE
# Libraries locations
get_lib_base libgcc 4.9.x
get_lib_base glibc 2.20
get_lib_base snappy LATEST gcc-4.9-glibc-2.20
get_lib_base zlib LATEST
get_lib_base bzip2 LATEST
get_lib_base lz4 LATEST
get_lib_base zstd LATEST
get_lib_base gflags LATEST
get_lib_base jemalloc LATEST
get_lib_base numa LATEST
get_lib_base libunwind LATEST
get_lib_base tbb 4.0_update2 gcc-4.9-glibc-2.20
get_lib_base kernel-headers LATEST
get_lib_base binutils LATEST centos6-native
get_lib_base valgrind 3.10.0 gcc-4.9-glibc-2.20
get_lib_base lua 5.2.3 gcc-4.9-glibc-2.20
git diff $OUTPUT
###########################################################
# 4.8.1 dependencies #
###########################################################
OUTPUT="$BASEDIR/dependencies_4.8.1.sh"
rm -f "$OUTPUT"
touch "$OUTPUT"
echo "Writing 4.8.1 dependencies to $OUTPUT"
# Compilers locations
GCC_BASE=`readlink -f $TP2_LATEST/gcc/4.8.1/centos6-native/*/`
CLANG_BASE=`readlink -f $TP2_LATEST/llvm-fb/stable/centos6-native/*/`
log_variable GCC_BASE
log_variable CLANG_BASE
# Libraries locations
get_lib_base libgcc 4.8.1 gcc-4.8.1-glibc-2.17
get_lib_base glibc 2.17 gcc-4.8.1-glibc-2.17
get_lib_base snappy LATEST gcc-4.8.1-glibc-2.17
get_lib_base zlib LATEST gcc-4.8.1-glibc-2.17
get_lib_base bzip2 LATEST gcc-4.8.1-glibc-2.17
get_lib_base lz4 LATEST gcc-4.8.1-glibc-2.17
get_lib_base zstd LATEST gcc-4.8.1-glibc-2.17
get_lib_base gflags LATEST gcc-4.8.1-glibc-2.17
get_lib_base jemalloc LATEST gcc-4.8.1-glibc-2.17
get_lib_base numa LATEST gcc-4.8.1-glibc-2.17
get_lib_base libunwind LATEST gcc-4.8.1-glibc-2.17
get_lib_base tbb 4.0_update2 gcc-4.8.1-glibc-2.17
get_lib_base kernel-headers LATEST gcc-4.8.1-glibc-2.17
get_lib_base binutils LATEST centos6-native
get_lib_base valgrind 3.8.1 gcc-4.8.1-glibc-2.17
get_lib_base lua 5.2.3 centos6-native
git diff $OUTPUT
-22
View File
@@ -1,22 +0,0 @@
# - Find JeMalloc library
# Find the native JeMalloc includes and library
# This module defines
# JEMALLOC_INCLUDE_DIRS, where to find jemalloc.h, Set when
# JEMALLOC_INCLUDE_DIR is found.
# JEMALLOC_LIBRARIES, libraries to link against to use JeMalloc.
# JEMALLOC_FOUND, If false, do not try to use JeMalloc.
#
find_path(JEMALLOC_INCLUDE_DIR
jemalloc/jemalloc.h)
find_library(JEMALLOC_LIBRARIES
jemalloc)
include(FindPackageHandleStandardArgs)
find_package_handle_standard_args(JeMalloc DEFAULT_MSG
JEMALLOC_LIBRARIES JEMALLOC_INCLUDE_DIR)
MARK_AS_ADVANCED(
JEMALLOC_INCLUDE_DIR
JEMALLOC_LIBRARIES
)
-23
View File
@@ -1,23 +0,0 @@
# - Find Snappy
# Find the snappy compression library and includes
#
# SNAPPY_INCLUDE_DIR - where to find snappy.h, etc.
# SNAPPY_LIBRARIES - List of libraries when using snappy.
# SNAPPY_FOUND - True if snappy found.
find_path(SNAPPY_INCLUDE_DIR
NAMES snappy.h
HINTS ${SNAPPY_ROOT_DIR}/include)
find_library(SNAPPY_LIBRARIES
NAMES snappy
HINTS ${SNAPPY_ROOT_DIR}/lib)
# handle the QUIETLY and REQUIRED arguments and set UUID_FOUND to TRUE if
# all listed variables are TRUE
include(FindPackageHandleStandardArgs)
find_package_handle_standard_args(snappy DEFAULT_MSG SNAPPY_LIBRARIES SNAPPY_INCLUDE_DIR)
mark_as_advanced(
SNAPPY_LIBRARIES
SNAPPY_INCLUDE_DIR)
+20 -65
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -15,7 +15,6 @@
#include "db/compaction_iterator.h"
#include "db/dbformat.h"
#include "db/event_helpers.h"
#include "db/filename.h"
#include "db/internal_stats.h"
#include "db/merge_helper.h"
@@ -42,72 +41,43 @@ TableBuilder* NewTableBuilder(
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,
WritableFileWriter* file, const CompressionType compression_type,
const CompressionOptions& compression_opts,
int level,
const std::string* compression_dict, const bool skip_filters) {
assert((column_family_id ==
TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
column_family_name.empty());
uint32_t column_family_id, WritableFileWriter* file,
const CompressionType compression_type,
const CompressionOptions& compression_opts, const bool skip_filters) {
return ioptions.table_factory->NewTableBuilder(
TableBuilderOptions(ioptions, internal_comparator,
int_tbl_prop_collector_factories, compression_type,
compression_opts, compression_dict, skip_filters,
column_family_name, level),
compression_opts, skip_filters),
column_family_id, file);
}
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::unique_ptr<InternalIterator> range_del_iter, FileMetaData* meta,
const EnvOptions& env_options, TableCache* table_cache,
InternalIterator* 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,
std::vector<SequenceNumber> snapshots,
SequenceNumber earliest_write_conflict_snapshot,
uint32_t column_family_id, std::vector<SequenceNumber> snapshots,
const CompressionType compression,
const CompressionOptions& compression_opts, bool paranoid_file_checks,
InternalStats* internal_stats, TableFileCreationReason reason,
EventLogger* event_logger, int job_id, const Env::IOPriority io_priority,
TableProperties* table_properties, int level) {
assert((column_family_id ==
TablePropertiesCollectorFactory::Context::kUnknownColumnFamily) ==
column_family_name.empty());
InternalStats* internal_stats, const Env::IOPriority io_priority,
TableProperties* table_properties) {
// Reports the IOStats for flush for every following bytes.
const size_t kReportFlushIOStatsEvery = 1048576;
Status s;
meta->fd.file_size = 0;
iter->SeekToFirst();
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.db_paths, meta->fd.GetNumber(),
meta->fd.GetPathId());
#ifndef ROCKSDB_LITE
EventHelpers::NotifyTableFileCreationStarted(
ioptions.listeners, dbname, column_family_name, fname, job_id, reason);
#endif // !ROCKSDB_LITE
TableProperties tp;
if (iter->Valid() || range_del_agg->ShouldAddTombstones()) {
if (iter->Valid()) {
TableBuilder* builder;
unique_ptr<WritableFileWriter> file_writer;
{
unique_ptr<WritableFile> file;
s = NewWritableFile(env, fname, &file, env_options);
if (!s.ok()) {
EventHelpers::LogAndNotifyTableFileCreationFinished(
event_logger, ioptions.listeners, dbname, column_family_name, fname,
job_id, meta->fd, tp, reason, s);
return s;
}
file->SetIOPriority(io_priority);
@@ -116,20 +86,18 @@ Status BuildTable(
builder = NewTableBuilder(
ioptions, internal_comparator, int_tbl_prop_collector_factories,
column_family_id, column_family_name, file_writer.get(), compression,
compression_opts, level);
column_family_id, file_writer.get(), compression, compression_opts);
}
MergeHelper merge(env, internal_comparator.user_comparator(),
ioptions.merge_operator, nullptr, ioptions.info_log,
mutable_cf_options.min_partial_merge_operands,
ioptions.min_partial_merge_operands,
true /* internal key corruption is not ok */,
snapshots.empty() ? 0 : snapshots.back());
CompactionIterator c_iter(
iter, internal_comparator.user_comparator(), &merge, kMaxSequenceNumber,
&snapshots, earliest_write_conflict_snapshot, env,
true /* internal key corruption is not ok */, range_del_agg.get());
CompactionIterator c_iter(iter, internal_comparator.user_comparator(),
&merge, kMaxSequenceNumber, &snapshots, env,
true /* internal key corruption is not ok */);
c_iter.SeekToFirst();
for (; c_iter.Valid(); c_iter.Next()) {
const Slice& key = c_iter.key();
@@ -144,9 +112,6 @@ Status BuildTable(
ThreadStatus::FLUSH_BYTES_WRITTEN, IOSTATS(bytes_written));
}
}
// nullptr for table_{min,max} so all range tombstones will be flushed
range_del_agg->AddToBuilder(builder, nullptr /* lower_bound */,
nullptr /* upper_bound */, meta);
// Finish and check for builder errors
bool empty = builder->NumEntries() == 0;
@@ -158,13 +123,11 @@ Status BuildTable(
}
if (s.ok() && !empty) {
uint64_t file_size = builder->FileSize();
meta->fd.file_size = file_size;
meta->fd.file_size = builder->FileSize();
meta->marked_for_compaction = builder->NeedCompact();
assert(meta->fd.GetFileSize() > 0);
tp = builder->GetTableProperties();
if (table_properties) {
*table_properties = tp;
*table_properties = builder->GetTableProperties();
}
}
delete builder;
@@ -181,12 +144,10 @@ Status BuildTable(
if (s.ok() && !empty) {
// Verify that the table is usable
std::unique_ptr<InternalIterator> it(table_cache->NewIterator(
ReadOptions(), env_options, internal_comparator, meta->fd,
nullptr /* range_del_agg */, nullptr,
ReadOptions(), env_options, internal_comparator, meta->fd, nullptr,
(internal_stats == nullptr) ? nullptr
: internal_stats->GetFileReadHist(0),
false /* for_compaction */, nullptr /* arena */,
false /* skip_filter */, level));
false));
s = it->status();
if (s.ok() && paranoid_file_checks) {
for (it->SeekToFirst(); it->Valid(); it->Next()) {
@@ -204,12 +165,6 @@ Status BuildTable(
if (!s.ok() || meta->fd.GetFileSize() == 0) {
env->DeleteFile(fname);
}
// Output to event logger and fire events.
EventHelpers::LogAndNotifyTableFileCreationFinished(
event_logger, ioptions.listeners, dbname, column_family_name, fname,
job_id, meta->fd, tp, reason, s);
return s;
}
+11 -28
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -12,14 +12,11 @@
#include "db/table_properties_collector.h"
#include "rocksdb/comparator.h"
#include "rocksdb/env.h"
#include "rocksdb/listener.h"
#include "rocksdb/options.h"
#include "rocksdb/status.h"
#include "rocksdb/table_properties.h"
#include "rocksdb/types.h"
#include "table/scoped_arena_iterator.h"
#include "util/cf_options.h"
#include "util/event_logger.h"
#include "rocksdb/options.h"
#include "rocksdb/immutable_options.h"
#include "rocksdb/table_properties.h"
namespace rocksdb {
@@ -36,21 +33,14 @@ 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
// TableBuilder returned by this function.
// @param compression_dict Data for presetting the compression library's
// dictionary, or nullptr.
TableBuilder* NewTableBuilder(
const ImmutableCFOptions& options,
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,
WritableFileWriter* file, const CompressionType compression_type,
uint32_t column_family_id, WritableFileWriter* file,
const CompressionType compression_type,
const CompressionOptions& compression_opts,
int level,
const std::string* compression_dict = nullptr,
const bool skip_filters = false);
// Build a Table file from the contents of *iter. The generated file
@@ -58,25 +48,18 @@ TableBuilder* NewTableBuilder(
// *meta will be filled with metadata about the generated table.
// If no data is present in *iter, meta->file_size will be set to
// zero, and no Table file will be produced.
//
// @param column_family_name Name of the column family that is also identified
// by column_family_id, or empty string if unknown.
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::unique_ptr<InternalIterator> range_del_iter, FileMetaData* meta,
const EnvOptions& env_options, TableCache* table_cache,
InternalIterator* 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,
std::vector<SequenceNumber> snapshots,
SequenceNumber earliest_write_conflict_snapshot,
uint32_t column_family_id, std::vector<SequenceNumber> snapshots,
const CompressionType compression,
const CompressionOptions& compression_opts, bool paranoid_file_checks,
InternalStats* internal_stats, TableFileCreationReason reason,
EventLogger* event_logger = nullptr, int job_id = 0,
InternalStats* internal_stats,
const Env::IOPriority io_priority = Env::IO_HIGH,
TableProperties* table_properties = nullptr, int level = -1);
TableProperties* table_properties = nullptr);
} // namespace rocksdb
+68 -307
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -30,7 +30,6 @@
#include "rocksdb/statistics.h"
#include "rocksdb/slice_transform.h"
#include "rocksdb/table.h"
#include "rocksdb/rate_limiter.h"
#include "rocksdb/utilities/backupable_db.h"
#include "utilities/merge_operators.h"
@@ -44,16 +43,13 @@ using rocksdb::CompactionFilterContext;
using rocksdb::CompactionOptionsFIFO;
using rocksdb::Comparator;
using rocksdb::CompressionType;
using rocksdb::WALRecoveryMode;
using rocksdb::DB;
using rocksdb::DBOptions;
using rocksdb::Env;
using rocksdb::EnvOptions;
using rocksdb::InfoLogLevel;
using rocksdb::FileLock;
using rocksdb::FilterPolicy;
using rocksdb::FlushOptions;
using rocksdb::IngestExternalFileOptions;
using rocksdb::Iterator;
using rocksdb::Logger;
using rocksdb::MergeOperator;
@@ -71,7 +67,6 @@ using rocksdb::Slice;
using rocksdb::SliceParts;
using rocksdb::SliceTransform;
using rocksdb::Snapshot;
using rocksdb::SstFileWriter;
using rocksdb::Status;
using rocksdb::WritableFile;
using rocksdb::WriteBatch;
@@ -82,8 +77,6 @@ using rocksdb::BackupableDBOptions;
using rocksdb::BackupInfo;
using rocksdb::RestoreOptions;
using rocksdb::CompactRangeOptions;
using rocksdb::RateLimiter;
using rocksdb::NewGenericRateLimiter;
using std::shared_ptr;
@@ -114,10 +107,6 @@ 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 { RateLimiter* rep; };
struct rocksdb_compactionfiltercontext_t {
CompactionFilter::Context rep;
@@ -134,7 +123,6 @@ struct rocksdb_compactionfilter_t : public CompactionFilter {
char** new_value, size_t *new_value_length,
unsigned char* value_changed);
const char* (*name_)(void*);
unsigned char ignore_snapshots_;
virtual ~rocksdb_compactionfilter_t() {
(*destructor_)(state_);
@@ -160,8 +148,6 @@ struct rocksdb_compactionfilter_t : public CompactionFilter {
}
virtual const char* Name() const override { return (*name_)(state_); }
virtual bool IgnoreSnapshots() const override { return ignore_snapshots_; }
};
struct rocksdb_compactionfilterfactory_t : public CompactionFilterFactory {
@@ -283,31 +269,33 @@ struct rocksdb_mergeoperator_t : public MergeOperator {
virtual const char* Name() const override { return (*name_)(state_); }
virtual bool FullMergeV2(const MergeOperationInput& merge_in,
MergeOperationOutput* merge_out) const override {
size_t n = merge_in.operand_list.size();
virtual bool FullMerge(const Slice& key, const Slice* existing_value,
const std::deque<std::string>& operand_list,
std::string* new_value,
Logger* logger) const override {
size_t n = operand_list.size();
std::vector<const char*> operand_pointers(n);
std::vector<size_t> operand_sizes(n);
for (size_t i = 0; i < n; i++) {
Slice operand(merge_in.operand_list[i]);
Slice operand(operand_list[i]);
operand_pointers[i] = operand.data();
operand_sizes[i] = operand.size();
}
const char* existing_value_data = nullptr;
size_t existing_value_len = 0;
if (merge_in.existing_value != nullptr) {
existing_value_data = merge_in.existing_value->data();
existing_value_len = merge_in.existing_value->size();
if (existing_value != nullptr) {
existing_value_data = existing_value->data();
existing_value_len = existing_value->size();
}
unsigned char success;
size_t new_value_len;
char* tmp_new_value = (*full_merge_)(
state_, merge_in.key.data(), merge_in.key.size(), existing_value_data,
existing_value_len, &operand_pointers[0], &operand_sizes[0],
static_cast<int>(n), &success, &new_value_len);
merge_out->new_value.assign(tmp_new_value, new_value_len);
state_, key.data(), key.size(), existing_value_data, existing_value_len,
&operand_pointers[0], &operand_sizes[0], static_cast<int>(n), &success,
&new_value_len);
new_value->assign(tmp_new_value, new_value_len);
if (delete_value_ != nullptr) {
(*delete_value_)(state_, tmp_new_value, new_value_len);
@@ -458,12 +446,6 @@ void rocksdb_backup_engine_create_new_backup(rocksdb_backup_engine_t* be,
SaveError(errptr, be->rep->CreateNewBackup(db->rep));
}
void rocksdb_backup_engine_purge_old_backups(rocksdb_backup_engine_t* be,
uint32_t num_backups_to_keep,
char** errptr) {
SaveError(errptr, be->rep->PurgeOldBackups(num_backups_to_keep));
}
rocksdb_restore_options_t* rocksdb_restore_options_create() {
return new rocksdb_restore_options_t;
}
@@ -843,31 +825,6 @@ rocksdb_iterator_t* rocksdb_create_iterator_cf(
return result;
}
void rocksdb_create_iterators(
rocksdb_t *db,
rocksdb_readoptions_t* opts,
rocksdb_column_family_handle_t** column_families,
rocksdb_iterator_t** iterators,
size_t size,
char** errptr) {
std::vector<ColumnFamilyHandle*> column_families_vec;
for (size_t i = 0; i < size; i++) {
column_families_vec.push_back(column_families[i]->rep);
}
std::vector<Iterator*> res;
Status status = db->rep->NewIterators(opts->rep, column_families_vec, &res);
assert(res.size() == size);
if (SaveError(errptr, status)) {
return;
}
for (size_t i = 0; i < size; i++) {
iterators[i] = new rocksdb_iterator_t;
iterators[i]->rep = res[i];
}
}
const rocksdb_snapshot_t* rocksdb_create_snapshot(
rocksdb_t* db) {
rocksdb_snapshot_t* result = new rocksdb_snapshot_t;
@@ -1031,11 +988,6 @@ void rocksdb_iter_seek(rocksdb_iterator_t* iter, const char* k, size_t klen) {
iter->rep->Seek(Slice(k, klen));
}
void rocksdb_iter_seek_for_prev(rocksdb_iterator_t* iter, const char* k,
size_t klen) {
iter->rep->SeekForPrev(Slice(k, klen));
}
void rocksdb_iter_next(rocksdb_iterator_t* iter) {
iter->rep->Next();
}
@@ -1336,11 +1288,6 @@ void rocksdb_block_based_options_set_cache_index_and_filter_blocks(
options->rep.cache_index_and_filter_blocks = v;
}
void rocksdb_block_based_options_set_pin_l0_filter_and_index_blocks_in_cache(
rocksdb_block_based_table_options_t* options, unsigned char v) {
options->rep.pin_l0_filter_and_index_blocks_in_cache = v;
}
void rocksdb_block_based_options_set_skip_table_builder_flush(
rocksdb_block_based_table_options_t* options, unsigned char v) {
options->rep.skip_table_builder_flush = v;
@@ -1441,11 +1388,6 @@ void rocksdb_options_set_compaction_filter_factory(
std::shared_ptr<CompactionFilterFactory>(factory);
}
void rocksdb_options_compaction_readahead_size(
rocksdb_options_t* opt, size_t s) {
opt->rep.compaction_readahead_size = s;
}
void rocksdb_options_set_comparator(
rocksdb_options_t* opt,
rocksdb_comparator_t* cmp) {
@@ -1526,19 +1468,19 @@ void rocksdb_options_set_max_bytes_for_level_base(
opt->rep.max_bytes_for_level_base = n;
}
void rocksdb_options_set_level_compaction_dynamic_level_bytes(
rocksdb_options_t* opt, unsigned char v) {
opt->rep.level_compaction_dynamic_level_bytes = v;
}
void rocksdb_options_set_max_bytes_for_level_multiplier(rocksdb_options_t* opt,
double n) {
void rocksdb_options_set_max_bytes_for_level_multiplier(
rocksdb_options_t* opt, int n) {
opt->rep.max_bytes_for_level_multiplier = n;
}
void rocksdb_options_set_max_compaction_bytes(rocksdb_options_t* opt,
uint64_t n) {
opt->rep.max_compaction_bytes = n;
void rocksdb_options_set_expanded_compaction_factor(
rocksdb_options_t* opt, int n) {
opt->rep.expanded_compaction_factor = n;
}
void rocksdb_options_set_max_grandparent_overlap_factor(
rocksdb_options_t* opt, int n) {
opt->rep.max_grandparent_overlap_factor = n;
}
void rocksdb_options_set_max_bytes_for_level_multiplier_additional(
@@ -1575,10 +1517,6 @@ void rocksdb_options_set_level0_stop_writes_trigger(
void rocksdb_options_set_max_mem_compaction_level(rocksdb_options_t* opt,
int n) {}
void rocksdb_options_set_wal_recovery_mode(rocksdb_options_t* opt,int mode) {
opt->rep.wal_recovery_mode = static_cast<WALRecoveryMode>(mode);
}
void rocksdb_options_set_compression(rocksdb_options_t* opt, int t) {
opt->rep.compression = static_cast<CompressionType>(t);
}
@@ -1593,13 +1531,11 @@ void rocksdb_options_set_compression_per_level(rocksdb_options_t* opt,
}
}
void rocksdb_options_set_compression_options(rocksdb_options_t* opt, int w_bits,
int level, int strategy,
int max_dict_bytes) {
void rocksdb_options_set_compression_options(
rocksdb_options_t* opt, int w_bits, int level, int strategy) {
opt->rep.compression_opts.window_bits = w_bits;
opt->rep.compression_opts.level = level;
opt->rep.compression_opts.strategy = strategy;
opt->rep.compression_opts.max_dict_bytes = max_dict_bytes;
}
void rocksdb_options_set_prefix_extractor(
@@ -1708,21 +1644,16 @@ void rocksdb_options_set_bytes_per_sync(
opt->rep.bytes_per_sync = v;
}
void rocksdb_options_set_allow_concurrent_memtable_write(rocksdb_options_t* opt,
unsigned char v) {
opt->rep.allow_concurrent_memtable_write = v;
}
void rocksdb_options_set_enable_write_thread_adaptive_yield(
rocksdb_options_t* opt, unsigned char v) {
opt->rep.enable_write_thread_adaptive_yield = v;
}
void rocksdb_options_set_verify_checksums_in_compaction(
rocksdb_options_t* opt, unsigned char v) {
opt->rep.verify_checksums_in_compaction = v;
}
void rocksdb_options_set_filter_deletes(
rocksdb_options_t* opt, unsigned char v) {
opt->rep.filter_deletes = v;
}
void rocksdb_options_set_max_sequential_skip_in_iterations(
rocksdb_options_t* opt, uint64_t v) {
opt->rep.max_sequential_skip_in_iterations = v;
@@ -1745,11 +1676,6 @@ void rocksdb_options_set_max_background_compactions(rocksdb_options_t* opt, int
opt->rep.max_background_compactions = n;
}
void rocksdb_options_set_base_background_compactions(rocksdb_options_t* opt,
int n) {
opt->rep.base_background_compactions = n;
}
void rocksdb_options_set_max_background_flushes(rocksdb_options_t* opt, int n) {
opt->rep.max_background_flushes = n;
}
@@ -1813,47 +1739,66 @@ void rocksdb_options_set_delete_obsolete_files_period_micros(
opt->rep.delete_obsolete_files_period_micros = v;
}
void rocksdb_options_set_source_compaction_factor(
rocksdb_options_t* opt, int n) {
opt->rep.expanded_compaction_factor = n;
}
void rocksdb_options_prepare_for_bulk_load(rocksdb_options_t* opt) {
opt->rep.PrepareForBulkLoad();
}
void rocksdb_options_set_memtable_vector_rep(rocksdb_options_t *opt) {
opt->rep.memtable_factory.reset(new rocksdb::VectorRepFactory);
static rocksdb::VectorRepFactory* factory = 0;
if (!factory) {
factory = new rocksdb::VectorRepFactory;
}
opt->rep.memtable_factory.reset(factory);
}
void rocksdb_options_set_memtable_prefix_bloom_size_ratio(
rocksdb_options_t* opt, double v) {
opt->rep.memtable_prefix_bloom_size_ratio = v;
void rocksdb_options_set_memtable_prefix_bloom_bits(
rocksdb_options_t* opt, uint32_t v) {
opt->rep.memtable_prefix_bloom_bits = v;
}
void rocksdb_options_set_memtable_huge_page_size(rocksdb_options_t* opt,
size_t v) {
opt->rep.memtable_huge_page_size = v;
void rocksdb_options_set_memtable_prefix_bloom_probes(
rocksdb_options_t* opt, uint32_t v) {
opt->rep.memtable_prefix_bloom_probes = v;
}
void rocksdb_options_set_hash_skip_list_rep(
rocksdb_options_t *opt, size_t bucket_count,
int32_t skiplist_height, int32_t skiplist_branching_factor) {
rocksdb::MemTableRepFactory* factory = rocksdb::NewHashSkipListRepFactory(
bucket_count, skiplist_height, skiplist_branching_factor);
static rocksdb::MemTableRepFactory* factory = 0;
if (!factory) {
factory = rocksdb::NewHashSkipListRepFactory(
bucket_count, skiplist_height, skiplist_branching_factor);
}
opt->rep.memtable_factory.reset(factory);
}
void rocksdb_options_set_hash_link_list_rep(
rocksdb_options_t *opt, size_t bucket_count) {
opt->rep.memtable_factory.reset(rocksdb::NewHashLinkListRepFactory(bucket_count));
static rocksdb::MemTableRepFactory* factory = 0;
if (!factory) {
factory = rocksdb::NewHashLinkListRepFactory(bucket_count);
}
opt->rep.memtable_factory.reset(factory);
}
void rocksdb_options_set_plain_table_factory(
rocksdb_options_t *opt, uint32_t user_key_len, int bloom_bits_per_key,
double hash_table_ratio, size_t index_sparseness) {
rocksdb::PlainTableOptions options;
options.user_key_len = user_key_len;
options.bloom_bits_per_key = bloom_bits_per_key;
options.hash_table_ratio = hash_table_ratio;
options.index_sparseness = index_sparseness;
static rocksdb::TableFactory* factory = 0;
if (!factory) {
rocksdb::PlainTableOptions options;
options.user_key_len = user_key_len;
options.bloom_bits_per_key = bloom_bits_per_key;
options.hash_table_ratio = hash_table_ratio;
options.index_sparseness = index_sparseness;
rocksdb::TableFactory* factory = rocksdb::NewPlainTableFactory(options);
factory = rocksdb::NewPlainTableFactory(options);
}
opt->rep.table_factory.reset(factory);
}
@@ -1882,11 +1827,6 @@ void rocksdb_options_set_inplace_update_num_locks(
opt->rep.inplace_update_num_locks = v;
}
void rocksdb_options_set_report_bg_io_stats(
rocksdb_options_t* opt, int v) {
opt->rep.report_bg_io_stats = v;
}
void rocksdb_options_set_compaction_style(rocksdb_options_t *opt, int style) {
opt->rep.compaction_style = static_cast<rocksdb::CompactionStyle>(style);
}
@@ -1909,28 +1849,6 @@ char *rocksdb_options_statistics_get_string(rocksdb_options_t *opt) {
return nullptr;
}
void rocksdb_options_set_ratelimiter(rocksdb_options_t *opt, rocksdb_ratelimiter_t *limiter) {
opt->rep.rate_limiter.reset(limiter->rep);
limiter->rep = nullptr;
}
rocksdb_ratelimiter_t* rocksdb_ratelimiter_create(
int64_t rate_bytes_per_sec,
int64_t refill_period_us,
int32_t fairness) {
rocksdb_ratelimiter_t* rate_limiter = new rocksdb_ratelimiter_t;
rate_limiter->rep = NewGenericRateLimiter(rate_bytes_per_sec,
refill_period_us, fairness);
return rate_limiter;
}
void rocksdb_ratelimiter_destroy(rocksdb_ratelimiter_t *limiter) {
if (limiter->rep) {
delete limiter->rep;
}
delete limiter;
}
/*
TODO:
DB::OpenForReadOnly
@@ -1960,17 +1878,10 @@ rocksdb_compactionfilter_t* rocksdb_compactionfilter_create(
result->state_ = state;
result->destructor_ = destructor;
result->filter_ = filter;
result->ignore_snapshots_ = false;
result->name_ = name;
return result;
}
void rocksdb_compactionfilter_set_ignore_snapshots(
rocksdb_compactionfilter_t* filter,
unsigned char whether_ignore) {
filter->ignore_snapshots_ = whether_ignore;
}
void rocksdb_compactionfilter_destroy(rocksdb_compactionfilter_t* filter) {
delete filter;
}
@@ -2054,7 +1965,7 @@ void rocksdb_filterpolicy_destroy(rocksdb_filterpolicy_t* filter) {
delete filter;
}
rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_format(int bits_per_key, bool original_format) {
rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom(int bits_per_key) {
// Make a rocksdb_filterpolicy_t, but override all of its methods so
// they delegate to a NewBloomFilterPolicy() instead of user
// supplied C functions.
@@ -2072,21 +1983,13 @@ rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_format(int bits_per_ke
static void DoNothing(void*) { }
};
Wrapper* wrapper = new Wrapper;
wrapper->rep_ = NewBloomFilterPolicy(bits_per_key, original_format);
wrapper->rep_ = NewBloomFilterPolicy(bits_per_key);
wrapper->state_ = nullptr;
wrapper->delete_filter_ = nullptr;
wrapper->destructor_ = &Wrapper::DoNothing;
return wrapper;
}
rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom_full(int bits_per_key) {
return rocksdb_filterpolicy_create_bloom_format(bits_per_key, false);
}
rocksdb_filterpolicy_t* rocksdb_filterpolicy_create_bloom(int bits_per_key) {
return rocksdb_filterpolicy_create_bloom_format(bits_per_key, true);
}
rocksdb_mergeoperator_t* rocksdb_mergeoperator_create(
void* state, void (*destructor)(void*),
char* (*full_merge)(void*, const char* key, size_t key_length,
@@ -2163,11 +2066,6 @@ void rocksdb_readoptions_set_tailing(
opt->rep.tailing = v;
}
void rocksdb_readoptions_set_readahead_size(
rocksdb_readoptions_t* opt, size_t v) {
opt->rep.readahead_size = v;
}
rocksdb_writeoptions_t* rocksdb_writeoptions_create() {
return new rocksdb_writeoptions_t;
}
@@ -2209,10 +2107,6 @@ void rocksdb_cache_destroy(rocksdb_cache_t* cache) {
delete cache;
}
void rocksdb_cache_set_capacity(rocksdb_cache_t* cache, size_t capacity) {
cache->rep->SetCapacity(capacity);
}
rocksdb_env_t* rocksdb_create_default_env() {
rocksdb_env_t* result = new rocksdb_env_t;
result->rep = Env::Default();
@@ -2220,13 +2114,6 @@ rocksdb_env_t* rocksdb_create_default_env() {
return result;
}
rocksdb_env_t* rocksdb_create_mem_env() {
rocksdb_env_t* result = new rocksdb_env_t;
result->rep = rocksdb::NewMemEnv(Env::Default());
result->is_default = false;
return result;
}
void rocksdb_env_set_background_threads(rocksdb_env_t* env, int n) {
env->rep->SetBackgroundThreads(n);
}
@@ -2244,107 +2131,6 @@ void rocksdb_env_destroy(rocksdb_env_t* env) {
delete env;
}
rocksdb_envoptions_t* rocksdb_envoptions_create() {
rocksdb_envoptions_t* opt = new rocksdb_envoptions_t;
return opt;
}
void rocksdb_envoptions_destroy(rocksdb_envoptions_t* opt) { delete opt; }
rocksdb_sstfilewriter_t* rocksdb_sstfilewriter_create(
const rocksdb_envoptions_t* env, const rocksdb_options_t* io_options) {
rocksdb_sstfilewriter_t* writer = new rocksdb_sstfilewriter_t;
writer->rep =
new SstFileWriter(env->rep, io_options->rep, io_options->rep.comparator);
return writer;
}
rocksdb_sstfilewriter_t* rocksdb_sstfilewriter_create_with_comparator(
const rocksdb_envoptions_t* env, const rocksdb_options_t* io_options,
const rocksdb_comparator_t* comparator) {
rocksdb_sstfilewriter_t* writer = new rocksdb_sstfilewriter_t;
writer->rep =
new SstFileWriter(env->rep, io_options->rep, io_options->rep.comparator);
return writer;
}
void rocksdb_sstfilewriter_open(rocksdb_sstfilewriter_t* writer,
const char* name, char** errptr) {
SaveError(errptr, writer->rep->Open(std::string(name)));
}
void rocksdb_sstfilewriter_add(rocksdb_sstfilewriter_t* writer, const char* key,
size_t keylen, const char* val, size_t vallen,
char** errptr) {
SaveError(errptr, writer->rep->Add(Slice(key, keylen), Slice(val, vallen)));
}
void rocksdb_sstfilewriter_finish(rocksdb_sstfilewriter_t* writer,
char** errptr) {
SaveError(errptr, writer->rep->Finish(NULL));
}
void rocksdb_sstfilewriter_destroy(rocksdb_sstfilewriter_t* writer) {
delete writer->rep;
delete writer;
}
rocksdb_ingestexternalfileoptions_t*
rocksdb_ingestexternalfileoptions_create() {
rocksdb_ingestexternalfileoptions_t* opt =
new rocksdb_ingestexternalfileoptions_t;
return opt;
}
void rocksdb_ingestexternalfileoptions_set_move_files(
rocksdb_ingestexternalfileoptions_t* opt, unsigned char move_files) {
opt->rep.move_files = move_files;
}
void rocksdb_ingestexternalfileoptions_set_snapshot_consistency(
rocksdb_ingestexternalfileoptions_t* opt,
unsigned char snapshot_consistency) {
opt->rep.snapshot_consistency = snapshot_consistency;
}
void rocksdb_ingestexternalfileoptions_set_allow_global_seqno(
rocksdb_ingestexternalfileoptions_t* opt,
unsigned char allow_global_seqno) {
opt->rep.allow_global_seqno = allow_global_seqno;
}
void rocksdb_ingestexternalfileoptions_set_allow_blocking_flush(
rocksdb_ingestexternalfileoptions_t* opt,
unsigned char allow_blocking_flush) {
opt->rep.allow_blocking_flush = allow_blocking_flush;
}
void rocksdb_ingestexternalfileoptions_destroy(
rocksdb_ingestexternalfileoptions_t* opt) {
delete opt;
}
void rocksdb_ingest_external_file(
rocksdb_t* db, const char* const* file_list, const size_t list_len,
const rocksdb_ingestexternalfileoptions_t* opt, char** errptr) {
std::vector<std::string> files(list_len);
for (size_t i = 0; i < list_len; ++i) {
files[i] = std::string(file_list[i]);
}
SaveError(errptr, db->rep->IngestExternalFile(files, opt->rep));
}
void rocksdb_ingest_external_file_cf(
rocksdb_t* db, rocksdb_column_family_handle_t* handle,
const char* const* file_list, const size_t list_len,
const rocksdb_ingestexternalfileoptions_t* opt, char** errptr) {
std::vector<std::string> files(list_len);
for (size_t i = 0; i < list_len; ++i) {
files[i] = std::string(file_list[i]);
}
SaveError(errptr, db->rep->IngestExternalFile(handle->rep, files, opt->rep));
}
rocksdb_slicetransform_t* rocksdb_slicetransform_create(
void* state,
void (*destructor)(void*),
@@ -2539,31 +2325,6 @@ void rocksdb_get_options_from_string(const rocksdb_options_t* base_options,
&new_options->rep));
}
void rocksdb_delete_file_in_range(rocksdb_t* db, const char* start_key,
size_t start_key_len, const char* limit_key,
size_t limit_key_len, char** errptr) {
Slice a, b;
SaveError(
errptr,
DeleteFilesInRange(
db->rep, db->rep->DefaultColumnFamily(),
(start_key ? (a = Slice(start_key, start_key_len), &a) : nullptr),
(limit_key ? (b = Slice(limit_key, limit_key_len), &b) : nullptr)));
}
void rocksdb_delete_file_in_range_cf(
rocksdb_t* db, rocksdb_column_family_handle_t* column_family,
const char* start_key, size_t start_key_len, const char* limit_key,
size_t limit_key_len, char** errptr) {
Slice a, b;
SaveError(
errptr,
DeleteFilesInRange(
db->rep, column_family->rep,
(start_key ? (a = Slice(start_key, start_key_len), &a) : nullptr),
(limit_key ? (b = Slice(limit_key, limit_key_len), &b) : nullptr)));
}
void rocksdb_free(void* ptr) { free(ptr); }
} // end extern "C"
+8 -140
View File
@@ -2,18 +2,17 @@
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 <stdio.h>
#ifndef ROCKSDB_LITE // Lite does not support C API
#include "rocksdb/c.h"
#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/types.h>
#ifndef OS_WIN
#include <unistd.h>
# include <unistd.h>
#endif
#include <inttypes.h>
@@ -22,10 +21,11 @@
#include <Windows.h>
#define snprintf _snprintf
# define snprintf _snprintf
// Ok for uniqueness
int geteuid() {
int result = 0;
result = ((int)GetCurrentProcessId() << 16);
@@ -38,13 +38,13 @@ int geteuid() {
const char* phase = "";
static char dbname[200];
static char sstfilename[200];
static char dbbackupname[200];
static void StartPhase(const char* name) {
fprintf(stderr, "=== Test %s\n", name);
phase = name;
}
static const char* GetTempDir(void) {
const char* ret = getenv("TEST_TMPDIR");
if (ret == NULL || ret[0] == '\0')
@@ -291,7 +291,6 @@ int main(int argc, char** argv) {
rocksdb_block_based_table_options_t* table_options;
rocksdb_readoptions_t* roptions;
rocksdb_writeoptions_t* woptions;
rocksdb_ratelimiter_t* rate_limiter;
char* err = NULL;
int run = -1;
@@ -305,11 +304,6 @@ int main(int argc, char** argv) {
GetTempDir(),
((int) geteuid()));
snprintf(sstfilename, sizeof(sstfilename),
"%s/rocksdb_c_test-%d-sst",
GetTempDir(),
((int)geteuid()));
StartPhase("create_objects");
cmp = rocksdb_comparator_create(NULL, CmpDestroy, CmpCompare, CmpName);
env = rocksdb_create_default_env();
@@ -323,19 +317,15 @@ int main(int argc, char** argv) {
rocksdb_options_set_write_buffer_size(options, 100000);
rocksdb_options_set_paranoid_checks(options, 1);
rocksdb_options_set_max_open_files(options, 10);
rocksdb_options_set_base_background_compactions(options, 1);
table_options = rocksdb_block_based_options_create();
rocksdb_block_based_options_set_block_cache(table_options, cache);
rocksdb_options_set_block_based_table_factory(options, table_options);
rocksdb_options_set_compression(options, rocksdb_no_compression);
rocksdb_options_set_compression_options(options, -14, -1, 0, 0);
rocksdb_options_set_compression_options(options, -14, -1, 0);
int compression_levels[] = {rocksdb_no_compression, rocksdb_no_compression,
rocksdb_no_compression, rocksdb_no_compression};
rocksdb_options_set_compression_per_level(options, compression_levels, 4);
rate_limiter = rocksdb_ratelimiter_create(1000 * 1024 * 1024, 100 * 1000, 10);
rocksdb_options_set_ratelimiter(options, rate_limiter);
rocksdb_ratelimiter_destroy(rate_limiter);
roptions = rocksdb_readoptions_create();
rocksdb_readoptions_set_verify_checksums(roptions, 1);
@@ -375,24 +365,6 @@ int main(int argc, char** argv) {
rocksdb_backup_engine_create_new_backup(be, db, &err);
CheckNoError(err);
// need a change to trigger a new backup
rocksdb_delete(db, woptions, "does-not-exist", 14, &err);
CheckNoError(err);
rocksdb_backup_engine_create_new_backup(be, db, &err);
CheckNoError(err);
const rocksdb_backup_engine_info_t* bei = rocksdb_backup_engine_get_backup_info(be);
CheckCondition(rocksdb_backup_engine_info_count(bei) > 1);
rocksdb_backup_engine_info_destroy(bei);
rocksdb_backup_engine_purge_old_backups(be, 1, &err);
CheckNoError(err);
bei = rocksdb_backup_engine_get_backup_info(be);
CheckCondition(rocksdb_backup_engine_info_count(bei) == 1);
rocksdb_backup_engine_info_destroy(bei);
rocksdb_delete(db, woptions, "foo", 3, &err);
CheckNoError(err);
@@ -497,10 +469,6 @@ int main(int argc, char** argv) {
CheckIter(iter, "foo", "hello");
rocksdb_iter_seek(iter, "b", 1);
CheckIter(iter, "box", "c");
rocksdb_iter_seek_for_prev(iter, "g", 1);
CheckIter(iter, "foo", "hello");
rocksdb_iter_seek_for_prev(iter, "box", 3);
CheckIter(iter, "box", "c");
rocksdb_iter_get_error(iter, &err);
CheckNoError(err);
rocksdb_iter_destroy(iter);
@@ -579,59 +547,6 @@ int main(int argc, char** argv) {
rocksdb_release_snapshot(db, snap);
}
StartPhase("addfile");
{
rocksdb_envoptions_t* env_opt = rocksdb_envoptions_create();
rocksdb_options_t* io_options = rocksdb_options_create();
rocksdb_sstfilewriter_t* writer =
rocksdb_sstfilewriter_create(env_opt, io_options);
unlink(sstfilename);
rocksdb_sstfilewriter_open(writer, sstfilename, &err);
CheckNoError(err);
rocksdb_sstfilewriter_add(writer, "sstk1", 5, "v1", 2, &err);
CheckNoError(err);
rocksdb_sstfilewriter_add(writer, "sstk2", 5, "v2", 2, &err);
CheckNoError(err);
rocksdb_sstfilewriter_add(writer, "sstk3", 5, "v3", 2, &err);
CheckNoError(err);
rocksdb_sstfilewriter_finish(writer, &err);
CheckNoError(err);
rocksdb_ingestexternalfileoptions_t* ing_opt =
rocksdb_ingestexternalfileoptions_create();
const char* file_list[1] = {sstfilename};
rocksdb_ingest_external_file(db, file_list, 1, ing_opt, &err);
CheckNoError(err);
CheckGet(db, roptions, "sstk1", "v1");
CheckGet(db, roptions, "sstk2", "v2");
CheckGet(db, roptions, "sstk3", "v3");
unlink(sstfilename);
rocksdb_sstfilewriter_open(writer, sstfilename, &err);
CheckNoError(err);
rocksdb_sstfilewriter_add(writer, "sstk2", 5, "v4", 2, &err);
CheckNoError(err);
rocksdb_sstfilewriter_add(writer, "sstk22", 6, "v5", 2, &err);
CheckNoError(err);
rocksdb_sstfilewriter_add(writer, "sstk3", 5, "v6", 2, &err);
CheckNoError(err);
rocksdb_sstfilewriter_finish(writer, &err);
CheckNoError(err);
rocksdb_ingest_external_file(db, file_list, 1, ing_opt, &err);
CheckNoError(err);
CheckGet(db, roptions, "sstk1", "v1");
CheckGet(db, roptions, "sstk2", "v4");
CheckGet(db, roptions, "sstk22", "v5");
CheckGet(db, roptions, "sstk3", "v6");
rocksdb_ingestexternalfileoptions_destroy(ing_opt);
rocksdb_sstfilewriter_destroy(writer);
rocksdb_options_destroy(io_options);
rocksdb_envoptions_destroy(env_opt);
}
StartPhase("repair");
{
// If we do not compact here, then the lazy deletion of
@@ -642,7 +557,6 @@ int main(int argc, char** argv) {
rocksdb_close(db);
rocksdb_options_set_create_if_missing(options, 0);
rocksdb_options_set_error_if_exists(options, 0);
rocksdb_options_set_wal_recovery_mode(options, 2);
rocksdb_repair_db(options, dbname, &err);
CheckNoError(err);
db = rocksdb_open(options, dbname, &err);
@@ -767,7 +681,7 @@ int main(int argc, char** argv) {
rocksdb_destroy_db(options, dbname, &err);
CheckNoError(err)
rocksdb_options_t* db_options = rocksdb_options_create();
rocksdb_options_t* db_options = rocksdb_options_create();
rocksdb_options_set_create_if_missing(db_options, 1);
db = rocksdb_open(db_options, dbname, &err);
CheckNoError(err)
@@ -854,30 +768,6 @@ int main(int argc, char** argv) {
CheckNoError(err);
rocksdb_iter_destroy(iter);
rocksdb_column_family_handle_t* iters_cf_handles[2] = { handles[0], handles[1] };
rocksdb_iterator_t* iters_handles[2];
rocksdb_create_iterators(db, roptions, iters_cf_handles, iters_handles, 2, &err);
CheckNoError(err);
iter = iters_handles[0];
CheckCondition(!rocksdb_iter_valid(iter));
rocksdb_iter_seek_to_first(iter);
CheckCondition(!rocksdb_iter_valid(iter));
rocksdb_iter_destroy(iter);
iter = iters_handles[1];
CheckCondition(!rocksdb_iter_valid(iter));
rocksdb_iter_seek_to_first(iter);
CheckCondition(rocksdb_iter_valid(iter));
for (i = 0; rocksdb_iter_valid(iter) != 0; rocksdb_iter_next(iter)) {
i++;
}
CheckCondition(i == 1);
rocksdb_iter_get_error(iter, &err);
CheckNoError(err);
rocksdb_iter_destroy(iter);
rocksdb_drop_column_family(db, handles[1], &err);
CheckNoError(err);
for (i = 0; i < 2; i++) {
@@ -896,7 +786,6 @@ int main(int argc, char** argv) {
rocksdb_options_set_prefix_extractor(options, rocksdb_slicetransform_create_fixed_prefix(3));
rocksdb_options_set_hash_skip_list_rep(options, 5000, 4, 4);
rocksdb_options_set_plain_table_factory(options, 4, 10, 0.75, 16);
rocksdb_options_set_allow_concurrent_memtable_write(options, 0);
db = rocksdb_open(options, dbname, &err);
CheckNoError(err);
@@ -1012,28 +901,6 @@ int main(int argc, char** argv) {
}
}
// Simple sanity check that setting memtable rep works.
StartPhase("memtable_reps");
{
// Create database with vector memtable.
rocksdb_close(db);
rocksdb_destroy_db(options, dbname, &err);
CheckNoError(err);
rocksdb_options_set_memtable_vector_rep(options);
db = rocksdb_open(options, dbname, &err);
CheckNoError(err);
// Create database with hash skiplist memtable.
rocksdb_close(db);
rocksdb_destroy_db(options, dbname, &err);
CheckNoError(err);
rocksdb_options_set_hash_skip_list_rep(options, 5000, 4, 4);
db = rocksdb_open(options, dbname, &err);
CheckNoError(err);
}
StartPhase("cleanup");
rocksdb_close(db);
rocksdb_options_destroy(options);
@@ -1049,6 +916,7 @@ int main(int argc, char** argv) {
}
#else
#include <stdio.h>
int main() {
fprintf(stderr, "SKIPPED\n");
+72 -335
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -26,8 +26,8 @@
#include "db/table_properties_collector.h"
#include "db/version_set.h"
#include "db/write_controller.h"
#include "db/writebuffer.h"
#include "memtable/hash_skiplist_rep.h"
#include "table/block_based_table_factory.h"
#include "util/autovector.h"
#include "util/compression.h"
#include "util/options_helper.h"
@@ -46,11 +46,6 @@ ColumnFamilyHandleImpl::ColumnFamilyHandleImpl(
ColumnFamilyHandleImpl::~ColumnFamilyHandleImpl() {
if (cfd_ != nullptr) {
#ifndef ROCKSDB_LITE
for (auto& listener : cfd_->ioptions()->listeners) {
listener->OnColumnFamilyHandleDeletionStarted(this);
}
#endif // ROCKSDB_LITE
// Job id == 0 means that this is not our background process, but rather
// user thread
JobContext job_context(0);
@@ -73,27 +68,16 @@ const std::string& ColumnFamilyHandleImpl::GetName() const {
return cfd()->GetName();
}
Status ColumnFamilyHandleImpl::GetDescriptor(ColumnFamilyDescriptor* desc) {
#ifndef ROCKSDB_LITE
// accessing mutable cf-options requires db mutex.
InstrumentedMutexLock l(mutex_);
*desc = ColumnFamilyDescriptor(cfd()->GetName(), cfd()->GetLatestCFOptions());
return Status::OK();
#else
return Status::NotSupported();
#endif // !ROCKSDB_LITE
}
const Comparator* ColumnFamilyHandleImpl::GetComparator() const {
const Comparator* ColumnFamilyHandleImpl::user_comparator() const {
return cfd()->user_comparator();
}
void GetIntTblPropCollectorFactory(
const ImmutableCFOptions& ioptions,
const ColumnFamilyOptions& cf_options,
std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
int_tbl_prop_collector_factories) {
auto& collector_factories = ioptions.table_properties_collector_factories;
for (size_t i = 0; i < ioptions.table_properties_collector_factories.size();
auto& collector_factories = cf_options.table_properties_collector_factories;
for (size_t i = 0; i < cf_options.table_properties_collector_factories.size();
++i) {
assert(collector_factories[i]);
int_tbl_prop_collector_factories->emplace_back(
@@ -126,26 +110,18 @@ Status CheckCompressionSupported(const ColumnFamilyOptions& cf_options) {
return Status::OK();
}
Status CheckConcurrentWritesSupported(const ColumnFamilyOptions& cf_options) {
if (cf_options.inplace_update_support) {
return Status::InvalidArgument(
"In-place memtable updates (inplace_update_support) is not compatible "
"with concurrent writes (allow_concurrent_memtable_write)");
}
if (!cf_options.memtable_factory->IsInsertConcurrentlySupported()) {
return Status::InvalidArgument(
"Memtable doesn't concurrent writes (allow_concurrent_memtable_write)");
}
return Status::OK();
}
ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
ColumnFamilyOptions SanitizeOptions(const DBOptions& db_options,
const InternalKeyComparator* icmp,
const ColumnFamilyOptions& src) {
ColumnFamilyOptions result = src;
size_t clamp_max = std::conditional<
sizeof(size_t) == 4, std::integral_constant<size_t, 0xffffffff>,
std::integral_constant<uint64_t, 64ull << 30>>::type::value;
ClipToRange(&result.write_buffer_size, ((size_t)64) << 10, clamp_max);
result.comparator = icmp;
#ifdef OS_MACOSX
// TODO(icanadi) make write_buffer_size uint64_t instead of size_t
ClipToRange(&result.write_buffer_size, ((size_t)64) << 10, ((size_t)1) << 30);
#else
ClipToRange(&result.write_buffer_size,
((size_t)64) << 10, ((size_t)64) << 30);
#endif
// if user sets arena_block_size, we trust user to use this value. Otherwise,
// calculate a proper value from writer_buffer_size;
if (result.arena_block_size <= 0) {
@@ -159,10 +135,6 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
result.min_write_buffer_number_to_merge =
std::min(result.min_write_buffer_number_to_merge,
result.max_write_buffer_number - 1);
if (result.min_write_buffer_number_to_merge < 1) {
result.min_write_buffer_number_to_merge = 1;
}
if (result.num_levels < 1) {
result.num_levels = 1;
}
@@ -176,12 +148,6 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
if (result.max_write_buffer_number_to_maintain < 0) {
result.max_write_buffer_number_to_maintain = result.max_write_buffer_number;
}
// bloom filter size shouldn't exceed 1/4 of memtable size.
if (result.memtable_prefix_bloom_size_ratio > 0.25) {
result.memtable_prefix_bloom_size_ratio = 0.25;
} else if (result.memtable_prefix_bloom_size_ratio < 0) {
result.memtable_prefix_bloom_size_ratio = 0;
}
XFUNC_TEST("memtablelist_history", "transaction_xftest_SanitizeOptions",
xf_transaction_set_memtable_history1,
xf_transaction_set_memtable_history,
@@ -209,10 +175,6 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
result.level0_stop_writes_trigger = std::numeric_limits<int>::max();
}
if (result.max_bytes_for_level_multiplier <= 0) {
result.max_bytes_for_level_multiplier = 1;
}
if (result.level0_file_num_compaction_trigger == 0) {
Warn(db_options.info_log.get(),
"level0_file_num_compaction_trigger cannot be 0");
@@ -249,17 +211,6 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
result.level0_slowdown_writes_trigger,
result.level0_file_num_compaction_trigger);
}
if (result.soft_pending_compaction_bytes_limit == 0) {
result.soft_pending_compaction_bytes_limit =
result.hard_pending_compaction_bytes_limit;
} else if (result.hard_pending_compaction_bytes_limit > 0 &&
result.soft_pending_compaction_bytes_limit >
result.hard_pending_compaction_bytes_limit) {
result.soft_pending_compaction_bytes_limit =
result.hard_pending_compaction_bytes_limit;
}
if (result.level_compaction_dynamic_level_bytes) {
if (result.compaction_style != kCompactionStyleLevel ||
db_options.db_paths.size() > 1U) {
@@ -271,10 +222,6 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
}
}
if (result.max_compaction_bytes == 0) {
result.max_compaction_bytes = result.target_file_size_base * 25;
}
return result;
}
@@ -342,8 +289,8 @@ void SuperVersionUnrefHandle(void* ptr) {
ColumnFamilyData::ColumnFamilyData(
uint32_t id, const std::string& name, Version* _dummy_versions,
Cache* _table_cache, WriteBufferManager* write_buffer_manager,
const ColumnFamilyOptions& cf_options, const ImmutableDBOptions& db_options,
Cache* _table_cache, WriteBuffer* write_buffer,
const ColumnFamilyOptions& cf_options, const DBOptions* db_options,
const EnvOptions& env_options, ColumnFamilySet* column_family_set)
: id_(id),
name_(name),
@@ -352,15 +299,14 @@ ColumnFamilyData::ColumnFamilyData(
refs_(0),
dropped_(false),
internal_comparator_(cf_options.comparator),
initial_cf_options_(SanitizeOptions(db_options, cf_options)),
ioptions_(db_options, initial_cf_options_),
mutable_cf_options_(initial_cf_options_),
is_delete_range_supported_(strcmp(cf_options.table_factory->Name(),
BlockBasedTableFactory().Name()) == 0),
write_buffer_manager_(write_buffer_manager),
options_(*db_options,
SanitizeOptions(*db_options, &internal_comparator_, cf_options)),
ioptions_(options_),
mutable_cf_options_(options_, ioptions_),
write_buffer_(write_buffer),
mem_(nullptr),
imm_(ioptions_.min_write_buffer_number_to_merge,
ioptions_.max_write_buffer_number_to_maintain),
imm_(options_.min_write_buffer_number_to_merge,
options_.max_write_buffer_number_to_maintain),
super_version_(nullptr),
super_version_number_(0),
local_sv_(new ThreadLocalPtr(&SuperVersionUnrefHandle)),
@@ -369,18 +315,16 @@ ColumnFamilyData::ColumnFamilyData(
log_number_(0),
column_family_set_(column_family_set),
pending_flush_(false),
pending_compaction_(false),
prev_compaction_needed_bytes_(0),
allow_2pc_(db_options.allow_2pc) {
pending_compaction_(false) {
Ref();
// Convert user defined table properties collector factories to internal ones.
GetIntTblPropCollectorFactory(ioptions_, &int_tbl_prop_collector_factories_);
GetIntTblPropCollectorFactory(options_, &int_tbl_prop_collector_factories_);
// if _dummy_versions is nullptr, then this is a dummy column family.
if (_dummy_versions != nullptr) {
internal_stats_.reset(
new InternalStats(ioptions_.num_levels, db_options.env, this));
new InternalStats(ioptions_.num_levels, db_options->env, this));
table_cache_.reset(new TableCache(ioptions_, env_options, _table_cache));
if (ioptions_.compaction_style == kCompactionStyleLevel) {
compaction_picker_.reset(
@@ -412,7 +356,7 @@ ColumnFamilyData::ColumnFamilyData(
if (column_family_set_->NumberOfColumnFamilies() < 10) {
Log(InfoLogLevel::INFO_LEVEL, ioptions_.info_log,
"--------------- Options for column family [%s]:\n", name.c_str());
initial_cf_options_.Dump(ioptions_.info_log);
options_.DumpCFOptions(ioptions_.info_log);
} else {
Log(InfoLogLevel::INFO_LEVEL, ioptions_.info_log,
"\t(skipping printing options)\n");
@@ -489,142 +433,13 @@ void ColumnFamilyData::SetDropped() {
column_family_set_->RemoveColumnFamily(this);
}
ColumnFamilyOptions ColumnFamilyData::GetLatestCFOptions() const {
return BuildColumnFamilyOptions(initial_cf_options_, mutable_cf_options_);
}
uint64_t ColumnFamilyData::OldestLogToKeep() {
auto current_log = GetLogNumber();
if (allow_2pc_) {
auto imm_prep_log = imm()->GetMinLogContainingPrepSection();
auto mem_prep_log = mem()->GetMinLogContainingPrepSection();
if (imm_prep_log > 0 && imm_prep_log < current_log) {
current_log = imm_prep_log;
}
if (mem_prep_log > 0 && mem_prep_log < current_log) {
current_log = mem_prep_log;
}
}
return current_log;
}
const double kIncSlowdownRatio = 0.8;
const double kDecSlowdownRatio = 1 / kIncSlowdownRatio;
const double kNearStopSlowdownRatio = 0.6;
const double kDelayRecoverSlowdownRatio = 1.4;
namespace {
// If penalize_stop is true, we further reduce slowdown rate.
std::unique_ptr<WriteControllerToken> SetupDelay(
WriteController* write_controller, uint64_t compaction_needed_bytes,
uint64_t prev_compaction_need_bytes, bool penalize_stop,
bool auto_comapctions_disabled) {
const uint64_t kMinWriteRate = 16 * 1024u; // Minimum write rate 16KB/s.
uint64_t max_write_rate = write_controller->max_delayed_write_rate();
uint64_t write_rate = write_controller->delayed_write_rate();
if (auto_comapctions_disabled) {
// When auto compaction is disabled, always use the value user gave.
write_rate = max_write_rate;
} else if (write_controller->NeedsDelay() && max_write_rate > kMinWriteRate) {
// If user gives rate less than kMinWriteRate, don't adjust it.
//
// If already delayed, need to adjust based on previous compaction debt.
// When there are two or more column families require delay, we always
// increase or reduce write rate based on information for one single
// column family. It is likely to be OK but we can improve if there is a
// problem.
// Ignore compaction_needed_bytes = 0 case because compaction_needed_bytes
// is only available in level-based compaction
//
// If the compaction debt stays the same as previously, we also further slow
// down. It usually means a mem table is full. It's mainly for the case
// where both of flush and compaction are much slower than the speed we
// insert to mem tables, so we need to actively slow down before we get
// feedback signal from compaction and flushes to avoid the full stop
// because of hitting the max write buffer number.
//
// If DB just falled into the stop condition, we need to further reduce
// the write rate to avoid the stop condition.
if (penalize_stop) {
// Penalize the near stop or stop condition by more agressive slowdown.
// This is to provide the long term slowdown increase signal.
// The penalty is more than the reward of recovering to the normal
// condition.
write_rate = static_cast<uint64_t>(static_cast<double>(write_rate) *
kNearStopSlowdownRatio);
if (write_rate < kMinWriteRate) {
write_rate = kMinWriteRate;
}
} else if (prev_compaction_need_bytes > 0 &&
prev_compaction_need_bytes <= compaction_needed_bytes) {
write_rate = static_cast<uint64_t>(static_cast<double>(write_rate) *
kIncSlowdownRatio);
if (write_rate < kMinWriteRate) {
write_rate = kMinWriteRate;
}
} else if (prev_compaction_need_bytes > compaction_needed_bytes) {
// We are speeding up by ratio of kSlowdownRatio when we have paid
// compaction debt. But we'll never speed up to faster than the write rate
// given by users.
write_rate = static_cast<uint64_t>(static_cast<double>(write_rate) *
kDecSlowdownRatio);
if (write_rate > max_write_rate) {
write_rate = max_write_rate;
}
}
}
return write_controller->GetDelayToken(write_rate);
}
int GetL0ThresholdSpeedupCompaction(int level0_file_num_compaction_trigger,
int level0_slowdown_writes_trigger) {
// SanitizeOptions() ensures it.
assert(level0_file_num_compaction_trigger <= level0_slowdown_writes_trigger);
if (level0_file_num_compaction_trigger < 0) {
return std::numeric_limits<int>::max();
}
const int64_t twice_level0_trigger =
static_cast<int64_t>(level0_file_num_compaction_trigger) * 2;
const int64_t one_fourth_trigger_slowdown =
static_cast<int64_t>(level0_file_num_compaction_trigger) +
((level0_slowdown_writes_trigger - level0_file_num_compaction_trigger) /
4);
assert(twice_level0_trigger >= 0);
assert(one_fourth_trigger_slowdown >= 0);
// 1/4 of the way between L0 compaction trigger threshold and slowdown
// condition.
// Or twice as compaction trigger, if it is smaller.
int64_t res = std::min(twice_level0_trigger, one_fourth_trigger_slowdown);
if (res >= port::kMaxInt32) {
return port::kMaxInt32;
} else {
// res fits in int
return static_cast<int>(res);
}
}
} // namespace
void ColumnFamilyData::RecalculateWriteStallConditions(
const MutableCFOptions& mutable_cf_options) {
if (current_ != nullptr) {
auto* vstorage = current_->storage_info();
const double score = vstorage->max_compaction_score();
const int max_level = vstorage->max_compaction_score_level();
auto write_controller = column_family_set_->write_controller_;
uint64_t compaction_needed_bytes =
vstorage->estimated_compaction_needed_bytes();
bool was_stopped = write_controller->IsStopped();
bool needed_delay = write_controller->NeedsDelay();
if (imm()->NumNotFlushed() >= mutable_cf_options.max_write_buffer_number) {
write_controller_token_ = write_controller->GetStopToken();
@@ -634,9 +449,8 @@ void ColumnFamilyData::RecalculateWriteStallConditions(
"(waiting for flush), max_write_buffer_number is set to %d",
name_.c_str(), imm()->NumNotFlushed(),
mutable_cf_options.max_write_buffer_number);
} else if (!mutable_cf_options.disable_auto_compactions &&
vstorage->l0_delay_trigger_count() >=
mutable_cf_options.level0_stop_writes_trigger) {
} else if (vstorage->l0_delay_trigger_count() >=
mutable_cf_options.level0_stop_writes_trigger) {
write_controller_token_ = write_controller->GetStopToken();
internal_stats_->AddCFStats(InternalStats::LEVEL0_NUM_FILES_TOTAL, 1);
if (compaction_picker_->IsLevel0CompactionInProgress()) {
@@ -646,118 +460,38 @@ void ColumnFamilyData::RecalculateWriteStallConditions(
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Stopping writes because we have %d level-0 files",
name_.c_str(), vstorage->l0_delay_trigger_count());
} else if (!mutable_cf_options.disable_auto_compactions &&
mutable_cf_options.hard_pending_compaction_bytes_limit > 0 &&
compaction_needed_bytes >=
} else if (mutable_cf_options.hard_pending_compaction_bytes_limit > 0 &&
vstorage->estimated_compaction_needed_bytes() >=
mutable_cf_options.hard_pending_compaction_bytes_limit) {
write_controller_token_ = write_controller->GetStopToken();
internal_stats_->AddCFStats(
InternalStats::HARD_PENDING_COMPACTION_BYTES_LIMIT, 1);
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Stopping writes because of estimated pending compaction "
"bytes %" PRIu64,
name_.c_str(), compaction_needed_bytes);
} else if (mutable_cf_options.max_write_buffer_number > 3 &&
imm()->NumNotFlushed() >=
mutable_cf_options.max_write_buffer_number - 1) {
write_controller_token_ =
SetupDelay(write_controller, compaction_needed_bytes,
prev_compaction_needed_bytes_, was_stopped,
mutable_cf_options.disable_auto_compactions);
internal_stats_->AddCFStats(InternalStats::MEMTABLE_SLOWDOWN, 1);
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Stalling writes because we have %d immutable memtables "
"(waiting for flush), max_write_buffer_number is set to %d "
"rate %" PRIu64,
name_.c_str(), imm()->NumNotFlushed(),
mutable_cf_options.max_write_buffer_number,
write_controller->delayed_write_rate());
} else if (!mutable_cf_options.disable_auto_compactions &&
mutable_cf_options.level0_slowdown_writes_trigger >= 0 &&
"[%s] Stopping writes because estimated pending compaction "
"bytes exceed %" PRIu64,
name_.c_str(), vstorage->estimated_compaction_needed_bytes());
} else if (mutable_cf_options.level0_slowdown_writes_trigger >= 0 &&
vstorage->l0_delay_trigger_count() >=
mutable_cf_options.level0_slowdown_writes_trigger) {
// L0 is the last two files from stopping.
bool near_stop = vstorage->l0_delay_trigger_count() >=
mutable_cf_options.level0_stop_writes_trigger - 2;
write_controller_token_ =
SetupDelay(write_controller, compaction_needed_bytes,
prev_compaction_needed_bytes_, was_stopped || near_stop,
mutable_cf_options.disable_auto_compactions);
write_controller_token_ = write_controller->GetDelayToken();
internal_stats_->AddCFStats(InternalStats::LEVEL0_SLOWDOWN_TOTAL, 1);
if (compaction_picker_->IsLevel0CompactionInProgress()) {
internal_stats_->AddCFStats(
InternalStats::LEVEL0_SLOWDOWN_WITH_COMPACTION, 1);
}
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Stalling writes because we have %d level-0 files "
"rate %" PRIu64,
name_.c_str(), vstorage->l0_delay_trigger_count(),
write_controller->delayed_write_rate());
} else if (!mutable_cf_options.disable_auto_compactions &&
mutable_cf_options.soft_pending_compaction_bytes_limit > 0 &&
vstorage->estimated_compaction_needed_bytes() >=
mutable_cf_options.soft_pending_compaction_bytes_limit) {
// If the distance to hard limit is less than 1/4 of the gap between soft
// and
// hard bytes limit, we think it is near stop and speed up the slowdown.
bool near_stop =
mutable_cf_options.hard_pending_compaction_bytes_limit > 0 &&
(compaction_needed_bytes -
mutable_cf_options.soft_pending_compaction_bytes_limit) >
3 * (mutable_cf_options.hard_pending_compaction_bytes_limit -
mutable_cf_options.soft_pending_compaction_bytes_limit) /
4;
write_controller_token_ =
SetupDelay(write_controller, compaction_needed_bytes,
prev_compaction_needed_bytes_, was_stopped || near_stop,
mutable_cf_options.disable_auto_compactions);
internal_stats_->AddCFStats(
InternalStats::SOFT_PENDING_COMPACTION_BYTES_LIMIT, 1);
"[%s] Stalling writes because we have %d level-0 files",
name_.c_str(), vstorage->l0_delay_trigger_count());
} else if (mutable_cf_options.soft_rate_limit > 0.0 &&
score > mutable_cf_options.soft_rate_limit) {
write_controller_token_ = write_controller->GetDelayToken();
internal_stats_->RecordLevelNSlowdown(max_level, true);
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Stalling writes because of estimated pending compaction "
"bytes %" PRIu64 " rate %" PRIu64,
name_.c_str(), vstorage->estimated_compaction_needed_bytes(),
write_controller->delayed_write_rate());
"[%s] Stalling writes because we hit soft limit on level %d",
name_.c_str(), max_level);
} else {
if (vstorage->l0_delay_trigger_count() >=
GetL0ThresholdSpeedupCompaction(
mutable_cf_options.level0_file_num_compaction_trigger,
mutable_cf_options.level0_slowdown_writes_trigger)) {
write_controller_token_ =
write_controller->GetCompactionPressureToken();
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Increasing compaction threads because we have %d level-0 "
"files ",
name_.c_str(), vstorage->l0_delay_trigger_count());
} else if (vstorage->estimated_compaction_needed_bytes() >=
mutable_cf_options.soft_pending_compaction_bytes_limit / 4) {
// Increase compaction threads if bytes needed for compaction exceeds
// 1/4 of threshold for slowing down.
// If soft pending compaction byte limit is not set, always speed up
// compaction.
write_controller_token_ =
write_controller->GetCompactionPressureToken();
if (mutable_cf_options.soft_pending_compaction_bytes_limit > 0) {
Log(InfoLogLevel::WARN_LEVEL, ioptions_.info_log,
"[%s] Increasing compaction threads because of estimated pending "
"compaction "
"bytes %" PRIu64,
name_.c_str(), vstorage->estimated_compaction_needed_bytes());
}
} else {
write_controller_token_.reset();
}
// If the DB recovers from delay conditions, we reward with reducing
// double the slowdown ratio. This is to balance the long term slowdown
// increase signal.
if (needed_delay) {
uint64_t write_rate = write_controller->delayed_write_rate();
write_controller->set_delayed_write_rate(static_cast<uint64_t>(
static_cast<double>(write_rate) * kDelayRecoverSlowdownRatio));
}
write_controller_token_.reset();
}
prev_compaction_needed_bytes_ = compaction_needed_bytes;
}
}
@@ -781,7 +515,7 @@ MemTable* ColumnFamilyData::ConstructNewMemtable(
const MutableCFOptions& mutable_cf_options, SequenceNumber earliest_seq) {
assert(current_ != nullptr);
return new MemTable(internal_comparator_, ioptions_, mutable_cf_options,
write_buffer_manager_, earliest_seq);
write_buffer_, earliest_seq);
}
void ColumnFamilyData::CreateNewMemtable(
@@ -807,23 +541,17 @@ Compaction* ColumnFamilyData::PickCompaction(
return result;
}
bool ColumnFamilyData::RangeOverlapWithCompaction(
const Slice& smallest_user_key, const Slice& largest_user_key,
int level) const {
return compaction_picker_->RangeOverlapWithCompaction(
smallest_user_key, largest_user_key, level);
}
const int ColumnFamilyData::kCompactAllLevels = -1;
const int ColumnFamilyData::kCompactToBaseLevel = -2;
Compaction* ColumnFamilyData::CompactRange(
const MutableCFOptions& mutable_cf_options, int input_level,
int output_level, uint32_t output_path_id, const InternalKey* begin,
const InternalKey* end, InternalKey** compaction_end, bool* conflict) {
const MutableCFOptions& mutable_cf_options,
int input_level, int output_level, uint32_t output_path_id,
const InternalKey* begin, const InternalKey* end,
InternalKey** compaction_end) {
auto* result = compaction_picker_->CompactRange(
GetName(), mutable_cf_options, current_->storage_info(), input_level,
output_level, output_path_id, begin, end, compaction_end, conflict);
output_level, output_path_id, begin, end, compaction_end);
if (result != nullptr) {
result->SetInputVersion(current_);
}
@@ -964,21 +692,21 @@ Status ColumnFamilyData::SetOptions(
#endif // ROCKSDB_LITE
ColumnFamilySet::ColumnFamilySet(const std::string& dbname,
const ImmutableDBOptions* db_options,
const DBOptions* db_options,
const EnvOptions& env_options,
Cache* table_cache,
WriteBufferManager* write_buffer_manager,
WriteBuffer* write_buffer,
WriteController* write_controller)
: max_column_family_(0),
dummy_cfd_(new ColumnFamilyData(0, "", nullptr, nullptr, nullptr,
ColumnFamilyOptions(), *db_options,
ColumnFamilyOptions(), db_options,
env_options, nullptr)),
default_cfd_cache_(nullptr),
db_name_(dbname),
db_options_(db_options),
env_options_(env_options),
table_cache_(table_cache),
write_buffer_manager_(write_buffer_manager),
write_buffer_(write_buffer),
write_controller_(write_controller) {
// initialize linked list
dummy_cfd_->prev_ = dummy_cfd_;
@@ -1041,9 +769,10 @@ ColumnFamilyData* ColumnFamilySet::CreateColumnFamily(
const std::string& name, uint32_t id, Version* dummy_versions,
const ColumnFamilyOptions& options) {
assert(column_families_.find(name) == column_families_.end());
ColumnFamilyData* new_cfd = new ColumnFamilyData(
id, name, dummy_versions, table_cache_, write_buffer_manager_, options,
*db_options_, env_options_, this);
ColumnFamilyData* new_cfd =
new ColumnFamilyData(id, name, dummy_versions, table_cache_,
write_buffer_, options, db_options_,
env_options_, this);
column_families_.insert({name, id});
column_family_data_.insert({id, new_cfd});
max_column_family_ = std::max(max_column_family_, id);
@@ -1108,6 +837,13 @@ ColumnFamilyHandle* ColumnFamilyMemTablesImpl::GetColumnFamilyHandle() {
return &handle_;
}
void ColumnFamilyMemTablesImpl::CheckMemtableFull() {
if (current_ != nullptr && current_->mem()->ShouldScheduleFlush()) {
flush_scheduler_->ScheduleFlush(current_);
current_->mem()->MarkFlushScheduled();
}
}
uint32_t GetColumnFamilyID(ColumnFamilyHandle* column_family) {
uint32_t column_family_id = 0;
if (column_family != nullptr) {
@@ -1120,7 +856,8 @@ uint32_t GetColumnFamilyID(ColumnFamilyHandle* column_family) {
const Comparator* GetColumnFamilyUserComparator(
ColumnFamilyHandle* column_family) {
if (column_family != nullptr) {
return column_family->GetComparator();
auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
return cfh->user_comparator();
}
return nullptr;
}
+53 -91
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -15,15 +15,17 @@
#include <atomic>
#include "db/memtable_list.h"
#include "db/table_cache.h"
#include "db/table_properties_collector.h"
#include "db/write_batch_internal.h"
#include "db/write_controller.h"
#include "db/table_cache.h"
#include "db/table_properties_collector.h"
#include "db/flush_scheduler.h"
#include "rocksdb/compaction_job_stats.h"
#include "rocksdb/db.h"
#include "rocksdb/env.h"
#include "rocksdb/options.h"
#include "util/cf_options.h"
#include "util/instrumented_mutex.h"
#include "util/mutable_cf_options.h"
#include "util/thread_local.h"
namespace rocksdb {
@@ -42,8 +44,6 @@ class LogBuffer;
class InstrumentedMutex;
class InstrumentedMutexLock;
extern const double kIncSlowdownRatio;
// ColumnFamilyHandleImpl is the class that clients use to access different
// column families. It has non-trivial destructor, which gets called when client
// is done using the column family
@@ -55,11 +55,10 @@ class ColumnFamilyHandleImpl : public ColumnFamilyHandle {
// destroy without mutex
virtual ~ColumnFamilyHandleImpl();
virtual ColumnFamilyData* cfd() const { return cfd_; }
virtual const Comparator* user_comparator() const;
virtual uint32_t GetID() const override;
virtual const std::string& GetName() const override;
virtual Status GetDescriptor(ColumnFamilyDescriptor* desc) override;
virtual const Comparator* GetComparator() const override;
private:
ColumnFamilyData* cfd_;
@@ -133,16 +132,14 @@ struct SuperVersion {
extern Status CheckCompressionSupported(const ColumnFamilyOptions& cf_options);
extern Status CheckConcurrentWritesSupported(
const ColumnFamilyOptions& cf_options);
extern ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
extern ColumnFamilyOptions SanitizeOptions(const DBOptions& db_options,
const InternalKeyComparator* icmp,
const ColumnFamilyOptions& src);
// Wrap user defined table proproties collector factories `from cf_options`
// into internal ones in int_tbl_prop_collector_factories. Add a system internal
// one too.
extern void GetIntTblPropCollectorFactory(
const ImmutableCFOptions& ioptions,
const ColumnFamilyOptions& cf_options,
std::vector<std::unique_ptr<IntTblPropCollectorFactory>>*
int_tbl_prop_collector_factories);
@@ -159,16 +156,14 @@ class ColumnFamilyData {
// thread-safe
const std::string& GetName() const { return name_; }
// Ref() can only be called from a context where the caller can guarantee
// that ColumnFamilyData is alive (while holding a non-zero ref already,
// holding a DB mutex, or as the leader in a write batch group).
// Ref() can only be called whily holding a DB mutex or during a
// single-threaded write.
void Ref() { refs_.fetch_add(1, std::memory_order_relaxed); }
// Unref decreases the reference count, but does not handle deletion
// when the count goes to 0. If this method returns true then the
// caller should delete the instance immediately, or later, by calling
// FreeDeadColumnFamilies(). Unref() can only be called while holding
// a DB mutex, or during single-threaded recovery.
// will just decrease reference count to 0, but will not delete it. returns
// true if the ref count was decreased to zero. in that case, it can be
// deleted by the caller immediately, or later, by calling
// FreeDeadColumnFamilies()
// Unref() can only be called while holding a DB mutex
bool Unref() {
int old_refs = refs_.fetch_sub(1, std::memory_order_relaxed);
assert(old_refs > 0);
@@ -200,12 +195,15 @@ class ColumnFamilyData {
void SetLogNumber(uint64_t log_number) { log_number_ = log_number; }
uint64_t GetLogNumber() const { return log_number_; }
// !!! To be deprecated! Please don't not use this function anymore!
const Options* options() const { return &options_; }
// thread-safe
const EnvOptions* soptions() const;
const ImmutableCFOptions* ioptions() const { return &ioptions_; }
// REQUIRES: DB mutex held
// This returns the MutableCFOptions used by current SuperVersion
// You should use this API to reference MutableCFOptions most of the time.
// You shoul use this API to reference MutableCFOptions most of the time.
const MutableCFOptions* GetCurrentMutableCFOptions() const {
return &(super_version_->mutable_cf_options);
}
@@ -214,14 +212,6 @@ class ColumnFamilyData {
const MutableCFOptions* GetLatestMutableCFOptions() const {
return &mutable_cf_options_;
}
// REQUIRES: DB mutex held
// Build ColumnFamiliesOptions with immutable options and latest mutable
// options.
ColumnFamilyOptions GetLatestCFOptions() const;
bool is_delete_range_supported() { return is_delete_range_supported_; }
#ifndef ROCKSDB_LITE
// REQUIRES: DB mutex held
Status SetOptions(
@@ -234,14 +224,11 @@ class ColumnFamilyData {
MemTable* mem() { return mem_; }
Version* current() { return current_; }
Version* dummy_versions() { return dummy_versions_; }
void SetCurrent(Version* _current);
void SetCurrent(Version* current);
uint64_t GetNumLiveVersions() const; // REQUIRE: DB mutex held
uint64_t GetTotalSstFilesSize() const; // REQUIRE: DB mutex held
void SetMemtable(MemTable* new_mem) { mem_ = new_mem; }
// calculate the oldest log needed for the durability of this column family
uint64_t OldestLogToKeep();
// See Memtable constructor for explanation of earliest_seq param.
MemTable* ConstructNewMemtable(const MutableCFOptions& mutable_cf_options,
SequenceNumber earliest_seq);
@@ -256,24 +243,17 @@ class ColumnFamilyData {
// REQUIRES: DB mutex held
Compaction* PickCompaction(const MutableCFOptions& mutable_options,
LogBuffer* log_buffer);
// Check if the passed range overlap with any running compactions.
// REQUIRES: DB mutex held
bool RangeOverlapWithCompaction(const Slice& smallest_user_key,
const Slice& largest_user_key,
int level) const;
// A flag to tell a manual compaction is to compact all levels together
// instad of for specific level.
static const int kCompactAllLevels;
// A flag to tell a manual compaction's output is base level.
static const int kCompactToBaseLevel;
// REQUIRES: DB mutex held
Compaction* CompactRange(const MutableCFOptions& mutable_cf_options,
int input_level, int output_level,
uint32_t output_path_id, const InternalKey* begin,
const InternalKey* end, InternalKey** compaction_end,
bool* manual_conflict);
Compaction* CompactRange(
const MutableCFOptions& mutable_cf_options,
int input_level, int output_level, uint32_t output_path_id,
const InternalKey* begin, const InternalKey* end,
InternalKey** compaction_end);
CompactionPicker* compaction_picker() { return compaction_picker_.get(); }
// thread-safe
@@ -325,6 +305,15 @@ class ColumnFamilyData {
bool pending_flush() { return pending_flush_; }
bool pending_compaction() { return pending_compaction_; }
private:
friend class ColumnFamilySet;
ColumnFamilyData(uint32_t id, const std::string& name,
Version* dummy_versions, Cache* table_cache,
WriteBuffer* write_buffer,
const ColumnFamilyOptions& options,
const DBOptions* db_options, const EnvOptions& env_options,
ColumnFamilySet* column_family_set);
// Recalculate some small conditions, which are changed only during
// compaction, adding new memtable and/or
// recalculation of compaction score. These values are used in
@@ -333,16 +322,6 @@ class ColumnFamilyData {
void RecalculateWriteStallConditions(
const MutableCFOptions& mutable_cf_options);
private:
friend class ColumnFamilySet;
ColumnFamilyData(uint32_t id, const std::string& name,
Version* dummy_versions, Cache* table_cache,
WriteBufferManager* write_buffer_manager,
const ColumnFamilyOptions& options,
const ImmutableDBOptions& db_options,
const EnvOptions& env_options,
ColumnFamilySet* column_family_set);
uint32_t id_;
const std::string name_;
Version* dummy_versions_; // Head of circular doubly-linked list of versions.
@@ -355,17 +334,15 @@ class ColumnFamilyData {
std::vector<std::unique_ptr<IntTblPropCollectorFactory>>
int_tbl_prop_collector_factories_;
const ColumnFamilyOptions initial_cf_options_;
const Options options_;
const ImmutableCFOptions ioptions_;
MutableCFOptions mutable_cf_options_;
const bool is_delete_range_supported_;
std::unique_ptr<TableCache> table_cache_;
std::unique_ptr<InternalStats> internal_stats_;
WriteBufferManager* write_buffer_manager_;
WriteBuffer* write_buffer_;
MemTable* mem_;
MemTableList imm_;
@@ -405,11 +382,6 @@ class ColumnFamilyData {
// If true --> this ColumnFamily is currently present in
// DBImpl::compaction_queue_
bool pending_compaction_;
uint64_t prev_compaction_needed_bytes_;
// if the database was opened with 2pc enabled
bool allow_2pc_;
};
// ColumnFamilySet has interesting thread-safety requirements
@@ -456,11 +428,9 @@ class ColumnFamilySet {
ColumnFamilyData* current_;
};
ColumnFamilySet(const std::string& dbname,
const ImmutableDBOptions* db_options,
ColumnFamilySet(const std::string& dbname, const DBOptions* db_options,
const EnvOptions& env_options, Cache* table_cache,
WriteBufferManager* write_buffer_manager,
WriteController* write_controller);
WriteBuffer* write_buffer, WriteController* write_controller);
~ColumnFamilySet();
ColumnFamilyData* GetDefault() const;
@@ -487,8 +457,6 @@ class ColumnFamilySet {
// Don't call while iterating over ColumnFamilySet
void FreeDeadColumnFamilies();
Cache* get_table_cache() { return table_cache_; }
private:
friend class ColumnFamilyData;
// helper function that gets called from cfd destructor
@@ -514,10 +482,10 @@ class ColumnFamilySet {
ColumnFamilyData* default_cfd_cache_;
const std::string db_name_;
const ImmutableDBOptions* const db_options_;
const DBOptions* const db_options_;
const EnvOptions env_options_;
Cache* table_cache_;
WriteBufferManager* write_buffer_manager_;
WriteBuffer* write_buffer_;
WriteController* write_controller_;
};
@@ -525,18 +493,15 @@ class ColumnFamilySet {
// memtables of different column families (specified by ID in the write batch)
class ColumnFamilyMemTablesImpl : public ColumnFamilyMemTables {
public:
explicit ColumnFamilyMemTablesImpl(ColumnFamilySet* column_family_set)
: column_family_set_(column_family_set), current_(nullptr) {}
// Constructs a ColumnFamilyMemTablesImpl equivalent to one constructed
// with the arguments used to construct *orig.
explicit ColumnFamilyMemTablesImpl(ColumnFamilyMemTablesImpl* orig)
: column_family_set_(orig->column_family_set_), current_(nullptr) {}
explicit ColumnFamilyMemTablesImpl(ColumnFamilySet* column_family_set,
FlushScheduler* flush_scheduler)
: column_family_set_(column_family_set),
current_(nullptr),
flush_scheduler_(flush_scheduler) {}
// sets current_ to ColumnFamilyData with column_family_id
// returns false if column family doesn't exist
// REQUIRES: use this function of DBImpl::column_family_memtables_ should be
// under a DB mutex OR from a write thread
// REQUIRES: under a DB mutex OR from a write thread
bool Seek(uint32_t column_family_id) override;
// Returns log number of the selected column family
@@ -544,23 +509,20 @@ class ColumnFamilyMemTablesImpl : public ColumnFamilyMemTables {
uint64_t GetLogNumber() const override;
// REQUIRES: Seek() called first
// REQUIRES: use this function of DBImpl::column_family_memtables_ should be
// under a DB mutex OR from a write thread
// REQUIRES: under a DB mutex OR from a write thread
virtual MemTable* GetMemTable() const override;
// Returns column family handle for the selected column family
// REQUIRES: use this function of DBImpl::column_family_memtables_ should be
// under a DB mutex OR from a write thread
// REQUIRES: under a DB mutex OR from a write thread
virtual ColumnFamilyHandle* GetColumnFamilyHandle() override;
// Cannot be called while another thread is calling Seek().
// REQUIRES: use this function of DBImpl::column_family_memtables_ should be
// under a DB mutex OR from a write thread
virtual ColumnFamilyData* current() override { return current_; }
// REQUIRES: under a DB mutex OR from a write thread
virtual void CheckMemtableFull() override;
private:
ColumnFamilySet* column_family_set_;
ColumnFamilyData* current_;
FlushScheduler* flush_scheduler_;
ColumnFamilyHandleInternal handle_;
};
+43 -1864
View File
File diff suppressed because it is too large. Load diff
+8 -108
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2014, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -7,10 +7,8 @@
#include <mutex>
#include <string>
#include <thread>
#include <vector>
#include "db/db_impl.h"
#include "rocksdb/db.h"
#include "rocksdb/env.h"
#include "util/string_util.h"
@@ -50,10 +48,6 @@ class FlushedFileCollector : public EventListener {
}
return result;
}
void ClearFlushedFiles() {
std::lock_guard<std::mutex> lock(mutex_);
flushed_files_.clear();
}
private:
std::vector<std::string> flushed_files_;
@@ -113,19 +107,19 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
break;
}
}
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
delete db;
}
TEST_F(CompactFilesTest, ObsoleteFiles) {
Options options;
// to trigger compaction more easily
const int kWriteBufferSize = 65536;
const int kWriteBufferSize = 10000;
options.create_if_missing = true;
// Disable RocksDB background compaction.
options.compaction_style = kCompactionStyleNone;
options.level0_slowdown_writes_trigger = (1 << 30);
options.level0_stop_writes_trigger = (1 << 30);
// Small slowdown and stop trigger for experimental purpose.
options.level0_slowdown_writes_trigger = 20;
options.level0_stop_writes_trigger = 20;
options.write_buffer_size = kWriteBufferSize;
options.max_write_buffer_number = 2;
options.compression = kNoCompression;
@@ -147,8 +141,10 @@ TEST_F(CompactFilesTest, ObsoleteFiles) {
}
auto l0_files = collector->GetFlushedFiles();
CompactionOptions compact_opt;
compact_opt.compression = kNoCompression;
compact_opt.output_file_size_limit = kWriteBufferSize * 5;
ASSERT_OK(db->CompactFiles(CompactionOptions(), l0_files, 1));
reinterpret_cast<DBImpl*>(db)->TEST_WaitForCompact();
// verify all compaction input files are deleted
for (auto fname : l0_files) {
@@ -157,102 +153,6 @@ TEST_F(CompactFilesTest, ObsoleteFiles) {
delete db;
}
TEST_F(CompactFilesTest, NotCutOutputOnLevel0) {
Options options;
options.create_if_missing = true;
// Disable RocksDB background compaction.
options.compaction_style = kCompactionStyleNone;
options.level0_slowdown_writes_trigger = 1000;
options.level0_stop_writes_trigger = 1000;
options.write_buffer_size = 65536;
options.max_write_buffer_number = 2;
options.compression = kNoCompression;
options.max_compaction_bytes = 5000;
// Add listener
FlushedFileCollector* collector = new FlushedFileCollector();
options.listeners.emplace_back(collector);
DB* db = nullptr;
DestroyDB(db_name_, options);
Status s = DB::Open(options, db_name_, &db);
assert(s.ok());
assert(db);
// create couple files
for (int i = 0; i < 500; ++i) {
db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
}
reinterpret_cast<DBImpl*>(db)->TEST_WaitForFlushMemTable();
auto l0_files_1 = collector->GetFlushedFiles();
collector->ClearFlushedFiles();
for (int i = 0; i < 500; ++i) {
db->Put(WriteOptions(), ToString(i), std::string(1000, 'a' + (i % 26)));
}
reinterpret_cast<DBImpl*>(db)->TEST_WaitForFlushMemTable();
auto l0_files_2 = collector->GetFlushedFiles();
ASSERT_OK(db->CompactFiles(CompactionOptions(), l0_files_1, 0));
ASSERT_OK(db->CompactFiles(CompactionOptions(), l0_files_2, 0));
// no assertion failure
delete db;
}
TEST_F(CompactFilesTest, CapturingPendingFiles) {
Options options;
options.create_if_missing = true;
// Disable RocksDB background compaction.
options.compaction_style = kCompactionStyleNone;
// Always do full scans for obsolete files (needed to reproduce the issue).
options.delete_obsolete_files_period_micros = 0;
// Add listener.
FlushedFileCollector* collector = new FlushedFileCollector();
options.listeners.emplace_back(collector);
DB* db = nullptr;
DestroyDB(db_name_, options);
Status s = DB::Open(options, db_name_, &db);
assert(s.ok());
assert(db);
// Create 5 files.
for (int i = 0; i < 5; ++i) {
db->Put(WriteOptions(), "key" + ToString(i), "value");
db->Flush(FlushOptions());
}
auto l0_files = collector->GetFlushedFiles();
EXPECT_EQ(5, l0_files.size());
rocksdb::SyncPoint::GetInstance()->LoadDependency({
{"CompactFilesImpl:2", "CompactFilesTest.CapturingPendingFiles:0"},
{"CompactFilesTest.CapturingPendingFiles:1", "CompactFilesImpl:3"},
});
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
// Start compacting files.
std::thread compaction_thread(
[&] { EXPECT_OK(db->CompactFiles(CompactionOptions(), l0_files, 1)); });
// In the meantime flush another file.
TEST_SYNC_POINT("CompactFilesTest.CapturingPendingFiles:0");
db->Put(WriteOptions(), "key5", "value");
db->Flush(FlushOptions());
TEST_SYNC_POINT("CompactFilesTest.CapturingPendingFiles:1");
compaction_thread.join();
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
delete db;
// Make sure we can reopen the DB.
s = DB::Open(options, db_name_, &db);
ASSERT_TRUE(s.ok());
assert(db);
delete db;
}
} // namespace rocksdb
int main(int argc, char** argv) {
+6 -6
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2014, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -46,7 +46,7 @@ Status CompactedDBImpl::Get(const ReadOptions& options,
ColumnFamilyHandle*, const Slice& key, std::string* value) {
GetContext get_context(user_comparator_, nullptr, nullptr, nullptr,
GetContext::kNotFound, key, value, nullptr, nullptr,
nullptr, nullptr);
nullptr);
LookupKey lkey(key, kMaxSequenceNumber);
files_.files[FindFile(key)].fd.table_reader->Get(
options, lkey.internal_key(), &get_context);
@@ -77,7 +77,7 @@ std::vector<Status> CompactedDBImpl::MultiGet(const ReadOptions& options,
if (r != nullptr) {
GetContext get_context(user_comparator_, nullptr, nullptr, nullptr,
GetContext::kNotFound, keys[idx], &(*values)[idx],
nullptr, nullptr, nullptr, nullptr);
nullptr, nullptr, nullptr);
LookupKey lkey(keys[idx], kMaxSequenceNumber);
r->Get(options, lkey.internal_key(), &get_context);
if (get_context.State() == GetContext::kFound) {
@@ -93,7 +93,7 @@ Status CompactedDBImpl::Init(const Options& options) {
mutex_.Lock();
ColumnFamilyDescriptor cf(kDefaultColumnFamilyName,
ColumnFamilyOptions(options));
Status s = Recover({cf}, true /* read only */, false, true);
Status s = Recover({ cf }, true /* read only */, false);
if (s.ok()) {
cfd_ = reinterpret_cast<ColumnFamilyHandleImpl*>(
DefaultColumnFamily())->cfd();
@@ -151,9 +151,9 @@ Status CompactedDBImpl::Open(const Options& options,
std::unique_ptr<CompactedDBImpl> db(new CompactedDBImpl(db_options, dbname));
Status s = db->Init(options);
if (s.ok()) {
Log(INFO_LEVEL, db->immutable_db_options_.info_log,
Log(INFO_LEVEL, db->db_options_.info_log,
"Opened the db as fully compacted mode");
LogFlush(db->immutable_db_options_.info_log);
LogFlush(db->db_options_.info_log);
*dbptr = db.release();
}
return s;
+1 -1
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2014, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
+54 -34
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -16,8 +16,8 @@
#include <inttypes.h>
#include <vector>
#include "db/column_family.h"
#include "rocksdb/compaction_filter.h"
#include "db/column_family.h"
#include "util/logging.h"
#include "util/sync_point.h"
@@ -46,7 +46,7 @@ void Compaction::GetBoundaryKeys(
Slice* largest_user_key) {
bool initialized = false;
const Comparator* ucmp = vstorage->InternalComparator()->user_comparator();
for (size_t i = 0; i < inputs.size(); ++i) {
for (uint32_t i = 0; i < inputs.size(); ++i) {
if (inputs[i].files.empty()) {
continue;
}
@@ -128,8 +128,8 @@ bool Compaction::TEST_IsBottommostLevel(
bool Compaction::IsFullCompaction(
VersionStorageInfo* vstorage,
const std::vector<CompactionInputFiles>& inputs) {
size_t num_files_in_compaction = 0;
size_t total_num_files = 0;
int num_files_in_compaction = 0;
int total_num_files = 0;
for (int l = 0; l < vstorage->num_levels(); l++) {
total_num_files += vstorage->NumLevelFiles(l);
}
@@ -140,21 +140,18 @@ bool Compaction::IsFullCompaction(
}
Compaction::Compaction(VersionStorageInfo* vstorage,
const ImmutableCFOptions& _immutable_cf_options,
const MutableCFOptions& _mutable_cf_options,
std::vector<CompactionInputFiles> _inputs,
int _output_level, uint64_t _target_file_size,
uint64_t _max_compaction_bytes, uint32_t _output_path_id,
CompressionType _compression,
uint64_t _max_grandparent_overlap_bytes,
uint32_t _output_path_id, CompressionType _compression,
std::vector<FileMetaData*> _grandparents,
bool _manual_compaction, double _score,
bool _deletion_compaction,
CompactionReason _compaction_reason)
bool _deletion_compaction)
: start_level_(_inputs[0].level),
output_level_(_output_level),
max_output_file_size_(_target_file_size),
max_compaction_bytes_(_max_compaction_bytes),
immutable_cf_options_(_immutable_cf_options),
max_grandparent_overlap_bytes_(_max_grandparent_overlap_bytes),
mutable_cf_options_(_mutable_cf_options),
input_version_(nullptr),
number_levels_(vstorage->num_levels()),
@@ -164,15 +161,14 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
deletion_compaction_(_deletion_compaction),
inputs_(std::move(_inputs)),
grandparents_(std::move(_grandparents)),
grandparent_index_(0),
seen_key_(false),
overlapped_bytes_(0),
score_(_score),
bottommost_level_(IsBottommostLevel(output_level_, vstorage, inputs_)),
is_full_compaction_(IsFullCompaction(vstorage, inputs_)),
is_manual_compaction_(_manual_compaction),
compaction_reason_(_compaction_reason) {
is_manual_compaction_(_manual_compaction) {
MarkFilesBeingCompacted(true);
if (is_manual_compaction_) {
compaction_reason_ = CompactionReason::kManualCompaction;
}
#ifndef NDEBUG
for (size_t i = 1; i < inputs_.size(); ++i) {
@@ -188,8 +184,6 @@ Compaction::Compaction(VersionStorageInfo* vstorage,
&arena_);
}
}
GetBoundaryKeys(vstorage, inputs_, &smallest_user_key_, &largest_user_key_);
}
Compaction::~Compaction() {
@@ -204,11 +198,9 @@ Compaction::~Compaction() {
}
bool Compaction::InputCompressionMatchesOutput() const {
VersionStorageInfo* vstorage = input_version_->storage_info();
int base_level = vstorage->base_level();
bool matches =
(GetCompressionType(*cfd_->ioptions(), vstorage, mutable_cf_options_,
start_level_, base_level) == output_compression_);
int base_level = input_version_->storage_info()->base_level();
bool matches = (GetCompressionType(*cfd_->ioptions(), start_level_,
base_level) == output_compression_);
if (matches) {
TEST_SYNC_POINT("Compaction::InputCompressionMatchesOutput:Matches");
return true;
@@ -249,7 +241,7 @@ bool Compaction::IsTrivialMove() const {
return (start_level_ != output_level_ && num_input_levels() == 1 &&
input(0, 0)->fd.GetPathId() == output_path_id() &&
InputCompressionMatchesOutput() &&
TotalFileSize(grandparents_) <= max_compaction_bytes_);
TotalFileSize(grandparents_) <= max_grandparent_overlap_bytes_);
}
void Compaction::AddInputDeletions(VersionEdit* out_edit) {
@@ -290,12 +282,38 @@ bool Compaction::KeyNotExistsBeyondOutputLevel(
return true;
}
bool Compaction::ShouldStopBefore(const Slice& internal_key) {
// Scan to find earliest grandparent file that contains key.
const InternalKeyComparator* icmp = &cfd_->internal_comparator();
while (grandparent_index_ < grandparents_.size() &&
icmp->Compare(internal_key,
grandparents_[grandparent_index_]->largest.Encode()) > 0) {
if (seen_key_) {
overlapped_bytes_ += grandparents_[grandparent_index_]->fd.GetFileSize();
}
assert(grandparent_index_ + 1 >= grandparents_.size() ||
icmp->Compare(grandparents_[grandparent_index_]->largest.Encode(),
grandparents_[grandparent_index_+1]->smallest.Encode())
< 0);
grandparent_index_++;
}
seen_key_ = true;
if (overlapped_bytes_ > max_grandparent_overlap_bytes_) {
// Too much overlap for current output; start new output
overlapped_bytes_ = 0;
return true;
} else {
return false;
}
}
// Mark (or clear) each file that is being compacted
void Compaction::MarkFilesBeingCompacted(bool mark_as_compacted) {
for (size_t i = 0; i < num_input_levels(); i++) {
for (size_t j = 0; j < inputs_[i].size(); j++) {
assert(mark_as_compacted ? !inputs_[i][j]->being_compacted
: inputs_[i][j]->being_compacted);
for (unsigned int j = 0; j < inputs_[i].size(); j++) {
assert(mark_as_compacted ? !inputs_[i][j]->being_compacted :
inputs_[i][j]->being_compacted);
inputs_[i][j]->being_compacted = mark_as_compacted;
}
}
@@ -353,7 +371,7 @@ int InputSummary(const std::vector<FileMetaData*>& files, char* output,
int len) {
*output = '\0';
int write = 0;
for (size_t i = 0; i < files.size(); i++) {
for (unsigned int i = 0; i < files.size(); i++) {
int sz = len - write;
int ret;
char sztxt[16];
@@ -370,8 +388,10 @@ int InputSummary(const std::vector<FileMetaData*>& files, char* output,
void Compaction::Summary(char* output, int len) {
int write =
snprintf(output, len, "Base version %" PRIu64 " Base level %d, inputs: [",
input_version_->GetVersionNumber(), start_level_);
snprintf(output, len, "Base version %" PRIu64
" Base level %d, inputs: [",
input_version_->GetVersionNumber(),
start_level_);
if (write < 0 || write >= len) {
return;
}
@@ -393,7 +413,7 @@ void Compaction::Summary(char* output, int len) {
snprintf(output + write, len - write, "]");
}
uint64_t Compaction::OutputFilePreallocationSize() const {
uint64_t Compaction::OutputFilePreallocationSize() {
uint64_t preallocation_size = 0;
if (cfd_->ioptions()->compaction_style == kCompactionStyleLevel ||
@@ -408,7 +428,7 @@ uint64_t Compaction::OutputFilePreallocationSize() const {
}
// Over-estimate slightly so we don't end up just barely crossing
// the threshold
return preallocation_size + (preallocation_size / 10);
return preallocation_size * 1.1;
}
std::unique_ptr<CompactionFilter> Compaction::CreateCompactionFilter() const {
@@ -429,7 +449,7 @@ bool Compaction::IsOutputLevelEmpty() const {
}
bool Compaction::ShouldFormSubcompactions() const {
if (immutable_cf_options_.max_subcompactions <= 1 || cfd_ == nullptr) {
if (mutable_cf_options_.max_subcompactions <= 1 || cfd_ == nullptr) {
return false;
}
if (cfd_->ioptions()->compaction_style == kCompactionStyleLevel) {
+23 -51
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -8,10 +8,10 @@
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#pragma once
#include "db/version_set.h"
#include "util/arena.h"
#include "util/autovector.h"
#include "util/cf_options.h"
#include "util/mutable_cf_options.h"
#include "db/version_set.h"
namespace rocksdb {
@@ -35,15 +35,13 @@ class CompactionFilter;
class Compaction {
public:
Compaction(VersionStorageInfo* input_version,
const ImmutableCFOptions& immutable_cf_options,
const MutableCFOptions& mutable_cf_options,
std::vector<CompactionInputFiles> inputs, int output_level,
uint64_t target_file_size, uint64_t max_compaction_bytes,
uint64_t target_file_size, uint64_t max_grandparent_overlap_bytes,
uint32_t output_path_id, CompressionType compression,
std::vector<FileMetaData*> grandparents,
bool manual_compaction = false, double score = -1,
bool deletion_compaction = false,
CompactionReason compaction_reason = CompactionReason::kUnknown);
bool deletion_compaction = false);
// No copying allowed
Compaction(const Compaction&) = delete;
@@ -104,10 +102,8 @@ class Compaction {
return &inputs_[compaction_input_level].files;
}
const std::vector<CompactionInputFiles>* inputs() { return &inputs_; }
// Returns the LevelFilesBrief of the specified compaction input level.
const LevelFilesBrief* input_levels(size_t compaction_input_level) const {
LevelFilesBrief* input_levels(size_t compaction_input_level) {
return &input_levels_[compaction_input_level];
}
@@ -135,10 +131,12 @@ class Compaction {
bool KeyNotExistsBeyondOutputLevel(const Slice& user_key,
std::vector<size_t>* level_ptrs) const;
// Returns true iff we should stop building the current output
// before processing "internal_key".
bool ShouldStopBefore(const Slice& internal_key);
// Clear all files to indicate that they are not being compacted
// Delete this compaction from the list of running compactions.
//
// Requirement: DB mutex held
void ReleaseCompactionFiles(Status status);
// Returns the summary of the compaction in "output" with maximum "len"
@@ -150,13 +148,13 @@ class Compaction {
double score() const { return score_; }
// Is this compaction creating a file in the bottom most level?
bool bottommost_level() const { return bottommost_level_; }
bool bottommost_level() { return bottommost_level_; }
// Does this compaction include all sst files?
bool is_full_compaction() const { return is_full_compaction_; }
bool is_full_compaction() { return is_full_compaction_; }
// Was this compaction triggered manually by the client?
bool is_manual_compaction() const { return is_manual_compaction_; }
bool is_manual_compaction() { return is_manual_compaction_; }
// Used when allow_trivial_move option is set in
// Universal compaction. If all the input files are
@@ -169,27 +167,19 @@ class Compaction {
// Used when allow_trivial_move option is set in
// Universal compaction. Returns true, if the input files
// are non-overlapping and can be trivially moved.
bool is_trivial_move() const { return is_trivial_move_; }
bool is_trivial_move() { return is_trivial_move_; }
// How many total levels are there?
int number_levels() const { return number_levels_; }
// Return the ImmutableCFOptions that should be used throughout the compaction
// procedure
const ImmutableCFOptions* immutable_cf_options() const {
return &immutable_cf_options_;
}
// Return the MutableCFOptions that should be used throughout the compaction
// procedure
const MutableCFOptions* mutable_cf_options() const {
return &mutable_cf_options_;
}
const MutableCFOptions* mutable_cf_options() { return &mutable_cf_options_; }
// Returns the size in bytes that the output file should be preallocated to.
// In level compaction, that is max_file_size_. In universal compaction, that
// is the sum of all input file sizes.
uint64_t OutputFilePreallocationSize() const;
uint64_t OutputFilePreallocationSize();
void SetInputVersion(Version* input_version);
@@ -228,18 +218,6 @@ class Compaction {
output_table_properties_ = std::move(tp);
}
Slice GetSmallestUserKey() const { return smallest_user_key_; }
Slice GetLargestUserKey() const { return largest_user_key_; }
CompactionReason compaction_reason() { return compaction_reason_; }
const std::vector<FileMetaData*>& grandparents() const {
return grandparents_;
}
uint64_t max_compaction_bytes() const { return max_compaction_bytes_; }
private:
// mark (or clear) all files that are being compacted
void MarkFilesBeingCompacted(bool mark_as_compacted);
@@ -261,9 +239,8 @@ class Compaction {
const int start_level_; // the lowest level to be compacted
const int output_level_; // levels to which output files are stored
uint64_t max_output_file_size_;
uint64_t max_compaction_bytes_;
const ImmutableCFOptions immutable_cf_options_;
const MutableCFOptions mutable_cf_options_;
uint64_t max_grandparent_overlap_bytes_;
MutableCFOptions mutable_cf_options_;
Version* input_version_;
VersionEdit edit_;
const int number_levels_;
@@ -281,9 +258,13 @@ class Compaction {
// A copy of inputs_, organized more closely in memory
autovector<LevelFilesBrief, 2> input_levels_;
// State used to check for number of overlapping grandparent files
// State used to check for number of of overlapping grandparent files
// (grandparent == "output_level_ + 1")
std::vector<FileMetaData*> grandparents_;
size_t grandparent_index_; // Index in grandparent_starts_
bool seen_key_; // Some output key has been seen
uint64_t overlapped_bytes_; // Bytes of overlap between current output
// and grandparent files
const double score_; // score that was used to pick this compaction.
// Is this compaction creating a file in the bottom most level?
@@ -303,15 +284,6 @@ class Compaction {
// table properties of output files
TablePropertiesCollection output_table_properties_;
// smallest user keys in compaction
Slice smallest_user_key_;
// largest user keys in compaction
Slice largest_user_key_;
// Reason for compaction
CompactionReason compaction_reason_;
};
// Utility function
-33
View File
@@ -1,33 +0,0 @@
// Copyright (c) 2016-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
#pragma once
struct CompactionIterationStats {
// Compaction statistics
// Doesn't include records skipped because of
// CompactionFilter::Decision::kRemoveAndSkipUntil.
int64_t num_record_drop_user = 0;
int64_t num_record_drop_hidden = 0;
int64_t num_record_drop_obsolete = 0;
int64_t num_record_drop_range_del = 0;
int64_t num_range_del_drop_obsolete = 0;
uint64_t total_filter_time = 0;
// Input statistics
// TODO(noetzli): The stats are incomplete. They are lacking everything
// consumed by MergeHelper.
uint64_t num_input_records = 0;
uint64_t num_input_deletion_records = 0;
uint64_t num_input_corrupt_records = 0;
uint64_t total_input_raw_key_bytes = 0;
uint64_t total_input_raw_value_bytes = 0;
// Single-Delete diagnostics for exceptional situations
uint64_t num_single_del_fallthru = 0;
uint64_t num_single_del_mismatch = 0;
};
+42 -214
View File
@@ -1,6 +1,6 @@
// 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.
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -13,34 +13,15 @@ namespace rocksdb {
CompactionIterator::CompactionIterator(
InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
SequenceNumber last_sequence, std::vector<SequenceNumber>* snapshots,
SequenceNumber earliest_write_conflict_snapshot, Env* env,
bool expect_valid_internal_key, RangeDelAggregator* range_del_agg,
const Compaction* compaction, const CompactionFilter* compaction_filter,
LogBuffer* log_buffer)
: CompactionIterator(
input, cmp, merge_helper, last_sequence, snapshots,
earliest_write_conflict_snapshot, env, expect_valid_internal_key,
range_del_agg,
std::unique_ptr<CompactionProxy>(
compaction ? new CompactionProxy(compaction) : nullptr),
compaction_filter, log_buffer) {}
CompactionIterator::CompactionIterator(
InternalIterator* input, const Comparator* cmp, MergeHelper* merge_helper,
SequenceNumber last_sequence, std::vector<SequenceNumber>* snapshots,
SequenceNumber earliest_write_conflict_snapshot, Env* env,
bool expect_valid_internal_key, RangeDelAggregator* range_del_agg,
std::unique_ptr<CompactionProxy> compaction,
Env* env, bool expect_valid_internal_key, Compaction* compaction,
const CompactionFilter* compaction_filter, LogBuffer* log_buffer)
: input_(input),
cmp_(cmp),
merge_helper_(merge_helper),
snapshots_(snapshots),
earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
env_(env),
expect_valid_internal_key_(expect_valid_internal_key),
range_del_agg_(range_del_agg),
compaction_(std::move(compaction)),
compaction_(compaction),
compaction_filter_(compaction_filter),
log_buffer_(log_buffer),
merge_out_iter_(merge_helper_) {
@@ -53,11 +34,11 @@ CompactionIterator::CompactionIterator(
if (snapshots_->size() == 0) {
// optimize for fast path if there are no snapshots
visible_at_tip_ = true;
earliest_snapshot_ = last_sequence;
visible_at_tip_ = last_sequence;
earliest_snapshot_ = visible_at_tip_;
latest_snapshot_ = 0;
} else {
visible_at_tip_ = false;
visible_at_tip_ = 0;
earliest_snapshot_ = snapshots_->at(0);
latest_snapshot_ = snapshots_->back();
}
@@ -66,20 +47,12 @@ CompactionIterator::CompactionIterator(
} else {
ignore_snapshots_ = false;
}
input_->SetPinnedItersMgr(&pinned_iters_mgr_);
}
CompactionIterator::~CompactionIterator() {
// input_ Iteartor lifetime is longer than pinned_iters_mgr_ lifetime
input_->SetPinnedItersMgr(nullptr);
}
void CompactionIterator::ResetRecordCounts() {
iter_stats_.num_record_drop_user = 0;
iter_stats_.num_record_drop_hidden = 0;
iter_stats_.num_record_drop_obsolete = 0;
iter_stats_.num_record_drop_range_del = 0;
iter_stats_.num_range_del_drop_obsolete = 0;
}
void CompactionIterator::SeekToFirst() {
@@ -108,8 +81,6 @@ void CompactionIterator::Next() {
ikey_.user_key = current_key_.GetUserKey();
valid_ = true;
} else {
// We consumed all pinned merge operands, release pinned iterators
pinned_iters_mgr_.ReleasePinnedData();
// MergeHelper moves the iterator to the first record after the merged
// records, so even though we reached the end of the merge output, we do
// not want to advance the iterator.
@@ -124,11 +95,6 @@ void CompactionIterator::Next() {
NextFromInput();
}
if (valid_) {
// Record that we've outputted a record for the current key.
has_outputted_key_ = true;
}
PrepareOutput();
}
@@ -166,27 +132,17 @@ void CompactionIterator::NextFromInput() {
iter_stats_.total_input_raw_key_bytes += key_.size();
iter_stats_.total_input_raw_value_bytes += value_.size();
// If need_skip is true, we should seek the input iterator
// to internal key skip_until and continue from there.
bool need_skip = false;
// Points either into compaction_filter_skip_until_ or into
// merge_helper_->compaction_filter_skip_until_.
Slice skip_until;
// Check whether the user key changed. After this if statement current_key_
// is a copy of the current input key (maybe converted to a delete by the
// compaction filter). ikey_.user_key is pointing to the copy.
if (!has_current_user_key_ ||
!cmp_->Equal(ikey_.user_key, current_user_key_)) {
// First occurrence of this user key
// Copy key for output
key_ = current_key_.SetKey(key_, &ikey_);
current_user_key_ = ikey_.user_key;
has_current_user_key_ = true;
has_outputted_key_ = false;
current_user_key_sequence_ = kMaxSequenceNumber;
current_user_key_snapshot_ = 0;
// apply the compaction filter to the first occurrence of the user key
if (compaction_filter_ != nullptr && ikey_.type == kTypeValue &&
(visible_at_tip_ || ikey_.sequence > latest_snapshot_ ||
@@ -195,41 +151,26 @@ void CompactionIterator::NextFromInput() {
// number is greater than any external snapshot, then invoke the
// filter. If the return value of the compaction filter is true,
// replace the entry with a deletion marker.
CompactionFilter::Decision filter;
bool value_changed = false;
bool to_delete = false;
compaction_filter_value_.clear();
compaction_filter_skip_until_.Clear();
{
StopWatchNano timer(env_, true);
filter = compaction_filter_->FilterV2(
compaction_->level(), ikey_.user_key,
CompactionFilter::ValueType::kValue, value_,
&compaction_filter_value_, compaction_filter_skip_until_.rep());
to_delete = compaction_filter_->Filter(
compaction_->level(), ikey_.user_key, value_,
&compaction_filter_value_, &value_changed);
iter_stats_.total_filter_time +=
env_ != nullptr ? timer.ElapsedNanos() : 0;
}
if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil &&
cmp_->Compare(*compaction_filter_skip_until_.rep(),
ikey_.user_key) <= 0) {
// Can't skip to a key smaller than the current one.
// Keep the key as per FilterV2 documentation.
filter = CompactionFilter::Decision::kKeep;
}
if (filter == CompactionFilter::Decision::kRemove) {
if (to_delete) {
// convert the current key to a delete
ikey_.type = kTypeDeletion;
current_key_.UpdateInternalKey(ikey_.sequence, kTypeDeletion);
// no value associated with delete
value_.clear();
iter_stats_.num_record_drop_user++;
} else if (filter == CompactionFilter::Decision::kChangeValue) {
} else if (value_changed) {
value_ = compaction_filter_value_;
} else if (filter == CompactionFilter::Decision::kRemoveAndSkipUntil) {
need_skip = true;
compaction_filter_skip_until_.ConvertFromUserKey(kMaxSequenceNumber,
kValueTypeForSeek);
skip_until = compaction_filter_skip_until_.Encode();
}
}
} else {
@@ -252,111 +193,34 @@ void CompactionIterator::NextFromInput() {
SequenceNumber last_snapshot = current_user_key_snapshot_;
SequenceNumber prev_snapshot = 0; // 0 means no previous snapshot
current_user_key_snapshot_ =
visible_at_tip_
? earliest_snapshot_
: findEarliestVisibleSnapshot(ikey_.sequence, &prev_snapshot);
visible_at_tip_ ? visible_at_tip_ : findEarliestVisibleSnapshot(
ikey_.sequence, &prev_snapshot);
if (need_skip) {
// This case is handled below.
} else if (clear_and_output_next_key_) {
// In the previous iteration we encountered a single delete that we could
// not compact out. We will keep this Put, but can drop it's data.
// (See Optimization 3, below.)
assert(ikey_.type == kTypeValue);
assert(current_user_key_snapshot_ == last_snapshot);
value_.clear();
valid_ = true;
clear_and_output_next_key_ = false;
} else if (ikey_.type == kTypeSingleDeletion) {
// We can compact out a SingleDelete if:
// 1) We encounter the corresponding PUT -OR- we know that this key
// doesn't appear past this output level
// =AND=
// 2) We've already returned a record in this snapshot -OR-
// there are no earlier earliest_write_conflict_snapshot.
//
// Rule 1 is needed for SingleDelete correctness. Rule 2 is needed to
// allow Transactions to do write-conflict checking (if we compacted away
// all keys, then we wouldn't know that a write happened in this
// snapshot). If there is no earlier snapshot, then we know that there
// are no active transactions that need to know about any writes.
//
// Optimization 3:
// If we encounter a SingleDelete followed by a PUT and Rule 2 is NOT
// true, then we must output a SingleDelete. In this case, we will decide
// to also output the PUT. While we are compacting less by outputting the
// PUT now, hopefully this will lead to better compaction in the future
// when Rule 2 is later true (Ie, We are hoping we can later compact out
// both the SingleDelete and the Put, while we couldn't if we only
// outputted the SingleDelete now).
// In this case, we can save space by removing the PUT's value as it will
// never be read.
//
// Deletes and Merges are not supported on the same key that has a
// SingleDelete as it is not possible to correctly do any partial
// compaction of such a combination of operations. The result of mixing
// those operations for a given key is documented as being undefined. So
// we can choose how to handle such a combinations of operations. We will
// try to compact out as much as we can in these cases.
// We will report counts on these anomalous cases.
// The easiest way to process a SingleDelete during iteration is to peek
// ahead at the next key.
if (ikey_.type == kTypeSingleDeletion) {
ParsedInternalKey next_ikey;
input_->Next();
// Check whether the next key exists, is not corrupt, and is the same key
// Check whether the current key is valid, not corrupt and the same
// as the single delete.
if (input_->Valid() && ParseInternalKey(input_->key(), &next_ikey) &&
cmp_->Equal(ikey_.user_key, next_ikey.user_key)) {
// Check whether the next key belongs to the same snapshot as the
// SingleDelete.
// Mixing single deletes and merges is not supported. Consecutive
// single deletes are not valid.
if (next_ikey.type != kTypeValue) {
assert(false);
status_ =
Status::InvalidArgument("Put expected after single delete.");
break;
}
// Check whether the current key belongs to the same snapshot as the
// single delete.
if (prev_snapshot == 0 || next_ikey.sequence > prev_snapshot) {
if (next_ikey.type == kTypeSingleDeletion) {
// We encountered two SingleDeletes in a row. This could be due to
// unexpected user input.
// Skip the first SingleDelete and let the next iteration decide how
// to handle the second SingleDelete
// First SingleDelete has been skipped since we already called
// input_->Next().
++iter_stats_.num_record_drop_obsolete;
++iter_stats_.num_single_del_mismatch;
} else if ((ikey_.sequence <= earliest_write_conflict_snapshot_) ||
has_outputted_key_) {
// Found a matching value, we can drop the single delete and the
// value. It is safe to drop both records since we've already
// outputted a key in this snapshot, or there is no earlier
// snapshot (Rule 2 above).
// Note: it doesn't matter whether the second key is a Put or if it
// is an unexpected Merge or Delete. We will compact it out
// either way. We will maintain counts of how many mismatches
// happened
if (next_ikey.type != kTypeValue) {
++iter_stats_.num_single_del_mismatch;
}
++iter_stats_.num_record_drop_hidden;
++iter_stats_.num_record_drop_obsolete;
// Already called input_->Next() once. Call it a second time to
// skip past the second key.
input_->Next();
} else {
// Found a matching value, but we cannot drop both keys since
// there is an earlier snapshot and we need to leave behind a record
// to know that a write happened in this snapshot (Rule 2 above).
// Clear the value and output the SingleDelete. (The value will be
// outputted on the next iteration.)
// Setting valid_ to true will output the current SingleDelete
valid_ = true;
// Set up the Put to be outputted in the next iteration.
// (Optimization 3).
clear_and_output_next_key_ = true;
}
// Found the matching value, we can drop the single delete and the
// value.
++iter_stats_.num_record_drop_hidden;
++iter_stats_.num_record_drop_obsolete;
input_->Next();
} else {
// We hit the next snapshot without hitting a put, so the iterator
// returns the single delete.
@@ -364,22 +228,18 @@ void CompactionIterator::NextFromInput() {
}
} else {
// We are at the end of the input, could not parse the next key, or hit
// a different key. The iterator returns the single delete if the key
// the next key. The iterator returns the single delete if the key
// possibly exists beyond the current output level. We set
// has_current_user_key to false so that if the iterator is at the next
// key, we do not compare it again against the previous key at the next
// iteration. If the next key is corrupt, we return before the
// comparison, so the value of has_current_user_key does not matter.
has_current_user_key_ = false;
if (compaction_ != nullptr && ikey_.sequence <= earliest_snapshot_ &&
if (compaction_ != nullptr &&
compaction_->KeyNotExistsBeyondOutputLevel(ikey_.user_key,
&level_ptrs_)) {
// Key doesn't exist outside of this range.
// Can compact out this SingleDelete.
++iter_stats_.num_record_drop_obsolete;
++iter_stats_.num_single_del_fallthru;
} else {
// Output SingleDelete
valid_ = true;
}
}
@@ -393,10 +253,6 @@ void CompactionIterator::NextFromInput() {
// same key, then this kv is not visible in any snapshot.
// Hidden by an newer entry for same user key
// TODO: why not > ?
//
// Note: Dropping this key will not affect TransactionDB write-conflict
// checking since there has already been a record returned for this key
// in this snapshot.
assert(last_sequence >= current_user_key_sequence_);
++iter_stats_.num_record_drop_hidden; // (A)
input_->Next();
@@ -415,9 +271,6 @@ void CompactionIterator::NextFromInput() {
// smaller sequence numbers will be dropped in the next
// few iterations of this loop (by rule (A) above).
// Therefore this deletion marker is obsolete and can be dropped.
//
// Note: Dropping this Delete will not affect TransactionDB
// write-conflict checking since it is earlier than any snapshot.
++iter_stats_.num_record_drop_obsolete;
input_->Next();
} else if (ikey_.type == kTypeMerge) {
@@ -428,13 +281,11 @@ void CompactionIterator::NextFromInput() {
return;
}
pinned_iters_mgr_.StartPinning();
// We know the merge type entry is not hidden, otherwise we would
// have hit (A)
// We encapsulate the merge related state machine in a different
// object to minimize change to the existing flow.
merge_helper_->MergeUntil(input_, range_del_agg_, prev_snapshot,
bottommost_level_);
merge_helper_->MergeUntil(input_, prev_snapshot, bottommost_level_);
merge_out_iter_.SeekToFirst();
if (merge_out_iter_.Valid()) {
@@ -457,27 +308,9 @@ void CompactionIterator::NextFromInput() {
// batch consumed by the merge operator should not shadow any keys
// coming after the merges
has_current_user_key_ = false;
pinned_iters_mgr_.ReleasePinnedData();
if (merge_helper_->FilteredUntil(&skip_until)) {
need_skip = true;
}
}
} else {
// 1. new user key -OR-
// 2. different snapshot stripe
bool should_delete = range_del_agg_->ShouldDelete(key_);
if (should_delete) {
++iter_stats_.num_record_drop_hidden;
++iter_stats_.num_record_drop_range_del;
input_->Next();
} else {
valid_ = true;
}
}
if (need_skip) {
input_->Seek(skip_until);
valid_ = true;
}
}
}
@@ -486,14 +319,9 @@ void CompactionIterator::PrepareOutput() {
// Zeroing out the sequence number leads to better compression.
// If this is the bottommost level (no files in lower levels)
// and the earliest snapshot is larger than this seqno
// and the userkey differs from the last userkey in compaction
// then we can squash the seqno to zero.
// This is safe for TransactionDB write-conflict checking since transactions
// only care about sequence number larger than any active snapshots.
if (bottommost_level_ && valid_ && ikey_.sequence < earliest_snapshot_ &&
ikey_.type != kTypeMerge &&
!cmp_->Equal(compaction_->GetLargestUserKey(), ikey_.user_key)) {
ikey_.type != kTypeMerge) {
assert(ikey_.type != kTypeDeletion && ikey_.type != kTypeSingleDeletion);
ikey_.sequence = 0;
current_key_.UpdateInternalKey(0, ikey_.type);
@@ -503,15 +331,15 @@ void CompactionIterator::PrepareOutput() {
inline SequenceNumber CompactionIterator::findEarliestVisibleSnapshot(
SequenceNumber in, SequenceNumber* prev_snapshot) {
assert(snapshots_->size());
SequenceNumber prev __attribute__((__unused__)) = kMaxSequenceNumber;
SequenceNumber prev __attribute__((unused)) = 0;
for (const auto cur : *snapshots_) {
assert(prev == kMaxSequenceNumber || prev <= cur);
assert(prev <= cur);
if (cur >= in) {
*prev_snapshot = prev == kMaxSequenceNumber ? 0 : prev;
*prev_snapshot = prev;
return cur;
}
prev = cur;
assert(prev < kMaxSequenceNumber);
assert(prev);
}
*prev_snapshot = prev;
return kMaxSequenceNumber;
+24 -69
View File
@@ -1,6 +1,6 @@
// 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.
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -12,70 +12,39 @@
#include <vector>
#include "db/compaction.h"
#include "db/compaction_iteration_stats.h"
#include "db/merge_helper.h"
#include "db/pinned_iterators_manager.h"
#include "db/range_del_aggregator.h"
#include "rocksdb/compaction_filter.h"
#include "util/log_buffer.h"
namespace rocksdb {
struct CompactionIteratorStats {
// Compaction statistics
int64_t num_record_drop_user = 0;
int64_t num_record_drop_hidden = 0;
int64_t num_record_drop_obsolete = 0;
uint64_t total_filter_time = 0;
// Input statistics
// TODO(noetzli): The stats are incomplete. They are lacking everything
// consumed by MergeHelper.
uint64_t num_input_records = 0;
uint64_t num_input_deletion_records = 0;
uint64_t num_input_corrupt_records = 0;
uint64_t total_input_raw_key_bytes = 0;
uint64_t total_input_raw_value_bytes = 0;
};
class CompactionIterator {
public:
// A wrapper around Compaction. Has a much smaller interface, only what
// CompactionIterator uses. Tests can override it.
class CompactionProxy {
public:
explicit CompactionProxy(const Compaction* compaction)
: compaction_(compaction) {}
virtual ~CompactionProxy() = default;
virtual int level(size_t compaction_input_level = 0) const {
return compaction_->level();
}
virtual bool KeyNotExistsBeyondOutputLevel(
const Slice& user_key, std::vector<size_t>* level_ptrs) const {
return compaction_->KeyNotExistsBeyondOutputLevel(user_key, level_ptrs);
}
virtual bool bottommost_level() const {
return compaction_->bottommost_level();
}
virtual int number_levels() const { return compaction_->number_levels(); }
virtual Slice GetLargestUserKey() const {
return compaction_->GetLargestUserKey();
}
protected:
CompactionProxy() = default;
private:
const Compaction* compaction_;
};
CompactionIterator(InternalIterator* input, const Comparator* cmp,
MergeHelper* merge_helper, SequenceNumber last_sequence,
std::vector<SequenceNumber>* snapshots,
SequenceNumber earliest_write_conflict_snapshot, Env* env,
std::vector<SequenceNumber>* snapshots, Env* env,
bool expect_valid_internal_key,
RangeDelAggregator* range_del_agg,
const Compaction* compaction = nullptr,
Compaction* compaction = nullptr,
const CompactionFilter* compaction_filter = nullptr,
LogBuffer* log_buffer = nullptr);
// Constructor with custom CompactionProxy, used for tests.
CompactionIterator(InternalIterator* input, const Comparator* cmp,
MergeHelper* merge_helper, SequenceNumber last_sequence,
std::vector<SequenceNumber>* snapshots,
SequenceNumber earliest_write_conflict_snapshot, Env* env,
bool expect_valid_internal_key,
RangeDelAggregator* range_del_agg,
std::unique_ptr<CompactionProxy> compaction,
const CompactionFilter* compaction_filter = nullptr,
LogBuffer* log_buffer = nullptr);
~CompactionIterator();
void ResetRecordCounts();
// Seek to the beginning of the compaction iterator output.
@@ -95,7 +64,7 @@ class CompactionIterator {
const ParsedInternalKey& ikey() const { return ikey_; }
bool Valid() const { return valid_; }
const Slice& user_key() const { return current_user_key_; }
const CompactionIterationStats& iter_stats() const { return iter_stats_; }
const CompactionIteratorStats& iter_stats() const { return iter_stats_; }
private:
// Processes the input stream to find the next output
@@ -119,16 +88,14 @@ class CompactionIterator {
const Comparator* cmp_;
MergeHelper* merge_helper_;
const std::vector<SequenceNumber>* snapshots_;
const SequenceNumber earliest_write_conflict_snapshot_;
Env* env_;
bool expect_valid_internal_key_;
RangeDelAggregator* range_del_agg_;
std::unique_ptr<CompactionProxy> compaction_;
Compaction* compaction_;
const CompactionFilter* compaction_filter_;
LogBuffer* log_buffer_;
bool bottommost_level_;
bool valid_ = false;
bool visible_at_tip_;
SequenceNumber visible_at_tip_;
SequenceNumber earliest_snapshot_;
SequenceNumber latest_snapshot_;
bool ignore_snapshots_;
@@ -158,20 +125,8 @@ class CompactionIterator {
Slice current_user_key_;
SequenceNumber current_user_key_sequence_;
SequenceNumber current_user_key_snapshot_;
// True if the iterator has already returned a record for the current key.
bool has_outputted_key_ = false;
// truncated the value of the next key and output it without applying any
// compaction rules. This is used for outputting a put after a single delete.
bool clear_and_output_next_key_ = false;
MergeOutputIterator merge_out_iter_;
// PinnedIteratorsManager used to pin input_ Iterator blocks while reading
// merge operands and then releasing them after consuming them.
PinnedIteratorsManager pinned_iters_mgr_;
std::string compaction_filter_value_;
InternalKey compaction_filter_skip_until_;
// "level_ptrs" holds indices that remember which file of an associated
// level we were last checking during the last call to compaction->
// KeyNotExistsBeyondOutputLevel(). This allows future calls to the function
@@ -179,6 +134,6 @@ class CompactionIterator {
// increasing so a later call to the function must be looking for a key that
// is in or beyond the last file checked during the previous call
std::vector<size_t> level_ptrs_;
CompactionIterationStats iter_stats_;
CompactionIteratorStats iter_stats_;
};
} // namespace rocksdb
+18 -293
View File
@@ -1,151 +1,43 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
#include "db/compaction_iterator.h"
#include <string>
#include <vector>
#include "util/testharness.h"
#include "util/testutil.h"
namespace rocksdb {
class LoggingForwardVectorIterator : public InternalIterator {
public:
struct Action {
enum class Type {
SEEK_TO_FIRST,
SEEK,
NEXT,
};
Type type;
std::string arg;
explicit Action(Type _type, std::string _arg = "")
: type(_type), arg(_arg) {}
bool operator==(const Action& rhs) const {
return std::tie(type, arg) == std::tie(rhs.type, rhs.arg);
}
};
LoggingForwardVectorIterator(const std::vector<std::string>& keys,
const std::vector<std::string>& values)
: keys_(keys), values_(values), current_(keys.size()) {
assert(keys_.size() == values_.size());
}
virtual bool Valid() const override { return current_ < keys_.size(); }
virtual void SeekToFirst() override {
log.emplace_back(Action::Type::SEEK_TO_FIRST);
current_ = 0;
}
virtual void SeekToLast() override { assert(false); }
virtual void Seek(const Slice& target) override {
log.emplace_back(Action::Type::SEEK, target.ToString());
current_ = std::lower_bound(keys_.begin(), keys_.end(), target.ToString()) -
keys_.begin();
}
virtual void SeekForPrev(const Slice& target) override { assert(false); }
virtual void Next() override {
assert(Valid());
log.emplace_back(Action::Type::NEXT);
current_++;
}
virtual void Prev() override { assert(false); }
virtual Slice key() const override {
assert(Valid());
return Slice(keys_[current_]);
}
virtual Slice value() const override {
assert(Valid());
return Slice(values_[current_]);
}
virtual Status status() const override { return Status::OK(); }
std::vector<Action> log;
private:
std::vector<std::string> keys_;
std::vector<std::string> values_;
size_t current_;
};
class FakeCompaction : public CompactionIterator::CompactionProxy {
public:
FakeCompaction() = default;
virtual int level(size_t compaction_input_level) const { return 0; }
virtual bool KeyNotExistsBeyondOutputLevel(
const Slice& user_key, std::vector<size_t>* level_ptrs) const {
return false;
}
virtual bool bottommost_level() const { return false; }
virtual int number_levels() const { return 1; }
virtual Slice GetLargestUserKey() const {
return "\xff\xff\xff\xff\xff\xff\xff\xff\xff";
}
};
class CompactionIteratorTest : public testing::Test {
public:
CompactionIteratorTest()
: cmp_(BytewiseComparator()), icmp_(cmp_), snapshots_({}) {}
CompactionIteratorTest() : cmp_(BytewiseComparator()), snapshots_({}) {}
void InitIterators(const std::vector<std::string>& ks,
const std::vector<std::string>& vs,
const std::vector<std::string>& range_del_ks,
const std::vector<std::string>& range_del_vs,
SequenceNumber last_sequence,
MergeOperator* merge_op = nullptr,
CompactionFilter* filter = nullptr) {
std::unique_ptr<InternalIterator> range_del_iter(
new test::VectorIterator(range_del_ks, range_del_vs));
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) {
compaction.reset(new FakeCompaction());
}
merge_helper_.reset(new MergeHelper(Env::Default(), cmp_, merge_op, filter,
void InitIterator(const std::vector<std::string>& ks,
const std::vector<std::string>& vs,
SequenceNumber last_sequence) {
merge_helper_.reset(new MergeHelper(Env::Default(), cmp_, nullptr, nullptr,
nullptr, 0U, false, 0));
iter_.reset(new LoggingForwardVectorIterator(ks, vs));
iter_.reset(new test::VectorIterator(ks, vs));
iter_->SeekToFirst();
c_iter_.reset(new CompactionIterator(
iter_.get(), cmp_, merge_helper_.get(), last_sequence, &snapshots_,
kMaxSequenceNumber, Env::Default(), false, range_del_agg_.get(),
std::move(compaction), filter));
c_iter_.reset(new CompactionIterator(iter_.get(), cmp_, merge_helper_.get(),
last_sequence, &snapshots_,
Env::Default(), false));
}
void AddSnapshot(SequenceNumber snapshot) { snapshots_.push_back(snapshot); }
const Comparator* cmp_;
const InternalKeyComparator icmp_;
std::vector<SequenceNumber> snapshots_;
std::unique_ptr<MergeHelper> merge_helper_;
std::unique_ptr<LoggingForwardVectorIterator> iter_;
std::unique_ptr<test::VectorIterator> iter_;
std::unique_ptr<CompactionIterator> c_iter_;
std::unique_ptr<RangeDelAggregator> range_del_agg_;
};
// It is possible that the output of the compaction iterator is empty even if
// the input is not.
TEST_F(CompactionIteratorTest, EmptyResult) {
InitIterators({test::KeyStr("a", 5, kTypeSingleDeletion),
test::KeyStr("a", 3, kTypeValue)},
{"", "val"}, {}, {}, 5);
InitIterator({test::KeyStr("a", 5, kTypeSingleDeletion),
test::KeyStr("a", 3, kTypeValue)},
{"", "val"}, 5);
c_iter_->SeekToFirst();
ASSERT_FALSE(c_iter_->Valid());
}
@@ -153,10 +45,10 @@ TEST_F(CompactionIteratorTest, EmptyResult) {
// If there is a corruption after a single deletion, the corrupted key should
// be preserved.
TEST_F(CompactionIteratorTest, CorruptionAfterSingleDeletion) {
InitIterators({test::KeyStr("a", 5, kTypeSingleDeletion),
test::KeyStr("a", 3, kTypeValue, true),
test::KeyStr("b", 10, kTypeValue)},
{"", "val", "val2"}, {}, {}, 10);
InitIterator({test::KeyStr("a", 5, kTypeSingleDeletion),
test::KeyStr("a", 3, kTypeValue, true),
test::KeyStr("b", 10, kTypeValue)},
{"", "val", "val2"}, 10);
c_iter_->SeekToFirst();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("a", 5, kTypeSingleDeletion),
@@ -171,173 +63,6 @@ TEST_F(CompactionIteratorTest, CorruptionAfterSingleDeletion) {
ASSERT_FALSE(c_iter_->Valid());
}
TEST_F(CompactionIteratorTest, SimpleRangeDeletion) {
InitIterators({test::KeyStr("morning", 5, kTypeValue),
test::KeyStr("morning", 2, kTypeValue),
test::KeyStr("night", 3, kTypeValue)},
{"zao", "zao", "wan"},
{test::KeyStr("ma", 4, kTypeRangeDeletion)}, {"mz"}, 5);
c_iter_->SeekToFirst();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("morning", 5, kTypeValue), c_iter_->key().ToString());
c_iter_->Next();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("night", 3, kTypeValue), c_iter_->key().ToString());
c_iter_->Next();
ASSERT_FALSE(c_iter_->Valid());
}
TEST_F(CompactionIteratorTest, RangeDeletionWithSnapshots) {
AddSnapshot(10);
std::vector<std::string> ks1;
ks1.push_back(test::KeyStr("ma", 28, kTypeRangeDeletion));
std::vector<std::string> vs1{"mz"};
std::vector<std::string> ks2{test::KeyStr("morning", 15, kTypeValue),
test::KeyStr("morning", 5, kTypeValue),
test::KeyStr("night", 40, kTypeValue),
test::KeyStr("night", 20, kTypeValue)};
std::vector<std::string> vs2{"zao 15", "zao 5", "wan 40", "wan 20"};
InitIterators(ks2, vs2, ks1, vs1, 40);
c_iter_->SeekToFirst();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("morning", 5, kTypeValue), c_iter_->key().ToString());
c_iter_->Next();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("night", 40, kTypeValue), c_iter_->key().ToString());
c_iter_->Next();
ASSERT_FALSE(c_iter_->Valid());
}
TEST_F(CompactionIteratorTest, CompactionFilterSkipUntil) {
// Expect no merging attempts.
class MergeOp : public MergeOperator {
public:
bool FullMergeV2(const MergeOperationInput& merge_in,
MergeOperationOutput* merge_out) const override {
ADD_FAILURE();
return false;
}
bool PartialMergeMulti(const Slice& key,
const std::deque<Slice>& operand_list,
std::string* new_value,
Logger* logger) const override {
ADD_FAILURE();
return false;
}
const char* Name() const override {
return "CompactionIteratorTest.CompactionFilterSkipUntil::MergeOp";
}
};
class Filter : public CompactionFilter {
virtual Decision FilterV2(int level, const Slice& key, ValueType t,
const Slice& existing_value,
std::string* new_value,
std::string* skip_until) const override {
std::string k = key.ToString();
std::string v = existing_value.ToString();
// See InitIterators() call below for the sequence of keys and their
// filtering decisions. Here we closely assert that compaction filter is
// called with the expected keys and only them, and with the right values.
if (k == "a") {
EXPECT_EQ(ValueType::kValue, t);
EXPECT_EQ("av50", v);
return Decision::kKeep;
}
if (k == "b") {
EXPECT_EQ(ValueType::kValue, t);
EXPECT_EQ("bv60", v);
*skip_until = "d+";
return Decision::kRemoveAndSkipUntil;
}
if (k == "e") {
EXPECT_EQ(ValueType::kMergeOperand, t);
EXPECT_EQ("em71", v);
return Decision::kKeep;
}
if (k == "f") {
if (v == "fm65") {
EXPECT_EQ(ValueType::kMergeOperand, t);
*skip_until = "f";
} else {
EXPECT_EQ("fm30", v);
EXPECT_EQ(ValueType::kMergeOperand, t);
*skip_until = "g+";
}
return Decision::kRemoveAndSkipUntil;
}
if (k == "h") {
EXPECT_EQ(ValueType::kValue, t);
EXPECT_EQ("hv91", v);
return Decision::kKeep;
}
if (k == "i") {
EXPECT_EQ(ValueType::kMergeOperand, t);
EXPECT_EQ("im95", v);
*skip_until = "z";
return Decision::kRemoveAndSkipUntil;
}
ADD_FAILURE();
return Decision::kKeep;
}
const char* Name() const override {
return "CompactionIteratorTest.CompactionFilterSkipUntil::Filter";
}
};
MergeOp merge_op;
Filter filter;
InitIterators(
{test::KeyStr("a", 50, kTypeValue), // keep
test::KeyStr("a", 45, kTypeMerge),
test::KeyStr("b", 60, kTypeValue), // skip to "d+"
test::KeyStr("b", 40, kTypeValue), test::KeyStr("c", 35, kTypeValue),
test::KeyStr("d", 70, kTypeMerge),
test::KeyStr("e", 71, kTypeMerge), // keep
test::KeyStr("f", 65, kTypeMerge), // skip to "f", aka keep
test::KeyStr("f", 30, kTypeMerge), // skip to "g+"
test::KeyStr("f", 25, kTypeValue), test::KeyStr("g", 90, kTypeValue),
test::KeyStr("h", 91, kTypeValue), // keep
test::KeyStr("i", 95, kTypeMerge), // skip to "z"
test::KeyStr("j", 99, kTypeValue)},
{"av50", "am45", "bv60", "bv40", "cv35", "dm70", "em71", "fm65", "fm30",
"fv25", "gv90", "hv91", "im95", "jv99"},
{}, {}, kMaxSequenceNumber, &merge_op, &filter);
// Compaction should output just "a", "e" and "h" keys.
c_iter_->SeekToFirst();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("a", 50, kTypeValue), c_iter_->key().ToString());
ASSERT_EQ("av50", c_iter_->value().ToString());
c_iter_->Next();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("e", 71, kTypeMerge), c_iter_->key().ToString());
ASSERT_EQ("em71", c_iter_->value().ToString());
c_iter_->Next();
ASSERT_TRUE(c_iter_->Valid());
ASSERT_EQ(test::KeyStr("h", 91, kTypeValue), c_iter_->key().ToString());
ASSERT_EQ("hv91", c_iter_->value().ToString());
c_iter_->Next();
ASSERT_FALSE(c_iter_->Valid());
// Check that the compaction iterator did the correct sequence of calls on
// the underlying iterator.
using A = LoggingForwardVectorIterator::Action;
using T = A::Type;
std::vector<A> expected_actions = {
A(T::SEEK_TO_FIRST),
A(T::NEXT),
A(T::NEXT),
A(T::SEEK, test::KeyStr("d+", kMaxSequenceNumber, kValueTypeForSeek)),
A(T::NEXT),
A(T::NEXT),
A(T::SEEK, test::KeyStr("g+", kMaxSequenceNumber, kValueTypeForSeek)),
A(T::NEXT),
A(T::SEEK, test::KeyStr("z", kMaxSequenceNumber, kValueTypeForSeek))};
ASSERT_EQ(expected_actions, iter_->log);
}
} // namespace rocksdb
int main(int argc, char** argv) {
+85 -357
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -16,13 +16,12 @@
#include <inttypes.h>
#include <algorithm>
#include <functional>
#include <list>
#include <vector>
#include <memory>
#include <random>
#include <list>
#include <set>
#include <thread>
#include <utility>
#include <vector>
#include "db/builder.h"
#include "db/db_iter.h"
@@ -52,7 +51,6 @@
#include "util/iostats_context_imp.h"
#include "util/log_buffer.h"
#include "util/logging.h"
#include "util/sst_file_manager_impl.h"
#include "util/mutexlock.h"
#include "util/perf_context_imp.h"
#include "util/stop_watch.h"
@@ -64,7 +62,7 @@ namespace rocksdb {
// Maintains state for each sub-compaction
struct CompactionJob::SubcompactionState {
const Compaction* compaction;
Compaction* compaction;
std::unique_ptr<CompactionIterator> c_iter;
// The boundaries of the key-range this compaction is interested in. No two
@@ -99,22 +97,12 @@ struct CompactionJob::SubcompactionState {
}
}
uint64_t current_output_file_size;
// State during the subcompaction
uint64_t total_bytes;
uint64_t num_input_records;
uint64_t num_output_records;
CompactionJobStats compaction_job_stats;
uint64_t approx_size;
// An index that used to speed up ShouldStopBefore().
size_t grandparent_index = 0;
// The number of bytes overlapping between the current output and
// grandparent files used in ShouldStopBefore().
uint64_t overlapped_bytes = 0;
// A flag determine whether the key has been seen in ShouldStopBefore()
bool seen_key = false;
std::string compression_dict;
SubcompactionState(Compaction* c, Slice* _start, Slice* _end,
uint64_t size = 0)
@@ -123,15 +111,10 @@ struct CompactionJob::SubcompactionState {
end(_end),
outfile(nullptr),
builder(nullptr),
current_output_file_size(0),
total_bytes(0),
num_input_records(0),
num_output_records(0),
approx_size(size),
grandparent_index(0),
overlapped_bytes(0),
seen_key(false),
compression_dict() {
approx_size(size) {
assert(compaction != nullptr);
}
@@ -145,16 +128,11 @@ struct CompactionJob::SubcompactionState {
outputs = std::move(o.outputs);
outfile = std::move(o.outfile);
builder = std::move(o.builder);
current_output_file_size = std::move(o.current_output_file_size);
total_bytes = std::move(o.total_bytes);
num_input_records = std::move(o.num_input_records);
num_output_records = std::move(o.num_output_records);
compaction_job_stats = std::move(o.compaction_job_stats);
approx_size = std::move(o.approx_size);
grandparent_index = std::move(o.grandparent_index);
overlapped_bytes = std::move(o.overlapped_bytes);
seen_key = std::move(o.seen_key);
compression_dict = std::move(o.compression_dict);
return *this;
}
@@ -162,39 +140,6 @@ struct CompactionJob::SubcompactionState {
SubcompactionState(const SubcompactionState&) = delete;
SubcompactionState& operator=(const SubcompactionState&) = delete;
// Returns true iff we should stop building the current output
// before processing "internal_key".
bool ShouldStopBefore(const Slice& internal_key, uint64_t curr_file_size) {
const InternalKeyComparator* icmp =
&compaction->column_family_data()->internal_comparator();
const std::vector<FileMetaData*>& grandparents = compaction->grandparents();
// Scan to find earliest grandparent file that contains key.
while (grandparent_index < grandparents.size() &&
icmp->Compare(internal_key,
grandparents[grandparent_index]->largest.Encode()) >
0) {
if (seen_key) {
overlapped_bytes += grandparents[grandparent_index]->fd.GetFileSize();
}
assert(grandparent_index + 1 >= grandparents.size() ||
icmp->Compare(
grandparents[grandparent_index]->largest.Encode(),
grandparents[grandparent_index + 1]->smallest.Encode()) <= 0);
grandparent_index++;
}
seen_key = true;
if (overlapped_bytes + curr_file_size >
compaction->max_compaction_bytes()) {
// Too much overlap for current output; start new output
overlapped_bytes = 0;
return true;
}
return false;
}
};
// Maintains state for the entire compaction
@@ -262,13 +207,11 @@ void CompactionJob::AggregateStatistics() {
}
CompactionJob::CompactionJob(
int job_id, Compaction* compaction, const ImmutableDBOptions& db_options,
int job_id, Compaction* compaction, const DBOptions& db_options,
const EnvOptions& env_options, VersionSet* versions,
std::atomic<bool>* shutting_down, LogBuffer* log_buffer,
Directory* db_directory, Directory* output_directory, Statistics* stats,
InstrumentedMutex* db_mutex, Status* db_bg_error,
std::vector<SequenceNumber> existing_snapshots,
SequenceNumber earliest_write_conflict_snapshot,
std::shared_ptr<Cache> table_cache, EventLogger* event_logger,
bool paranoid_file_checks, bool measure_io_stats, const std::string& dbname,
CompactionJobStats* compaction_job_stats)
@@ -286,18 +229,13 @@ CompactionJob::CompactionJob(
db_directory_(db_directory),
output_directory_(output_directory),
stats_(stats),
db_mutex_(db_mutex),
db_bg_error_(db_bg_error),
existing_snapshots_(std::move(existing_snapshots)),
earliest_write_conflict_snapshot_(earliest_write_conflict_snapshot),
table_cache_(std::move(table_cache)),
event_logger_(event_logger),
paranoid_file_checks_(paranoid_file_checks),
measure_io_stats_(measure_io_stats) {
assert(log_buffer_ != nullptr);
const auto* cfd = compact_->compaction->column_family_data();
ThreadStatusUtil::SetColumnFamily(cfd, cfd->ioptions()->env,
db_options_.enable_thread_tracking);
ThreadStatusUtil::SetColumnFamily(compact_->compaction->column_family_data());
ThreadStatusUtil::SetThreadOperation(ThreadStatus::OP_COMPACTION);
ReportStartedCompaction(compaction);
}
@@ -309,9 +247,8 @@ CompactionJob::~CompactionJob() {
void CompactionJob::ReportStartedCompaction(
Compaction* compaction) {
const auto* cfd = compact_->compaction->column_family_data();
ThreadStatusUtil::SetColumnFamily(cfd, cfd->ioptions()->env,
db_options_.enable_thread_tracking);
ThreadStatusUtil::SetColumnFamily(
compact_->compaction->column_family_data());
ThreadStatusUtil::SetThreadOperationProperty(
ThreadStatus::COMPACTION_JOB_ID,
@@ -395,6 +332,11 @@ struct RangeWithSize {
: range(a, b), size(s) {}
};
bool SliceCompare(const Comparator* cmp, const Slice& a, const Slice& b) {
// Returns true if a < b
return cmp->Compare(ExtractUserKey(a), ExtractUserKey(b)) < 0;
}
// Generates a histogram representing potential divisions of key ranges from
// the input. It adds the starting and/or ending keys of certain input files
// to the working set and then finds the approximate size of data in between
@@ -403,13 +345,14 @@ struct RangeWithSize {
void CompactionJob::GenSubcompactionBoundaries() {
auto* c = compact_->compaction;
auto* cfd = c->column_family_data();
const Comparator* cfd_comparator = cfd->user_comparator();
std::vector<Slice> bounds;
std::set<Slice, std::function<bool(const Slice& a, const Slice& b)> > bounds(
std::bind(&SliceCompare, cfd->user_comparator(), std::placeholders::_1,
std::placeholders::_2));
int start_lvl = c->start_level();
int out_lvl = c->output_level();
// Add the starting and/or ending key of certain input files as a potential
// boundary
// boundary (because we're inserting into a set, it avoids duplicates)
for (size_t lvl_idx = 0; lvl_idx < c->num_input_levels(); lvl_idx++) {
int lvl = c->level(lvl_idx);
if (lvl >= start_lvl && lvl <= out_lvl) {
@@ -417,44 +360,34 @@ void CompactionJob::GenSubcompactionBoundaries() {
size_t num_files = flevel->num_files;
if (num_files == 0) {
continue;
break;
}
if (lvl == 0) {
// For level 0 add the starting and ending key of each file since the
// files may have greatly differing key ranges (not range-partitioned)
for (size_t i = 0; i < num_files; i++) {
bounds.emplace_back(flevel->files[i].smallest_key);
bounds.emplace_back(flevel->files[i].largest_key);
bounds.emplace(flevel->files[i].smallest_key);
bounds.emplace(flevel->files[i].largest_key);
}
} else {
// For all other levels add the smallest/largest key in the level to
// encompass the range covered by that level
bounds.emplace_back(flevel->files[0].smallest_key);
bounds.emplace_back(flevel->files[num_files - 1].largest_key);
bounds.emplace(flevel->files[0].smallest_key);
bounds.emplace(flevel->files[num_files - 1].largest_key);
if (lvl == out_lvl) {
// For the last level include the starting keys of all files since
// the last level is the largest and probably has the widest key
// range. Since it's range partitioned, the ending key of one file
// and the starting key of the next are very close (or identical).
for (size_t i = 1; i < num_files; i++) {
bounds.emplace_back(flevel->files[i].smallest_key);
bounds.emplace(flevel->files[i].smallest_key);
}
}
}
}
}
std::sort(bounds.begin(), bounds.end(),
[cfd_comparator] (const Slice& a, const Slice& b) -> bool {
return cfd_comparator->Compare(ExtractUserKey(a), ExtractUserKey(b)) < 0;
});
// Remove duplicated entries from bounds
bounds.erase(std::unique(bounds.begin(), bounds.end(),
[cfd_comparator] (const Slice& a, const Slice& b) -> bool {
return cfd_comparator->Compare(ExtractUserKey(a), ExtractUserKey(b)) == 0;
}), bounds.end());
// Combine consecutive pairs of boundaries into ranges with an approximate
// size of data covered by keys in that range
uint64_t sum = 0;
@@ -476,26 +409,27 @@ void CompactionJob::GenSubcompactionBoundaries() {
// Group the ranges into subcompactions
const double min_file_fill_percent = 4.0 / 5;
uint64_t max_output_files = static_cast<uint64_t>(
std::ceil(sum / min_file_fill_percent /
c->mutable_cf_options()->MaxFileSizeForLevel(out_lvl)));
uint64_t max_output_files = std::ceil(
sum / min_file_fill_percent /
cfd->GetCurrentMutableCFOptions()->MaxFileSizeForLevel(out_lvl));
uint64_t subcompactions =
std::min({static_cast<uint64_t>(ranges.size()),
static_cast<uint64_t>(db_options_.max_subcompactions),
max_output_files});
double mean = sum * 1.0 / subcompactions;
if (subcompactions > 1) {
double mean = sum * 1.0 / subcompactions;
// Greedily add ranges to the subcompaction until the sum of the ranges'
// sizes becomes >= the expected mean size of a subcompaction
sum = 0;
for (size_t i = 0; i < ranges.size() - 1; i++) {
sum += ranges[i].size;
if (subcompactions == 1) {
// If there's only one left to schedule then it goes to the end so no
// need to put an end boundary
continue;
break;
}
sum += ranges[i].size;
if (sum >= mean) {
boundaries_.emplace_back(ExtractUserKey(ranges[i].range.limit));
sizes_.emplace_back(sum);
@@ -538,7 +472,7 @@ Status CompactionJob::Run() {
thread.join();
}
if (output_directory_ && !db_options_.disable_data_sync) {
if (output_directory_ && !db_options_.disableDataSync) {
output_directory_->Fsync();
}
@@ -575,31 +509,22 @@ Status CompactionJob::Run() {
return status;
}
Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options,
InstrumentedMutex* db_mutex) {
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_INSTALL);
db_mutex_->AssertHeld();
db_mutex->AssertHeld();
Status status = compact_->status;
ColumnFamilyData* cfd = compact_->compaction->column_family_data();
cfd->internal_stats()->AddCompactionStats(
compact_->compaction->output_level(), compaction_stats_);
if (status.ok()) {
status = InstallCompactionResults(mutable_cf_options);
status = InstallCompactionResults(mutable_cf_options, db_mutex);
}
VersionStorageInfo::LevelSummaryStorage tmp;
auto vstorage = cfd->current()->storage_info();
const auto& stats = compaction_stats_;
double read_write_amp = 0.0;
double write_amp = 0.0;
if (stats.bytes_read_non_output_levels > 0) {
read_write_amp = (stats.bytes_written + stats.bytes_read_output_level +
stats.bytes_read_non_output_levels) /
static_cast<double>(stats.bytes_read_non_output_levels);
write_amp = stats.bytes_written /
static_cast<double>(stats.bytes_read_non_output_levels);
}
LogToBuffer(
log_buffer_,
"[%s] compacted to: %s, MB/sec: %.1f rd, %.1f wr, level %d, "
@@ -612,10 +537,16 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
stats.bytes_written / static_cast<double>(stats.micros),
compact_->compaction->output_level(),
stats.num_input_files_in_non_output_levels,
stats.num_input_files_in_output_level, stats.num_output_files,
stats.num_input_files_in_output_level,
stats.num_output_files,
stats.bytes_read_non_output_levels / 1048576.0,
stats.bytes_read_output_level / 1048576.0,
stats.bytes_written / 1048576.0, read_write_amp, write_amp,
stats.bytes_written / 1048576.0,
(stats.bytes_written + stats.bytes_read_output_level +
stats.bytes_read_non_output_levels) /
static_cast<double>(stats.bytes_read_non_output_levels),
stats.bytes_written /
static_cast<double>(stats.bytes_read_non_output_levels),
status.ToString().c_str(), stats.num_input_records,
stats.num_dropped_records);
@@ -632,13 +563,6 @@ Status CompactionJob::Install(const MutableCFOptions& mutable_cf_options) {
<< "num_output_records" << compact_->num_output_records
<< "num_subcompactions" << compact_->sub_compact_states.size();
if (compaction_job_stats_ != nullptr) {
stream << "num_single_delete_mismatches"
<< compaction_job_stats_->num_single_del_mismatch;
stream << "num_single_delete_fallthrough"
<< compaction_job_stats_->num_single_del_fallthru;
}
if (measure_io_stats_ && compaction_job_stats_ != nullptr) {
stream << "file_write_nanos" << compaction_job_stats_->file_write_nanos;
stream << "file_range_sync_nanos"
@@ -661,11 +585,8 @@ 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();
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.get()));
std::unique_ptr<InternalIterator> input(
versions_->MakeInputIterator(sub_compact->compaction));
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_PROCESS_KV);
@@ -686,31 +607,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
prev_prepare_write_nanos = IOSTATS(prepare_write_nanos);
}
const MutableCFOptions* mutable_cf_options =
sub_compact->compaction->mutable_cf_options();
// To build compression dictionary, we sample the first output file, assuming
// it'll reach the maximum length, and then use the dictionary for compressing
// subsequent output files. The dictionary may be less than max_dict_bytes if
// the first output file's length is less than the maximum.
const int kSampleLenShift = 6; // 2^6 = 64-byte samples
std::set<size_t> sample_begin_offsets;
if (bottommost_level_ &&
cfd->ioptions()->compression_opts.max_dict_bytes > 0) {
const size_t kMaxSamples =
cfd->ioptions()->compression_opts.max_dict_bytes >> kSampleLenShift;
const size_t kOutFileLen = mutable_cf_options->MaxFileSizeForLevel(
compact_->compaction->output_level());
if (kOutFileLen != port::kMaxSizet) {
const size_t kOutFileNumSamples = kOutFileLen >> kSampleLenShift;
Random64 generator{versions_->NewFileNumber()};
for (size_t i = 0; i < kMaxSamples; ++i) {
sample_begin_offsets.insert(generator.Uniform(kOutFileNumSamples)
<< kSampleLenShift);
}
}
}
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
auto compaction_filter = cfd->ioptions()->compaction_filter;
std::unique_ptr<CompactionFilter> compaction_filter_from_factory = nullptr;
if (compaction_filter == nullptr) {
@@ -721,7 +618,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
MergeHelper merge(
env_, cfd->user_comparator(), cfd->ioptions()->merge_operator,
compaction_filter, db_options_.info_log.get(),
mutable_cf_options->min_partial_merge_operands,
cfd->ioptions()->min_partial_merge_operands,
false /* internal key corruption is expected */,
existing_snapshots_.empty() ? 0 : existing_snapshots_.back(),
compact_->compaction->level(), db_options_.statistics.get());
@@ -741,18 +638,11 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
Status status;
sub_compact->c_iter.reset(new CompactionIterator(
input.get(), cfd->user_comparator(), &merge, versions_->LastSequence(),
&existing_snapshots_, earliest_write_conflict_snapshot_, env_, false,
range_del_agg.get(), sub_compact->compaction, compaction_filter));
&existing_snapshots_, env_, false, sub_compact->compaction,
compaction_filter));
auto c_iter = sub_compact->c_iter.get();
c_iter->SeekToFirst();
const auto& c_iter_stats = c_iter->iter_stats();
auto sample_begin_offset_iter = sample_begin_offsets.cbegin();
// data_begin_offset and compression_dict are only valid while generating
// dictionary from the first output file.
size_t data_begin_offset = 0;
std::string compression_dict;
compression_dict.reserve(cfd->ioptions()->compression_opts.max_dict_bytes);
// TODO(noetzli): check whether we could check !shutting_down_->... only
// only occasionally (see diff D42687)
while (status.ok() && !shutting_down_->load(std::memory_order_acquire) &&
@@ -767,16 +657,9 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
if (end != nullptr &&
cfd->user_comparator()->Compare(c_iter->user_key(), *end) >= 0) {
break;
} else if (sub_compact->compaction->output_level() != 0 &&
sub_compact->ShouldStopBefore(
key, sub_compact->current_output_file_size) &&
} else if (sub_compact->compaction->ShouldStopBefore(key) &&
sub_compact->builder != nullptr) {
CompactionIterationStats range_del_out_stats;
status = FinishCompactionOutputFile(input->status(), sub_compact,
range_del_agg.get(),
&range_del_out_stats, &key);
RecordDroppedKeys(range_del_out_stats,
&sub_compact->compaction_job_stats);
status = FinishCompactionOutputFile(input->status(), sub_compact);
if (!status.ok()) {
break;
}
@@ -799,89 +682,21 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
assert(sub_compact->builder != nullptr);
assert(sub_compact->current_output() != nullptr);
sub_compact->builder->Add(key, value);
sub_compact->current_output_file_size = sub_compact->builder->FileSize();
sub_compact->current_output()->meta.UpdateBoundaries(
key, c_iter->ikey().sequence);
sub_compact->num_output_records++;
if (sub_compact->outputs.size() == 1) { // first output file
// Check if this key/value overlaps any sample intervals; if so, appends
// overlapping portions to the dictionary.
for (const auto& data_elmt : {key, value}) {
size_t data_end_offset = data_begin_offset + data_elmt.size();
while (sample_begin_offset_iter != sample_begin_offsets.cend() &&
*sample_begin_offset_iter < data_end_offset) {
size_t sample_end_offset =
*sample_begin_offset_iter + (1 << kSampleLenShift);
// Invariant: Because we advance sample iterator while processing the
// data_elmt containing the sample's last byte, the current sample
// cannot end before the current data_elmt.
assert(data_begin_offset < sample_end_offset);
size_t data_elmt_copy_offset, data_elmt_copy_len;
if (*sample_begin_offset_iter <= data_begin_offset) {
// The sample starts before data_elmt starts, so take bytes starting
// at the beginning of data_elmt.
data_elmt_copy_offset = 0;
} else {
// data_elmt starts before the sample starts, so take bytes starting
// at the below offset into data_elmt.
data_elmt_copy_offset =
*sample_begin_offset_iter - data_begin_offset;
}
if (sample_end_offset <= data_end_offset) {
// The sample ends before data_elmt ends, so take as many bytes as
// needed.
data_elmt_copy_len =
sample_end_offset - (data_begin_offset + data_elmt_copy_offset);
} else {
// data_elmt ends before the sample ends, so take all remaining
// bytes in data_elmt.
data_elmt_copy_len =
data_end_offset - (data_begin_offset + data_elmt_copy_offset);
}
compression_dict.append(&data_elmt.data()[data_elmt_copy_offset],
data_elmt_copy_len);
if (sample_end_offset > data_end_offset) {
// Didn't finish sample. Try to finish it with the next data_elmt.
break;
}
// Next sample may require bytes from same data_elmt.
sample_begin_offset_iter++;
}
data_begin_offset = data_end_offset;
}
}
// Close output file if it is big enough
// TODO(aekmekji): determine if file should be closed earlier than this
// during subcompactions (i.e. if output size, estimated by input size, is
// going to be 1.2MB and max_output_file_size = 1MB, prefer to have 0.6MB
// and 0.6MB instead of 1MB and 0.2MB)
if (sub_compact->compaction->output_level() != 0 &&
sub_compact->current_output_file_size >=
sub_compact->compaction->max_output_file_size()) {
Status input_status = input->status();
c_iter->Next();
const Slice* next_key = nullptr;
if (c_iter->Valid()) {
next_key = &c_iter->key();
}
CompactionIterationStats range_del_out_stats;
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) {
// Use dictionary from first output file for compression of subsequent
// files.
sub_compact->compression_dict = std::move(compression_dict);
}
} else {
c_iter->Next();
if (sub_compact->builder->FileSize() >=
sub_compact->compaction->max_output_file_size()) {
status = FinishCompactionOutputFile(input->status(), sub_compact);
}
c_iter->Next();
}
sub_compact->num_input_records = c_iter_stats.num_input_records;
@@ -889,10 +704,6 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
c_iter_stats.num_input_deletion_records;
sub_compact->compaction_job_stats.num_corrupt_keys =
c_iter_stats.num_input_corrupt_records;
sub_compact->compaction_job_stats.num_single_del_fallthru =
c_iter_stats.num_single_del_fallthru;
sub_compact->compaction_job_stats.num_single_del_mismatch =
c_iter_stats.num_single_del_mismatch;
sub_compact->compaction_job_stats.total_input_raw_key_bytes +=
c_iter_stats.total_input_raw_key_bytes;
sub_compact->compaction_job_stats.total_input_raw_value_bytes +=
@@ -908,18 +719,8 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
status = Status::ShutdownInProgress(
"Database shutdown or Column family drop during compaction");
}
if (status.ok() && sub_compact->builder == nullptr &&
sub_compact->outputs.size() == 0 &&
range_del_agg->ShouldAddTombstones(bottommost_level_)) {
// handle subcompaction containing only range deletions
status = OpenCompactionOutputFile(sub_compact);
}
if (status.ok() && sub_compact->builder != nullptr) {
CompactionIterationStats range_del_out_stats;
status =
FinishCompactionOutputFile(input->status(), sub_compact,
range_del_agg.get(), &range_del_out_stats);
RecordDroppedKeys(range_del_out_stats, &sub_compact->compaction_job_stats);
status = FinishCompactionOutputFile(input->status(), sub_compact);
}
if (status.ok()) {
status = input->status();
@@ -945,7 +746,7 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
}
void CompactionJob::RecordDroppedKeys(
const CompactionIterationStats& c_iter_stats,
const CompactionIteratorStats& c_iter_stats,
CompactionJobStats* compaction_job_stats) {
if (c_iter_stats.num_record_drop_user > 0) {
RecordTick(stats_, COMPACTION_KEY_DROP_USER,
@@ -967,21 +768,10 @@ void CompactionJob::RecordDroppedKeys(
c_iter_stats.num_record_drop_obsolete;
}
}
if (c_iter_stats.num_record_drop_range_del > 0) {
RecordTick(stats_, COMPACTION_KEY_DROP_RANGE_DEL,
c_iter_stats.num_record_drop_range_del);
}
if (c_iter_stats.num_range_del_drop_obsolete > 0) {
RecordTick(stats_, COMPACTION_RANGE_DEL_DROP_OBSOLETE,
c_iter_stats.num_range_del_drop_obsolete);
}
}
Status CompactionJob::FinishCompactionOutputFile(
const Status& input_status, SubcompactionState* sub_compact,
RangeDelAggregator* range_del_agg,
CompactionIterationStats* range_del_out_stats,
const Slice* next_table_min_key /* = nullptr */) {
const Status& input_status, SubcompactionState* sub_compact) {
AutoThreadOperationStageUpdater stage_updater(
ThreadStatus::STAGE_COMPACTION_SYNC_FILE);
assert(sub_compact != nullptr);
@@ -996,36 +786,6 @@ Status CompactionJob::FinishCompactionOutputFile(
// Check for iterator errors
Status s = input_status;
auto meta = &sub_compact->current_output()->meta;
if (s.ok()) {
Slice lower_bound_guard, upper_bound_guard;
const Slice *lower_bound, *upper_bound;
if (sub_compact->outputs.size() == 1) {
// For the first output table, include range tombstones before the min key
// but after the subcompaction boundary.
lower_bound = sub_compact->start;
} else if (meta->smallest.size() > 0) {
// For subsequent output tables, only include range tombstones from min
// key onwards since the previous file was extended to contain range
// tombstones falling before min key.
lower_bound_guard = meta->smallest.user_key();
lower_bound = &lower_bound_guard;
} else {
lower_bound = nullptr;
}
if (next_table_min_key != nullptr) {
// This isn't the last file in the subcompaction, so extend until the next
// file starts.
upper_bound_guard = ExtractUserKey(*next_table_min_key);
upper_bound = &upper_bound_guard;
} else {
// This is the last file in the subcompaction, so extend until the
// subcompaction ends.
upper_bound = sub_compact->end;
}
range_del_agg->AddToBuilder(sub_compact->builder.get(), lower_bound,
upper_bound, meta, range_del_out_stats,
bottommost_level_);
}
const uint64_t current_entries = sub_compact->builder->NumEntries();
meta->marked_for_compaction = sub_compact->builder->NeedCompact();
if (s.ok()) {
@@ -1039,7 +799,7 @@ Status CompactionJob::FinishCompactionOutputFile(
sub_compact->total_bytes += current_bytes;
// Finish and check for file errors
if (s.ok() && !db_options_.disable_data_sync) {
if (s.ok() && !db_options_.disableDataSync) {
StopWatch sw(env_, stats_, COMPACTION_OUTFILE_SYNC_MICROS);
s = sub_compact->outfile->Sync(db_options_.use_fsync);
}
@@ -1048,15 +808,13 @@ Status CompactionJob::FinishCompactionOutputFile(
}
sub_compact->outfile.reset();
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
TableProperties tp;
if (s.ok() && current_entries > 0) {
// Verify that the table is usable
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
InternalIterator* iter = cfd->table_cache()->NewIterator(
ReadOptions(), env_options_, cfd->internal_comparator(), meta->fd,
nullptr /* range_del_agg */, nullptr,
cfd->internal_stats()->GetFileReadHist(
compact_->compaction->output_level()),
nullptr, cfd->internal_stats()->GetFileReadHist(
compact_->compaction->output_level()),
false);
s = iter->status();
@@ -1066,52 +824,35 @@ Status CompactionJob::FinishCompactionOutputFile(
}
delete iter;
// Output to event logger and fire events.
if (s.ok()) {
tp = sub_compact->builder->GetTableProperties();
auto tp = sub_compact->builder->GetTableProperties();
sub_compact->current_output()->table_properties =
std::make_shared<TableProperties>(tp);
TableFileCreationInfo info(std::move(tp));
info.db_name = dbname_;
info.cf_name = cfd->GetName();
info.file_path =
TableFileName(cfd->ioptions()->db_paths, meta->fd.GetNumber(),
meta->fd.GetPathId());
info.file_size = meta->fd.GetFileSize();
info.job_id = job_id_;
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"[%s] [JOB %d] Generated table #%" PRIu64 ": %" PRIu64
" keys, %" PRIu64 " bytes%s",
cfd->GetName().c_str(), job_id_, output_number, current_entries,
current_bytes,
meta->marked_for_compaction ? " (need compaction)" : "");
EventHelpers::LogAndNotifyTableFileCreation(
event_logger_, cfd->ioptions()->listeners, meta->fd, info);
}
}
std::string fname = TableFileName(db_options_.db_paths, meta->fd.GetNumber(),
meta->fd.GetPathId());
EventHelpers::LogAndNotifyTableFileCreationFinished(
event_logger_, cfd->ioptions()->listeners, dbname_, cfd->GetName(), fname,
job_id_, meta->fd, tp, TableFileCreationReason::kCompaction, s);
// Report new file to SstFileManagerImpl
auto sfm =
static_cast<SstFileManagerImpl*>(db_options_.sst_file_manager.get());
if (sfm && meta->fd.GetPathId() == 0) {
auto fn = TableFileName(cfd->ioptions()->db_paths, meta->fd.GetNumber(),
meta->fd.GetPathId());
sfm->OnAddFile(fn);
if (sfm->IsMaxAllowedSpaceReached()) {
InstrumentedMutexLock l(db_mutex_);
if (db_bg_error_->ok()) {
s = Status::IOError("Max allowed space was reached");
*db_bg_error_ = s;
TEST_SYNC_POINT(
"CompactionJob::FinishCompactionOutputFile:MaxAllowedSpaceReached");
}
}
}
sub_compact->builder.reset();
sub_compact->current_output_file_size = 0;
return s;
}
Status CompactionJob::InstallCompactionResults(
const MutableCFOptions& mutable_cf_options) {
db_mutex_->AssertHeld();
const MutableCFOptions& mutable_cf_options, InstrumentedMutex* db_mutex) {
db_mutex->AssertHeld();
auto* compaction = compact_->compaction;
// paranoia: verify that the files that we started with
@@ -1146,7 +887,7 @@ Status CompactionJob::InstallCompactionResults(
}
return versions_->LogAndApply(compaction->column_family_data(),
mutable_cf_options, compaction->edit(),
db_mutex_, db_directory_);
db_mutex, db_directory_);
}
void CompactionJob::RecordCompactionIOStats() {
@@ -1166,17 +907,10 @@ Status CompactionJob::OpenCompactionOutputFile(
assert(sub_compact->builder == nullptr);
// no need to lock because VersionSet::next_file_number_ is atomic
uint64_t file_number = versions_->NewFileNumber();
std::string fname = TableFileName(db_options_.db_paths, file_number,
sub_compact->compaction->output_path_id());
// Fire events.
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
#ifndef ROCKSDB_LITE
EventHelpers::NotifyTableFileCreationStarted(
cfd->ioptions()->listeners, dbname_, cfd->GetName(), fname, job_id_,
TableFileCreationReason::kCompaction);
#endif // !ROCKSDB_LITE
// Make the output file
unique_ptr<WritableFile> writable_file;
std::string fname = TableFileName(db_options_.db_paths, file_number,
sub_compact->compaction->output_path_id());
Status s = NewWritableFile(env_, fname, &writable_file, env_options_);
if (!s.ok()) {
Log(InfoLogLevel::ERROR_LEVEL, db_options_.info_log,
@@ -1185,13 +919,8 @@ Status CompactionJob::OpenCompactionOutputFile(
sub_compact->compaction->column_family_data()->GetName().c_str(),
job_id_, file_number, s.ToString().c_str());
LogFlush(db_options_.info_log);
EventHelpers::LogAndNotifyTableFileCreationFinished(
event_logger_, cfd->ioptions()->listeners, dbname_, cfd->GetName(),
fname, job_id_, FileDescriptor(), TableProperties(),
TableFileCreationReason::kCompaction, s);
return s;
}
SubcompactionState::Output out;
out.meta.fd =
FileDescriptor(file_number, sub_compact->compaction->output_path_id(), 0);
@@ -1204,6 +933,7 @@ Status CompactionJob::OpenCompactionOutputFile(
sub_compact->outfile.reset(
new WritableFileWriter(std::move(writable_file), env_options_));
ColumnFamilyData* cfd = sub_compact->compaction->column_family_data();
// If the Column family flag is to only optimize filters for hits,
// we can skip creating filters if this is the bottommost_level where
// data is going to be found
@@ -1211,12 +941,9 @@ Status CompactionJob::OpenCompactionOutputFile(
cfd->ioptions()->optimize_filters_for_hits && bottommost_level_;
sub_compact->builder.reset(NewTableBuilder(
*cfd->ioptions(), cfd->internal_comparator(),
cfd->int_tbl_prop_collector_factories(), cfd->GetID(), cfd->GetName(),
cfd->int_tbl_prop_collector_factories(), cfd->GetID(),
sub_compact->outfile.get(), sub_compact->compaction->output_compression(),
cfd->ioptions()->compression_opts,
sub_compact->compaction->output_level(),
&sub_compact->compression_dict,
skip_filters));
cfd->ioptions()->compression_opts, skip_filters));
LogFlush(db_options_.info_log);
return s;
}
@@ -1263,7 +990,8 @@ void CompactionJob::UpdateCompactionStats() {
for (int input_level = 0;
input_level < static_cast<int>(compaction->num_input_levels());
++input_level) {
if (compaction->level(input_level) != compaction->output_level()) {
if (compaction->start_level() + input_level
!= compaction->output_level()) {
UpdateCompactionInputStatsHelper(
&compaction_stats_.num_input_files_in_non_output_levels,
&compaction_stats_.bytes_read_non_output_levels,
+11 -26
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -25,7 +25,6 @@
#include "db/job_context.h"
#include "db/log_writer.h"
#include "db/memtable_list.h"
#include "db/range_del_aggregator.h"
#include "db/version_edit.h"
#include "db/write_controller.h"
#include "db/write_thread.h"
@@ -38,7 +37,6 @@
#include "rocksdb/transaction_log.h"
#include "table/scoped_arena_iterator.h"
#include "util/autovector.h"
#include "util/db_options.h"
#include "util/event_logger.h"
#include "util/stop_watch.h"
#include "util/thread_local.h"
@@ -54,15 +52,12 @@ class Arena;
class CompactionJob {
public:
CompactionJob(int job_id, Compaction* compaction,
const ImmutableDBOptions& db_options,
CompactionJob(int job_id, Compaction* compaction, const DBOptions& db_options,
const EnvOptions& env_options, VersionSet* versions,
std::atomic<bool>* shutting_down, LogBuffer* log_buffer,
Directory* db_directory, Directory* output_directory,
Statistics* stats, InstrumentedMutex* db_mutex,
Status* db_bg_error,
Statistics* stats,
std::vector<SequenceNumber> existing_snapshots,
SequenceNumber earliest_write_conflict_snapshot,
std::shared_ptr<Cache> table_cache, EventLogger* event_logger,
bool paranoid_file_checks, bool measure_io_stats,
const std::string& dbname,
@@ -81,7 +76,8 @@ class CompactionJob {
Status Run();
// REQUIRED: mutex held
Status Install(const MutableCFOptions& mutable_cf_options);
Status Install(const MutableCFOptions& mutable_cf_options,
InstrumentedMutex* db_mutex);
private:
struct SubcompactionState;
@@ -96,18 +92,16 @@ class CompactionJob {
// kv-pairs
void ProcessKeyValueCompaction(SubcompactionState* sub_compact);
Status FinishCompactionOutputFile(
const Status& input_status, SubcompactionState* sub_compact,
RangeDelAggregator* range_del_agg,
CompactionIterationStats* range_del_out_stats,
const Slice* next_table_min_key = nullptr);
Status InstallCompactionResults(const MutableCFOptions& mutable_cf_options);
Status FinishCompactionOutputFile(const Status& input_status,
SubcompactionState* sub_compact);
Status InstallCompactionResults(const MutableCFOptions& mutable_cf_options,
InstrumentedMutex* db_mutex);
void RecordCompactionIOStats();
Status OpenCompactionOutputFile(SubcompactionState* sub_compact);
void CleanupCompaction();
void UpdateCompactionJobStats(
const InternalStats::CompactionStats& stats) const;
void RecordDroppedKeys(const CompactionIterationStats& c_iter_stats,
void RecordDroppedKeys(const CompactionIteratorStats& c_iter_stats,
CompactionJobStats* compaction_job_stats = nullptr);
void UpdateCompactionStats();
@@ -126,9 +120,8 @@ class CompactionJob {
// DBImpl state
const std::string& dbname_;
const ImmutableDBOptions& db_options_;
const DBOptions& db_options_;
const EnvOptions& env_options_;
Env* env_;
VersionSet* versions_;
std::atomic<bool>* shutting_down_;
@@ -136,19 +129,11 @@ class CompactionJob {
Directory* db_directory_;
Directory* output_directory_;
Statistics* stats_;
InstrumentedMutex* db_mutex_;
Status* db_bg_error_;
// If there were two snapshots with seq numbers s1 and
// s2 and s1 < s2, and if we find two instances of a key k1 then lies
// entirely within s1 and s2, then the earlier version of k1 can be safely
// deleted because that version is not visible in any snapshot.
std::vector<SequenceNumber> existing_snapshots_;
// This is the earliest snapshot that could be used for write-conflict
// checking by a transaction. For any user-key newer than this snapshot, we
// should make sure not to remove evidence that a write occurred.
SequenceNumber earliest_write_conflict_snapshot_;
std::shared_ptr<Cache> table_cache_;
EventLogger* event_logger_;
+5 -10
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -551,9 +551,8 @@ uint64_t EstimatedFileSize(
const size_t kFooterSize = 512;
uint64_t data_size =
static_cast<uint64_t>(
num_records * (key_size + value_size * compression_ratio +
kPerKeyOverhead));
kPerKeyOverhead);
return data_size + kFooterSize
+ num_records * bloom_bits_per_key / 8 // filter block
@@ -624,10 +623,7 @@ CompressionType GetAnyCompression() {
return kBZip2Compression;
} else if (LZ4_Supported()) {
return kLZ4Compression;
} else if (XPRESS_Supported()) {
return kXpressCompression;
}
return kNoCompression;
}
@@ -663,7 +659,7 @@ TEST_P(CompactionJobStatsTest, CompactionJobStatsTest) {
options.max_subcompactions = max_subcompactions_;
options.bytes_per_sync = 512 * 1024;
options.report_bg_io_stats = true;
options.compaction_measure_io_stats = true;
for (int test = 0; test < 2; ++test) {
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -957,7 +953,7 @@ TEST_P(CompactionJobStatsTest, UniversalCompactionTest) {
uint64_t key_base = 100000000l;
// Note: key_base must be multiple of num_keys_per_L0_file
int num_keys_per_table = 100;
const uint32_t kTestScale = 6;
const uint32_t kTestScale = 8;
const int kKeySize = 10;
const int kValueSize = 900;
double compression_ratio = 1.0;
@@ -1008,9 +1004,8 @@ TEST_P(CompactionJobStatsTest, UniversalCompactionTest) {
num_input_units,
num_keys_per_table * num_input_units,
1.0, 0, false));
dbfull()->TEST_WaitForCompact();
}
ASSERT_EQ(stats_checker->NumberOfUnverifiedStats(), 3U);
ASSERT_EQ(stats_checker->NumberOfUnverifiedStats(), 4U);
for (uint64_t start_key = key_base;
start_key <= key_base * kTestScale;
+55 -275
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -10,13 +10,13 @@
#include <string>
#include <tuple>
#include "db/column_family.h"
#include "db/compaction_job.h"
#include "db/column_family.h"
#include "db/version_set.h"
#include "db/writebuffer.h"
#include "rocksdb/cache.h"
#include "rocksdb/db.h"
#include "rocksdb/options.h"
#include "rocksdb/write_buffer_manager.h"
#include "table/mock_table.h"
#include "util/file_reader_writer.h"
#include "util/string_util.h"
@@ -68,12 +68,11 @@ class CompactionJobTest : public testing::Test {
CompactionJobTest()
: env_(Env::Default()),
dbname_(test::TmpDir() + "/compaction_job_test"),
db_options_(),
mutable_cf_options_(cf_options_),
mutable_cf_options_(Options(), ImmutableCFOptions(Options())),
table_cache_(NewLRUCache(50000, 16)),
write_buffer_manager_(db_options_.db_write_buffer_size),
write_buffer_(db_options_.db_write_buffer_size),
versions_(new VersionSet(dbname_, &db_options_, env_options_,
table_cache_.get(), &write_buffer_manager_,
table_cache_.get(), &write_buffer_,
&write_controller_)),
shutting_down_(false),
mock_table_factory_(new mock::MockTableFactory()) {
@@ -163,12 +162,9 @@ class CompactionJobTest : public testing::Test {
auto key = ToString(i * kMatchingKeys + k);
auto value = ToString(i * kKeysPerFile + k);
InternalKey internal_key(key, ++sequence_number, kTypeValue);
// This is how the key will look like once it's written in bottommost
// file
InternalKey bottommost_internal_key(
key, (key == "9999") ? sequence_number : 0, kTypeValue);
InternalKey bottommost_internal_key(key, 0, kTypeValue);
if (corrupt_id(k)) {
test::CorruptKeyType(&internal_key);
test::CorruptKeyType(&bottommost_internal_key);
@@ -221,11 +217,9 @@ class CompactionJobTest : public testing::Test {
cfd_ = versions_->GetColumnFamilySet()->GetDefault();
}
void RunCompaction(
const std::vector<std::vector<FileMetaData*>>& input_files,
const stl_wrappers::KVMap& expected_results,
const std::vector<SequenceNumber>& snapshots = {},
SequenceNumber earliest_write_conflict_snapshot = kMaxSequenceNumber) {
void RunCompaction(const std::vector<std::vector<FileMetaData*>>& input_files,
const stl_wrappers::KVMap& expected_results,
const std::vector<SequenceNumber>& snapshots = {}) {
auto cfd = versions_->GetColumnFamilySet()->GetDefault();
size_t num_input_files = 0;
@@ -240,20 +234,20 @@ class CompactionJobTest : public testing::Test {
num_input_files += level_files.size();
}
Compaction compaction(cfd->current()->storage_info(), *cfd->ioptions(),
Compaction compaction(cfd->current()->storage_info(),
*cfd->GetLatestMutableCFOptions(),
compaction_input_files, 1, 1024 * 1024,
10 * 1024 * 1024, 0, kNoCompression, {}, true);
compaction_input_files, 1, 1024 * 1024, 10, 0,
kNoCompression, {}, true);
compaction.SetInputVersion(cfd->current());
LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, db_options_.info_log.get());
mutex_.Lock();
EventLogger event_logger(db_options_.info_log.get());
CompactionJob compaction_job(
0, &compaction, db_options_, env_options_, versions_.get(),
&shutting_down_, &log_buffer, nullptr, nullptr, nullptr, &mutex_,
&bg_error_, snapshots, earliest_write_conflict_snapshot, table_cache_,
&event_logger, false, false, dbname_, &compaction_job_stats_);
CompactionJob compaction_job(0, &compaction, db_options_, env_options_,
versions_.get(), &shutting_down_, &log_buffer,
nullptr, nullptr, nullptr, snapshots,
table_cache_, &event_logger, false, false,
dbname_, &compaction_job_stats_);
VerifyInitializationOfCompactionJobStats(compaction_job_stats_);
@@ -263,7 +257,8 @@ class CompactionJobTest : public testing::Test {
s = compaction_job.Run();
ASSERT_OK(s);
mutex_.Lock();
ASSERT_OK(compaction_job.Install(*cfd->GetLatestMutableCFOptions()));
ASSERT_OK(compaction_job.Install(*cfd->GetLatestMutableCFOptions(),
&mutex_));
mutex_.Unlock();
if (expected_results.size() == 0) {
@@ -281,12 +276,12 @@ class CompactionJobTest : public testing::Test {
Env* env_;
std::string dbname_;
EnvOptions env_options_;
ImmutableDBOptions db_options_;
ColumnFamilyOptions cf_options_;
MutableCFOptions mutable_cf_options_;
std::shared_ptr<Cache> table_cache_;
WriteController write_controller_;
WriteBufferManager write_buffer_manager_;
DBOptions db_options_;
ColumnFamilyOptions cf_options_;
WriteBuffer write_buffer_;
std::unique_ptr<VersionSet> versions_;
InstrumentedMutex mutex_;
std::atomic<bool> shutting_down_;
@@ -295,7 +290,6 @@ class CompactionJobTest : public testing::Test {
ColumnFamilyData* cfd_;
std::unique_ptr<CompactionFilter> compaction_filter_;
std::shared_ptr<MergeOperator> merge_op_;
Status bg_error_;
};
TEST_F(CompactionJobTest, Simple) {
@@ -352,7 +346,7 @@ TEST_F(CompactionJobTest, SimpleOverwrite) {
auto expected_results =
mock::MakeMockFile({{KeyStr("a", 0U, kTypeValue), "val2"},
{KeyStr("b", 4U, kTypeValue), "val3"}});
{KeyStr("b", 0U, kTypeValue), "val3"}});
SetLastSequence(4U);
auto files = cfd_->current()->storage_info()->LevelFiles(0);
@@ -405,7 +399,7 @@ TEST_F(CompactionJobTest, SimpleMerge) {
auto expected_results =
mock::MakeMockFile({{KeyStr("a", 0U, kTypeValue), "3,4,5"},
{KeyStr("b", 2U, kTypeValue), "1,2"}});
{KeyStr("b", 0U, kTypeValue), "1,2"}});
SetLastSequence(5U);
auto files = cfd_->current()->storage_info()->LevelFiles(0);
@@ -565,144 +559,34 @@ TEST_F(CompactionJobTest, SimpleSingleDelete) {
TEST_F(CompactionJobTest, SingleDeleteSnapshots) {
NewDB();
auto file1 = mock::MakeMockFile({
{KeyStr("A", 12U, kTypeSingleDeletion), ""},
{KeyStr("a", 12U, kTypeSingleDeletion), ""},
{KeyStr("b", 21U, kTypeSingleDeletion), ""},
{KeyStr("c", 22U, kTypeSingleDeletion), ""},
{KeyStr("d", 9U, kTypeSingleDeletion), ""},
{KeyStr("f", 21U, kTypeSingleDeletion), ""},
{KeyStr("j", 11U, kTypeSingleDeletion), ""},
{KeyStr("j", 9U, kTypeSingleDeletion), ""},
{KeyStr("k", 12U, kTypeSingleDeletion), ""},
{KeyStr("k", 11U, kTypeSingleDeletion), ""},
{KeyStr("l", 3U, kTypeSingleDeletion), ""},
{KeyStr("l", 2U, kTypeSingleDeletion), ""},
});
auto file1 = mock::MakeMockFile({{KeyStr("A", 12U, kTypeSingleDeletion), ""},
{KeyStr("a", 12U, kTypeSingleDeletion), ""},
{KeyStr("b", 21U, kTypeSingleDeletion), ""},
{KeyStr("c", 22U, kTypeSingleDeletion), ""},
{KeyStr("d", 9U, kTypeSingleDeletion), ""}});
AddMockFile(file1);
auto file2 = mock::MakeMockFile({
{KeyStr("0", 2U, kTypeSingleDeletion), ""},
{KeyStr("a", 11U, kTypeValue), "val1"},
{KeyStr("b", 11U, kTypeValue), "val2"},
{KeyStr("c", 21U, kTypeValue), "val3"},
{KeyStr("d", 8U, kTypeValue), "val4"},
{KeyStr("e", 2U, kTypeSingleDeletion), ""},
{KeyStr("f", 1U, kTypeValue), "val1"},
{KeyStr("g", 11U, kTypeSingleDeletion), ""},
{KeyStr("h", 2U, kTypeSingleDeletion), ""},
{KeyStr("m", 12U, kTypeValue), "val1"},
{KeyStr("m", 11U, kTypeSingleDeletion), ""},
{KeyStr("m", 8U, kTypeValue), "val2"},
});
auto file2 = mock::MakeMockFile({{KeyStr("0", 2U, kTypeSingleDeletion), ""},
{KeyStr("a", 11U, kTypeValue), "val1"},
{KeyStr("b", 11U, kTypeValue), "val2"},
{KeyStr("c", 21U, kTypeValue), "val3"},
{KeyStr("d", 8U, kTypeValue), "val4"},
{KeyStr("e", 2U, kTypeSingleDeletion), ""}});
AddMockFile(file2);
auto file3 = mock::MakeMockFile({
{KeyStr("A", 1U, kTypeValue), "val"},
{KeyStr("e", 1U, kTypeValue), "val"},
});
auto file3 = mock::MakeMockFile({{KeyStr("A", 1U, kTypeValue), "val"},
{KeyStr("e", 1U, kTypeValue), "val"}});
AddMockFile(file3, 2);
auto expected_results = mock::MakeMockFile({
{KeyStr("A", 12U, kTypeSingleDeletion), ""},
{KeyStr("a", 12U, kTypeSingleDeletion), ""},
{KeyStr("a", 11U, kTypeValue), ""},
{KeyStr("b", 21U, kTypeSingleDeletion), ""},
{KeyStr("b", 11U, kTypeValue), "val2"},
{KeyStr("c", 22U, kTypeSingleDeletion), ""},
{KeyStr("c", 21U, kTypeValue), ""},
{KeyStr("e", 2U, kTypeSingleDeletion), ""},
{KeyStr("f", 21U, kTypeSingleDeletion), ""},
{KeyStr("f", 1U, kTypeValue), "val1"},
{KeyStr("g", 11U, kTypeSingleDeletion), ""},
{KeyStr("j", 11U, kTypeSingleDeletion), ""},
{KeyStr("k", 11U, kTypeSingleDeletion), ""},
{KeyStr("m", 12U, kTypeValue), "val1"},
{KeyStr("m", 11U, kTypeSingleDeletion), ""},
{KeyStr("m", 8U, kTypeValue), "val2"},
});
auto expected_results =
mock::MakeMockFile({{KeyStr("A", 12U, kTypeSingleDeletion), ""},
{KeyStr("b", 21U, kTypeSingleDeletion), ""},
{KeyStr("b", 11U, kTypeValue), "val2"},
{KeyStr("e", 2U, kTypeSingleDeletion), ""}});
SetLastSequence(22U);
auto files = cfd_->current()->storage_info()->LevelFiles(0);
RunCompaction({files}, expected_results, {10U, 20U}, 10U);
}
TEST_F(CompactionJobTest, EarliestWriteConflictSnapshot) {
NewDB();
// Test multiple snapshots where the earliest snapshot is not a
// write-conflic-snapshot.
auto file1 = mock::MakeMockFile({
{KeyStr("A", 24U, kTypeSingleDeletion), ""},
{KeyStr("A", 23U, kTypeValue), "val"},
{KeyStr("B", 24U, kTypeSingleDeletion), ""},
{KeyStr("B", 23U, kTypeValue), "val"},
{KeyStr("D", 24U, kTypeSingleDeletion), ""},
{KeyStr("G", 32U, kTypeSingleDeletion), ""},
{KeyStr("G", 31U, kTypeValue), "val"},
{KeyStr("G", 24U, kTypeSingleDeletion), ""},
{KeyStr("G", 23U, kTypeValue), "val2"},
{KeyStr("H", 31U, kTypeValue), "val"},
{KeyStr("H", 24U, kTypeSingleDeletion), ""},
{KeyStr("H", 23U, kTypeValue), "val"},
{KeyStr("I", 35U, kTypeSingleDeletion), ""},
{KeyStr("I", 34U, kTypeValue), "val2"},
{KeyStr("I", 33U, kTypeSingleDeletion), ""},
{KeyStr("I", 32U, kTypeValue), "val3"},
{KeyStr("I", 31U, kTypeSingleDeletion), ""},
{KeyStr("J", 34U, kTypeValue), "val"},
{KeyStr("J", 33U, kTypeSingleDeletion), ""},
{KeyStr("J", 25U, kTypeValue), "val2"},
{KeyStr("J", 24U, kTypeSingleDeletion), ""},
});
AddMockFile(file1);
auto file2 = mock::MakeMockFile({
{KeyStr("A", 14U, kTypeSingleDeletion), ""},
{KeyStr("A", 13U, kTypeValue), "val2"},
{KeyStr("C", 14U, kTypeSingleDeletion), ""},
{KeyStr("C", 13U, kTypeValue), "val"},
{KeyStr("E", 12U, kTypeSingleDeletion), ""},
{KeyStr("F", 4U, kTypeSingleDeletion), ""},
{KeyStr("F", 3U, kTypeValue), "val"},
{KeyStr("G", 14U, kTypeSingleDeletion), ""},
{KeyStr("G", 13U, kTypeValue), "val3"},
{KeyStr("H", 14U, kTypeSingleDeletion), ""},
{KeyStr("H", 13U, kTypeValue), "val2"},
{KeyStr("I", 13U, kTypeValue), "val4"},
{KeyStr("I", 12U, kTypeSingleDeletion), ""},
{KeyStr("I", 11U, kTypeValue), "val5"},
{KeyStr("J", 15U, kTypeValue), "val3"},
{KeyStr("J", 14U, kTypeSingleDeletion), ""},
});
AddMockFile(file2);
auto expected_results = mock::MakeMockFile({
{KeyStr("A", 24U, kTypeSingleDeletion), ""},
{KeyStr("A", 23U, kTypeValue), ""},
{KeyStr("B", 24U, kTypeSingleDeletion), ""},
{KeyStr("B", 23U, kTypeValue), ""},
{KeyStr("D", 24U, kTypeSingleDeletion), ""},
{KeyStr("E", 12U, kTypeSingleDeletion), ""},
{KeyStr("G", 32U, kTypeSingleDeletion), ""},
{KeyStr("G", 31U, kTypeValue), ""},
{KeyStr("H", 31U, kTypeValue), "val"},
{KeyStr("I", 35U, kTypeSingleDeletion), ""},
{KeyStr("I", 34U, kTypeValue), ""},
{KeyStr("I", 31U, kTypeSingleDeletion), ""},
{KeyStr("I", 13U, kTypeValue), "val4"},
{KeyStr("J", 34U, kTypeValue), "val"},
{KeyStr("J", 33U, kTypeSingleDeletion), ""},
{KeyStr("J", 25U, kTypeValue), "val2"},
{KeyStr("J", 24U, kTypeSingleDeletion), ""},
{KeyStr("J", 15U, kTypeValue), "val3"},
{KeyStr("J", 14U, kTypeSingleDeletion), ""},
});
SetLastSequence(24U);
auto files = cfd_->current()->storage_info()->LevelFiles(0);
RunCompaction({files}, expected_results, {10U, 20U, 30U}, 20U);
RunCompaction({files}, expected_results, {10U, 20U});
}
TEST_F(CompactionJobTest, SingleDeleteZeroSeq) {
@@ -720,7 +604,7 @@ TEST_F(CompactionJobTest, SingleDeleteZeroSeq) {
AddMockFile(file2);
auto expected_results = mock::MakeMockFile({
{KeyStr("dummy", 5U, kTypeValue), "val2"},
{KeyStr("dummy", 0U, kTypeValue), "val2"},
});
SetLastSequence(22U);
@@ -732,22 +616,9 @@ TEST_F(CompactionJobTest, MultiSingleDelete) {
// Tests three scenarios involving multiple single delete/put pairs:
//
// A: Put Snapshot SDel Put SDel -> Put Snapshot SDel
// B: Snapshot Put SDel Put SDel Snapshot -> Snapshot SDel Snapshot
// B: Put SDel Put SDel -> (Removed)
// C: SDel Put SDel Snapshot Put -> Snapshot Put
// D: (Put) SDel Snapshot Put SDel -> (Put) SDel Snapshot SDel
// E: Put SDel Snapshot Put SDel -> Snapshot SDel
// F: Put SDel Put Sdel Snapshot -> removed
// G: Snapshot SDel Put SDel Put -> Snapshot Put SDel
// H: (Put) Put SDel Put Sdel Snapshot -> Removed
// I: (Put) Snapshot Put SDel Put SDel -> SDel
// J: Put Put SDel Put SDel SDel Snapshot Put Put SDel SDel Put
// -> Snapshot Put
// K: SDel SDel Put SDel Put Put Snapshot SDel Put SDel SDel Put SDel
// -> Snapshot Put Snapshot SDel
// L: SDel Put Del Put SDel Snapshot Del Put Del SDel Put SDel
// -> Snapshot SDel
// M: (Put) SDel Put Del Put SDel Snapshot Put Del SDel Put SDel Del
// -> SDel Snapshot Del
// D: (Put) SDel Snapshot Put SDel -> (Put) SDel Snapshot
NewDB();
auto file1 = mock::MakeMockFile({
@@ -759,34 +630,6 @@ TEST_F(CompactionJobTest, MultiSingleDelete) {
{KeyStr("C", 14U, kTypeValue), "val3"},
{KeyStr("D", 12U, kTypeSingleDeletion), ""},
{KeyStr("D", 11U, kTypeValue), "val4"},
{KeyStr("G", 15U, kTypeValue), "val"},
{KeyStr("G", 14U, kTypeSingleDeletion), ""},
{KeyStr("G", 13U, kTypeValue), "val"},
{KeyStr("I", 14U, kTypeSingleDeletion), ""},
{KeyStr("I", 13U, kTypeValue), "val"},
{KeyStr("J", 15U, kTypeValue), "val"},
{KeyStr("J", 14U, kTypeSingleDeletion), ""},
{KeyStr("J", 13U, kTypeSingleDeletion), ""},
{KeyStr("J", 12U, kTypeValue), "val"},
{KeyStr("J", 11U, kTypeValue), "val"},
{KeyStr("K", 16U, kTypeSingleDeletion), ""},
{KeyStr("K", 15U, kTypeValue), "val1"},
{KeyStr("K", 14U, kTypeSingleDeletion), ""},
{KeyStr("K", 13U, kTypeSingleDeletion), ""},
{KeyStr("K", 12U, kTypeValue), "val2"},
{KeyStr("K", 11U, kTypeSingleDeletion), ""},
{KeyStr("L", 16U, kTypeSingleDeletion), ""},
{KeyStr("L", 15U, kTypeValue), "val"},
{KeyStr("L", 14U, kTypeSingleDeletion), ""},
{KeyStr("L", 13U, kTypeDeletion), ""},
{KeyStr("L", 12U, kTypeValue), "val"},
{KeyStr("L", 11U, kTypeDeletion), ""},
{KeyStr("M", 16U, kTypeDeletion), ""},
{KeyStr("M", 15U, kTypeSingleDeletion), ""},
{KeyStr("M", 14U, kTypeValue), "val"},
{KeyStr("M", 13U, kTypeSingleDeletion), ""},
{KeyStr("M", 12U, kTypeDeletion), ""},
{KeyStr("M", 11U, kTypeValue), "val"},
});
AddMockFile(file1);
@@ -798,87 +641,24 @@ TEST_F(CompactionJobTest, MultiSingleDelete) {
{KeyStr("C", 9U, kTypeValue), "val6"},
{KeyStr("C", 8U, kTypeSingleDeletion), ""},
{KeyStr("D", 10U, kTypeSingleDeletion), ""},
{KeyStr("E", 12U, kTypeSingleDeletion), ""},
{KeyStr("E", 11U, kTypeValue), "val"},
{KeyStr("E", 5U, kTypeSingleDeletion), ""},
{KeyStr("E", 4U, kTypeValue), "val"},
{KeyStr("F", 6U, kTypeSingleDeletion), ""},
{KeyStr("F", 5U, kTypeValue), "val"},
{KeyStr("F", 4U, kTypeSingleDeletion), ""},
{KeyStr("F", 3U, kTypeValue), "val"},
{KeyStr("G", 12U, kTypeSingleDeletion), ""},
{KeyStr("H", 6U, kTypeSingleDeletion), ""},
{KeyStr("H", 5U, kTypeValue), "val"},
{KeyStr("H", 4U, kTypeSingleDeletion), ""},
{KeyStr("H", 3U, kTypeValue), "val"},
{KeyStr("I", 12U, kTypeSingleDeletion), ""},
{KeyStr("I", 11U, kTypeValue), "val"},
{KeyStr("J", 6U, kTypeSingleDeletion), ""},
{KeyStr("J", 5U, kTypeSingleDeletion), ""},
{KeyStr("J", 4U, kTypeValue), "val"},
{KeyStr("J", 3U, kTypeSingleDeletion), ""},
{KeyStr("J", 2U, kTypeValue), "val"},
{KeyStr("K", 8U, kTypeValue), "val3"},
{KeyStr("K", 7U, kTypeValue), "val4"},
{KeyStr("K", 6U, kTypeSingleDeletion), ""},
{KeyStr("K", 5U, kTypeValue), "val5"},
{KeyStr("K", 2U, kTypeSingleDeletion), ""},
{KeyStr("K", 1U, kTypeSingleDeletion), ""},
{KeyStr("L", 5U, kTypeSingleDeletion), ""},
{KeyStr("L", 4U, kTypeValue), "val"},
{KeyStr("L", 3U, kTypeDeletion), ""},
{KeyStr("L", 2U, kTypeValue), "val"},
{KeyStr("L", 1U, kTypeSingleDeletion), ""},
{KeyStr("M", 10U, kTypeSingleDeletion), ""},
{KeyStr("M", 7U, kTypeValue), "val"},
{KeyStr("M", 5U, kTypeDeletion), ""},
{KeyStr("M", 4U, kTypeValue), "val"},
{KeyStr("M", 3U, kTypeSingleDeletion), ""},
});
AddMockFile(file2);
auto file3 = mock::MakeMockFile({
{KeyStr("D", 1U, kTypeValue), "val"},
{KeyStr("H", 1U, kTypeValue), "val"},
{KeyStr("I", 2U, kTypeValue), "val"},
{KeyStr("D", 11U, kTypeValue), "val"},
});
AddMockFile(file3, 2);
auto file4 = mock::MakeMockFile({
{KeyStr("M", 1U, kTypeValue), "val"},
auto expected_results = mock::MakeMockFile({
{KeyStr("A", 12U, kTypeSingleDeletion), ""},
{KeyStr("A", 10U, kTypeValue), "val"},
{KeyStr("C", 14U, kTypeValue), "val3"},
{KeyStr("D", 10U, kTypeSingleDeletion), ""},
});
AddMockFile(file4, 2);
auto expected_results =
mock::MakeMockFile({{KeyStr("A", 14U, kTypeSingleDeletion), ""},
{KeyStr("A", 13U, kTypeValue), ""},
{KeyStr("A", 12U, kTypeSingleDeletion), ""},
{KeyStr("A", 10U, kTypeValue), "val"},
{KeyStr("B", 14U, kTypeSingleDeletion), ""},
{KeyStr("B", 13U, kTypeValue), ""},
{KeyStr("C", 14U, kTypeValue), "val3"},
{KeyStr("D", 12U, kTypeSingleDeletion), ""},
{KeyStr("D", 11U, kTypeValue), ""},
{KeyStr("D", 10U, kTypeSingleDeletion), ""},
{KeyStr("E", 12U, kTypeSingleDeletion), ""},
{KeyStr("E", 11U, kTypeValue), ""},
{KeyStr("G", 15U, kTypeValue), "val"},
{KeyStr("G", 12U, kTypeSingleDeletion), ""},
{KeyStr("I", 14U, kTypeSingleDeletion), ""},
{KeyStr("I", 13U, kTypeValue), ""},
{KeyStr("J", 15U, kTypeValue), "val"},
{KeyStr("K", 16U, kTypeSingleDeletion), ""},
{KeyStr("K", 15U, kTypeValue), ""},
{KeyStr("K", 11U, kTypeSingleDeletion), ""},
{KeyStr("K", 8U, kTypeValue), "val3"},
{KeyStr("L", 16U, kTypeSingleDeletion), ""},
{KeyStr("L", 15U, kTypeValue), ""},
{KeyStr("M", 16U, kTypeDeletion), ""},
{KeyStr("M", 3U, kTypeSingleDeletion), ""}});
SetLastSequence(22U);
auto files = cfd_->current()->storage_info()->LevelFiles(0);
RunCompaction({files}, expected_results, {10U}, 10U);
RunCompaction({files}, expected_results, {10U});
}
// This test documents the behavior where a corrupt key follows a deletion or a
@@ -904,7 +684,7 @@ TEST_F(CompactionJobTest, CorruptionAfterDeletion) {
mock::MakeMockFile({{test::KeyStr("A", 0U, kTypeValue), "val3"},
{test::KeyStr("a", 0U, kTypeValue, true), "val"},
{test::KeyStr("b", 0U, kTypeValue, true), "val"},
{test::KeyStr("c", 1U, kTypeValue), "val2"}});
{test::KeyStr("c", 0U, kTypeValue), "val2"}});
SetLastSequence(6U);
auto files = cfd_->current()->storage_info()->LevelFiles(0);
+170 -382
View File
File diff suppressed because it is too large. Load diff
+46 -72
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -20,7 +20,8 @@
#include "rocksdb/env.h"
#include "rocksdb/options.h"
#include "rocksdb/status.h"
#include "util/cf_options.h"
#include "util/mutable_cf_options.h"
namespace rocksdb {
@@ -59,28 +60,29 @@ class CompactionPicker {
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
VersionStorageInfo* vstorage, int input_level, int output_level,
uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
InternalKey** compaction_end, bool* manual_conflict);
InternalKey** compaction_end);
// The maximum allowed output level. Default value is NumberLevels() - 1.
virtual int MaxOutputLevel() const { return NumberLevels() - 1; }
virtual int MaxOutputLevel() const {
return NumberLevels() - 1;
}
virtual bool NeedsCompaction(const VersionStorageInfo* vstorage) const = 0;
// Sanitize the input set of compaction input files.
// When the input parameters do not describe a valid compaction, the
// function will try to fix the input_files by adding necessary
// files. If it's not possible to conver an invalid input_files
// into a valid one by adding more files, the function will return a
// non-ok status with specific reason.
// Sanitize the input set of compaction input files.
// When the input parameters do not describe a valid compaction, the
// function will try to fix the input_files by adding necessary
// files. If it's not possible to conver an invalid input_files
// into a valid one by adding more files, the function will return a
// non-ok status with specific reason.
#ifndef ROCKSDB_LITE
Status SanitizeCompactionInputFiles(std::unordered_set<uint64_t>* input_files,
const ColumnFamilyMetaData& cf_meta,
const int output_level) const;
Status SanitizeCompactionInputFiles(
std::unordered_set<uint64_t>* input_files,
const ColumnFamilyMetaData& cf_meta,
const int output_level) const;
#endif // ROCKSDB_LITE
// Free up the files that participated in a compaction
//
// Requirement: DB mutex held
void ReleaseCompactionFiles(Compaction* c, Status status);
// Returns true if any one of the specified files are being compacted
@@ -113,33 +115,21 @@ class CompactionPicker {
return !level0_compactions_in_progress_.empty();
}
// Return true if the passed key range overlap with a compaction output
// that is currently running.
bool RangeOverlapWithCompaction(const Slice& smallest_user_key,
const Slice& largest_user_key,
int level) const;
protected:
int NumberLevels() const { return ioptions_.num_levels; }
// Stores the minimal range that covers all entries in inputs in
// *smallest, *largest.
// REQUIRES: inputs is not empty
void GetRange(const CompactionInputFiles& inputs, InternalKey* smallest,
InternalKey* largest) const;
void GetRange(const CompactionInputFiles& inputs,
InternalKey* smallest, InternalKey* largest);
// Stores the minimal range that covers all entries in inputs1 and inputs2
// in *smallest, *largest.
// REQUIRES: inputs is not empty
void GetRange(const CompactionInputFiles& inputs1,
const CompactionInputFiles& inputs2, InternalKey* smallest,
InternalKey* largest) const;
// Stores the minimal range that covers all entries in inputs
// in *smallest, *largest.
// REQUIRES: inputs is not empty (at least on entry have one file)
void GetRange(const std::vector<CompactionInputFiles>& inputs,
InternalKey* smallest, InternalKey* largest) const;
const CompactionInputFiles& inputs2,
InternalKey* smallest, InternalKey* largest);
// Add more files to the inputs on "level" to make sure that
// no newer version of a key is compacted to "level+1" while leaving an older
@@ -160,11 +150,6 @@ class CompactionPicker {
const InternalKey* smallest,
const InternalKey* largest, int level, int* index);
// Returns true if the key range that `inputs` files cover overlap with the
// key range of a currently running compaction.
bool FilesRangeOverlapWithCompaction(
const std::vector<CompactionInputFiles>& inputs, int level) const;
bool SetupOtherInputs(const std::string& cf_name,
const MutableCFOptions& mutable_cf_options,
VersionStorageInfo* vstorage,
@@ -179,28 +164,19 @@ class CompactionPicker {
const ImmutableCFOptions& ioptions_;
// A helper function to SanitizeCompactionInputFiles() that
// sanitizes "input_files" by adding necessary files.
// A helper function to SanitizeCompactionInputFiles() that
// sanitizes "input_files" by adding necessary files.
#ifndef ROCKSDB_LITE
virtual Status SanitizeCompactionInputFilesForAllLevels(
std::unordered_set<uint64_t>* input_files,
const ColumnFamilyMetaData& cf_meta, const int output_level) const;
const ColumnFamilyMetaData& cf_meta,
const int output_level) const;
#endif // ROCKSDB_LITE
// Register this compaction in the set of running compactions
void RegisterCompaction(Compaction* c);
// Remove this compaction from the set of running compactions
void UnregisterCompaction(Compaction* c);
// Keeps track of all compactions that are running on Level0.
// Protected by DB mutex
// It is protected by DB mutex
std::set<Compaction*> level0_compactions_in_progress_;
// Keeps track of all compactions that are running.
// Protected by DB mutex
std::unordered_set<Compaction*> compactions_in_progress_;
const InternalKeyComparator* const icmp_;
};
@@ -214,8 +190,8 @@ class LevelCompactionPicker : public CompactionPicker {
VersionStorageInfo* vstorage,
LogBuffer* log_buffer) override;
virtual bool NeedsCompaction(
const VersionStorageInfo* vstorage) const override;
virtual bool NeedsCompaction(const VersionStorageInfo* vstorage) const
override;
// Pick a path ID to place a newly generated file, with its level
static uint32_t GetPathId(const ImmutableCFOptions& ioptions,
@@ -254,8 +230,8 @@ class UniversalCompactionPicker : public CompactionPicker {
virtual int MaxOutputLevel() const override { return NumberLevels() - 1; }
virtual bool NeedsCompaction(
const VersionStorageInfo* vstorage) const override;
virtual bool NeedsCompaction(const VersionStorageInfo* vstorage) const
override;
private:
struct SortedRun {
@@ -275,7 +251,7 @@ class UniversalCompactionPicker : public CompactionPicker {
// sorted_run_count is added into the string to print
void DumpSizeInfo(char* out_buf, size_t out_buf_size,
size_t sorted_run_count) const;
int sorted_run_count) const;
int level;
// `file` Will be null for level > 0. For level = 0, the sorted run is
@@ -326,20 +302,22 @@ class FIFOCompactionPicker : public CompactionPicker {
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
VersionStorageInfo* vstorage, int input_level, int output_level,
uint32_t output_path_id, const InternalKey* begin, const InternalKey* end,
InternalKey** compaction_end, bool* manual_conflict) override;
InternalKey** compaction_end) override;
// The maximum allowed output level. Always returns 0.
virtual int MaxOutputLevel() const override { return 0; }
virtual int MaxOutputLevel() const override {
return 0;
}
virtual bool NeedsCompaction(
const VersionStorageInfo* vstorage) const override;
virtual bool NeedsCompaction(const VersionStorageInfo* vstorage) const
override;
};
class NullCompactionPicker : public CompactionPicker {
public:
NullCompactionPicker(const ImmutableCFOptions& ioptions,
const InternalKeyComparator* icmp)
: CompactionPicker(ioptions, icmp) {}
const InternalKeyComparator* icmp) :
CompactionPicker(ioptions, icmp) {}
virtual ~NullCompactionPicker() {}
// Always return "nullptr"
@@ -351,27 +329,23 @@ class NullCompactionPicker : public CompactionPicker {
}
// Always return "nullptr"
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,
const InternalKey* begin, const InternalKey* end,
InternalKey** compaction_end,
bool* manual_conflict) override {
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, const InternalKey* begin, const InternalKey* end,
InternalKey** compaction_end) override {
return nullptr;
}
// Always returns false.
virtual bool NeedsCompaction(
const VersionStorageInfo* vstorage) const override {
virtual bool NeedsCompaction(const VersionStorageInfo* vstorage) const
override {
return false;
}
};
#endif // !ROCKSDB_LITE
CompressionType GetCompressionType(const ImmutableCFOptions& ioptions,
const VersionStorageInfo* vstorage,
const MutableCFOptions& mutable_cf_options,
int level, int base_level,
const bool enable_compression = true);
+29 -383
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -48,7 +48,7 @@ class CompactionPickerTest : public testing::Test {
: ucmp_(BytewiseComparator()),
icmp_(ucmp_),
ioptions_(options_),
mutable_cf_options_(options_),
mutable_cf_options_(options_, ioptions_),
level_compaction_picker(ioptions_, &icmp_),
cf_name_("dummy"),
log_buffer_(InfoLogLevel::INFO_LEVEL, &logger_),
@@ -66,8 +66,8 @@ class CompactionPickerTest : public testing::Test {
void NewVersionStorage(int num_levels, CompactionStyle style) {
DeleteVersionStorage();
options_.num_levels = num_levels;
vstorage_.reset(new VersionStorageInfo(&icmp_, ucmp_, options_.num_levels,
style, nullptr, false));
vstorage_.reset(new VersionStorageInfo(
&icmp_, ucmp_, options_.num_levels, style, nullptr));
vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
}
@@ -79,7 +79,7 @@ class CompactionPickerTest : public testing::Test {
}
void Add(int level, uint32_t file_number, const char* smallest,
const char* largest, uint64_t file_size = 1, uint32_t path_id = 0,
const char* largest, uint64_t file_size = 0, uint32_t path_id = 0,
SequenceNumber smallest_seq = 100,
SequenceNumber largest_seq = 100) {
assert(level < vstorage_->num_levels());
@@ -115,13 +115,12 @@ class CompactionPickerTest : public testing::Test {
void UpdateVersionStorageInfo() {
vstorage_->CalculateBaseBytes(ioptions_, mutable_cf_options_);
vstorage_->UpdateFilesByCompactionPri(ioptions_.compaction_pri);
vstorage_->UpdateFilesByCompactionPri(mutable_cf_options_);
vstorage_->UpdateNumNonEmptyLevels();
vstorage_->GenerateFileIndexer();
vstorage_->GenerateLevelFilesBrief();
vstorage_->ComputeCompactionScore(ioptions_, mutable_cf_options_);
vstorage_->ComputeCompactionScore(mutable_cf_options_, fifo_options_);
vstorage_->GenerateLevel0NonOverlapping();
vstorage_->ComputeFilesMarkedForCompaction();
vstorage_->SetFinalized();
}
};
@@ -196,7 +195,6 @@ TEST_F(CompactionPickerTest, LevelMaxScore) {
NewVersionStorage(6, kCompactionStyleLevel);
mutable_cf_options_.target_file_size_base = 10000000;
mutable_cf_options_.target_file_size_multiplier = 10;
mutable_cf_options_.max_bytes_for_level_base = 10 * 1024 * 1024;
Add(0, 1U, "150", "200", 1000000000U);
// Level 1 score 1.2
Add(1, 66U, "150", "200", 6000000U);
@@ -321,7 +319,6 @@ TEST_F(CompactionPickerTest, Level0TriggerDynamic4) {
mutable_cf_options_.level0_file_num_compaction_trigger = 2;
mutable_cf_options_.max_bytes_for_level_base = 200;
mutable_cf_options_.max_bytes_for_level_multiplier = 10;
NewVersionStorage(num_levels, kCompactionStyleLevel);
Add(0, 1U, "150", "200");
Add(0, 2U, "200", "250");
@@ -351,7 +348,6 @@ TEST_F(CompactionPickerTest, Level0TriggerDynamic4) {
TEST_F(CompactionPickerTest, LevelTriggerDynamic4) {
int num_levels = ioptions_.num_levels;
ioptions_.level_compaction_dynamic_level_bytes = true;
ioptions_.compaction_pri = kMinOverlappingRatio;
mutable_cf_options_.level0_file_num_compaction_trigger = 2;
mutable_cf_options_.max_bytes_for_level_base = 200;
mutable_cf_options_.max_bytes_for_level_multiplier = 10;
@@ -370,9 +366,11 @@ TEST_F(CompactionPickerTest, LevelTriggerDynamic4) {
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(1U, compaction->num_input_files(0));
ASSERT_EQ(5U, compaction->input(0, 0)->fd.GetNumber());
ASSERT_EQ(0, compaction->num_input_files(1));
ASSERT_EQ(1U, compaction->num_input_levels());
ASSERT_EQ(6U, compaction->input(0, 0)->fd.GetNumber());
ASSERT_EQ(2U, compaction->num_input_files(1));
ASSERT_EQ(3U, compaction->input(1, 0)->fd.GetNumber());
ASSERT_EQ(4U, compaction->input(1, 1)->fd.GetNumber());
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_EQ(num_levels - 1, compaction->output_level());
}
@@ -468,6 +466,7 @@ TEST_F(CompactionPickerTest, NeedsCompactionFIFO) {
fifo_options_.max_table_files_size = kMaxSize;
ioptions_.compaction_options_fifo = fifo_options_;
FIFOCompactionPicker fifo_compaction_picker(ioptions_, &icmp_);
UpdateVersionStorageInfo();
// must return false when there's no files.
ASSERT_EQ(fifo_compaction_picker.NeedsCompaction(vstorage_.get()), false);
@@ -488,90 +487,6 @@ TEST_F(CompactionPickerTest, NeedsCompactionFIFO) {
}
#endif // ROCKSDB_LITE
TEST_F(CompactionPickerTest, CompactionPriMinOverlapping1) {
NewVersionStorage(6, kCompactionStyleLevel);
ioptions_.compaction_pri = kMinOverlappingRatio;
mutable_cf_options_.target_file_size_base = 10000000;
mutable_cf_options_.target_file_size_multiplier = 10;
mutable_cf_options_.max_bytes_for_level_base = 10 * 1024 * 1024;
Add(2, 6U, "150", "179", 50000000U);
Add(2, 7U, "180", "220", 50000000U);
Add(2, 8U, "321", "400", 50000000U); // File not overlapping
Add(2, 9U, "721", "800", 50000000U);
Add(3, 26U, "150", "170", 260000000U);
Add(3, 27U, "171", "179", 260000000U);
Add(3, 28U, "191", "220", 260000000U);
Add(3, 29U, "221", "300", 260000000U);
Add(3, 30U, "750", "900", 260000000U);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(1U, compaction->num_input_files(0));
// Pick file 8 because it overlaps with 0 files on level 3.
ASSERT_EQ(8U, compaction->input(0, 0)->fd.GetNumber());
}
TEST_F(CompactionPickerTest, CompactionPriMinOverlapping2) {
NewVersionStorage(6, kCompactionStyleLevel);
ioptions_.compaction_pri = kMinOverlappingRatio;
mutable_cf_options_.target_file_size_base = 10000000;
mutable_cf_options_.target_file_size_multiplier = 10;
mutable_cf_options_.max_bytes_for_level_base = 10 * 1024 * 1024;
Add(2, 6U, "150", "175",
60000000U); // Overlaps with file 26, 27, total size 521M
Add(2, 7U, "176", "200", 60000000U); // Overlaps with file 27, 28, total size
// 520M, the smalelst overlapping
Add(2, 8U, "201", "300",
60000000U); // Overlaps with file 28, 29, total size 521M
Add(3, 26U, "100", "110", 261000000U);
Add(3, 26U, "150", "170", 261000000U);
Add(3, 27U, "171", "179", 260000000U);
Add(3, 28U, "191", "220", 260000000U);
Add(3, 29U, "221", "300", 261000000U);
Add(3, 30U, "321", "400", 261000000U);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(1U, compaction->num_input_files(0));
// Picking file 7 because overlapping ratio is the biggest.
ASSERT_EQ(7U, compaction->input(0, 0)->fd.GetNumber());
}
TEST_F(CompactionPickerTest, CompactionPriMinOverlapping3) {
NewVersionStorage(6, kCompactionStyleLevel);
ioptions_.compaction_pri = kMinOverlappingRatio;
mutable_cf_options_.max_bytes_for_level_base = 10000000;
mutable_cf_options_.max_bytes_for_level_multiplier = 10;
// file 7 and 8 over lap with the same file, but file 8 is smaller so
// it will be picked.
Add(2, 6U, "150", "167", 60000000U); // Overlaps with file 26, 27
Add(2, 7U, "168", "169", 60000000U); // Overlaps with file 27
Add(2, 8U, "201", "300", 61000000U); // Overlaps with file 28, but the file
// itself is larger. Should be picked.
Add(3, 26U, "160", "165", 260000000U);
Add(3, 27U, "166", "170", 260000000U);
Add(3, 28U, "180", "400", 260000000U);
Add(3, 29U, "401", "500", 260000000U);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(1U, compaction->num_input_files(0));
// Picking file 8 because overlapping ratio is the biggest.
ASSERT_EQ(8U, compaction->input(0, 0)->fd.GetNumber());
}
// This test exhibits the bug where we don't properly reset parent_index in
// PickCompaction()
TEST_F(CompactionPickerTest, ParentIndexResetBug) {
@@ -599,8 +514,6 @@ TEST_F(CompactionPickerTest, ParentIndexResetBug) {
// ranges (with different sequence numbers) in the input files.
TEST_F(CompactionPickerTest, OverlappingUserKeys) {
NewVersionStorage(6, kCompactionStyleLevel);
ioptions_.compaction_pri = kByCompensatedSize;
Add(1, 1U, "100", "150", 1U);
// Overlapping user keys
Add(1, 2U, "200", "400", 1U);
@@ -622,9 +535,9 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys2) {
// Overlapping user keys on same level and output level
Add(1, 1U, "200", "400", 1000000000U);
Add(1, 2U, "400", "500", 1U, 0, 0);
Add(2, 3U, "000", "100", 1U);
Add(2, 4U, "100", "600", 1U, 0, 0);
Add(2, 5U, "600", "700", 1U, 0, 0);
Add(2, 3U, "400", "600", 1U);
// The following file is not in the compaction despite overlapping user keys
Add(2, 4U, "600", "700", 1U, 0, 0);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
@@ -632,12 +545,10 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys2) {
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_EQ(2U, compaction->num_input_files(0));
ASSERT_EQ(3U, compaction->num_input_files(1));
ASSERT_EQ(1U, compaction->num_input_files(1));
ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
ASSERT_EQ(2U, compaction->input(0, 1)->fd.GetNumber());
ASSERT_EQ(3U, compaction->input(1, 0)->fd.GetNumber());
ASSERT_EQ(4U, compaction->input(1, 1)->fd.GetNumber());
ASSERT_EQ(5U, compaction->input(1, 2)->fd.GetNumber());
}
TEST_F(CompactionPickerTest, OverlappingUserKeys3) {
@@ -669,190 +580,6 @@ TEST_F(CompactionPickerTest, OverlappingUserKeys3) {
ASSERT_EQ(7U, compaction->input(1, 1)->fd.GetNumber());
}
TEST_F(CompactionPickerTest, OverlappingUserKeys4) {
NewVersionStorage(6, kCompactionStyleLevel);
mutable_cf_options_.max_bytes_for_level_base = 1000000;
Add(1, 1U, "100", "150", 1U);
Add(1, 2U, "150", "199", 1U, 0, 0);
Add(1, 3U, "200", "250", 1100000U, 0, 0);
Add(1, 4U, "251", "300", 1U, 0, 0);
Add(1, 5U, "300", "350", 1U, 0, 0);
Add(2, 6U, "100", "115", 1U);
Add(2, 7U, "125", "325", 1U);
Add(2, 8U, "350", "400", 1U);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_EQ(1U, compaction->num_input_files(0));
ASSERT_EQ(1U, compaction->num_input_files(1));
ASSERT_EQ(3U, compaction->input(0, 0)->fd.GetNumber());
ASSERT_EQ(7U, compaction->input(1, 0)->fd.GetNumber());
}
TEST_F(CompactionPickerTest, OverlappingUserKeys5) {
NewVersionStorage(6, kCompactionStyleLevel);
// Overlapping user keys on same level and output level
Add(1, 1U, "200", "400", 1000000000U);
Add(1, 2U, "400", "500", 1U, 0, 0);
Add(2, 3U, "000", "100", 1U);
Add(2, 4U, "100", "600", 1U, 0, 0);
Add(2, 5U, "600", "700", 1U, 0, 0);
vstorage_->LevelFiles(2)[2]->being_compacted = true;
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() == nullptr);
}
TEST_F(CompactionPickerTest, OverlappingUserKeys6) {
NewVersionStorage(6, kCompactionStyleLevel);
// Overlapping user keys on same level and output level
Add(1, 1U, "200", "400", 1U, 0, 0);
Add(1, 2U, "401", "500", 1U, 0, 0);
Add(2, 3U, "000", "100", 1U);
Add(2, 4U, "100", "300", 1U, 0, 0);
Add(2, 5U, "305", "450", 1U, 0, 0);
Add(2, 6U, "460", "600", 1U, 0, 0);
Add(2, 7U, "600", "700", 1U, 0, 0);
vstorage_->LevelFiles(1)[0]->marked_for_compaction = true;
vstorage_->LevelFiles(1)[1]->marked_for_compaction = true;
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_EQ(1U, compaction->num_input_files(0));
ASSERT_EQ(3U, compaction->num_input_files(1));
}
TEST_F(CompactionPickerTest, OverlappingUserKeys7) {
NewVersionStorage(6, kCompactionStyleLevel);
mutable_cf_options_.max_compaction_bytes = 100000000000u;
// Overlapping user keys on same level and output level
Add(1, 1U, "200", "400", 1U, 0, 0);
Add(1, 2U, "401", "500", 1000000000U, 0, 0);
Add(2, 3U, "100", "250", 1U);
Add(2, 4U, "300", "600", 1U, 0, 0);
Add(2, 5U, "600", "800", 1U, 0, 0);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_GE(1U, compaction->num_input_files(0));
ASSERT_GE(2U, compaction->num_input_files(1));
// File 5 has to be included in the compaction
ASSERT_EQ(5U, compaction->inputs(1)->back()->fd.GetNumber());
}
TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri1) {
NewVersionStorage(6, kCompactionStyleLevel);
mutable_cf_options_.level0_file_num_compaction_trigger = 2;
mutable_cf_options_.max_bytes_for_level_base = 900000000U;
// 6 L0 files, score 3.
Add(0, 1U, "000", "400", 1U);
Add(0, 2U, "001", "400", 1U, 0, 0);
Add(0, 3U, "001", "400", 1000000000U, 0, 0);
Add(0, 31U, "001", "400", 1000000000U, 0, 0);
Add(0, 32U, "001", "400", 1000000000U, 0, 0);
Add(0, 33U, "001", "400", 1000000000U, 0, 0);
// L1 total size 2GB, score 2.2. If one file being comapcted, score 1.1.
Add(1, 4U, "050", "300", 1000000000U, 0, 0);
file_map_[4u].first->being_compacted = true;
Add(1, 5U, "301", "350", 1000000000U, 0, 0);
// Output level overlaps with the beginning and the end of the chain
Add(2, 6U, "050", "100", 1U);
Add(2, 7U, "300", "400", 1U);
// No compaction should be scheduled, if L0 has higher priority than L1
// but L0->L1 compaction is blocked by a file in L1 being compacted.
UpdateVersionStorageInfo();
ASSERT_EQ(0, vstorage_->CompactionScoreLevel(0));
ASSERT_EQ(1, vstorage_->CompactionScoreLevel(1));
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() == nullptr);
}
TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri2) {
NewVersionStorage(6, kCompactionStyleLevel);
mutable_cf_options_.level0_file_num_compaction_trigger = 2;
mutable_cf_options_.max_bytes_for_level_base = 900000000U;
// 6 L0 files, score 3.
Add(0, 1U, "000", "400", 1U);
Add(0, 2U, "001", "400", 1U, 0, 0);
Add(0, 3U, "001", "400", 1000000000U, 0, 0);
Add(0, 31U, "001", "400", 1000000000U, 0, 0);
Add(0, 32U, "001", "400", 1000000000U, 0, 0);
Add(0, 33U, "001", "400", 1000000000U, 0, 0);
// L1 total size 2GB, score 2.2. If one file being comapcted, score 1.1.
Add(1, 4U, "050", "300", 1000000000U, 0, 0);
Add(1, 5U, "301", "350", 1000000000U, 0, 0);
// Output level overlaps with the beginning and the end of the chain
Add(2, 6U, "050", "100", 1U);
Add(2, 7U, "300", "400", 1U);
// If no file in L1 being compacted, L0->L1 compaction will be scheduled.
UpdateVersionStorageInfo(); // being_compacted flag is cleared here.
ASSERT_EQ(0, vstorage_->CompactionScoreLevel(0));
ASSERT_EQ(1, vstorage_->CompactionScoreLevel(1));
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
}
TEST_F(CompactionPickerTest, NotScheduleL1IfL0WithHigherPri3) {
NewVersionStorage(6, kCompactionStyleLevel);
mutable_cf_options_.level0_file_num_compaction_trigger = 2;
mutable_cf_options_.max_bytes_for_level_base = 900000000U;
// 6 L0 files, score 3.
Add(0, 1U, "000", "400", 1U);
Add(0, 2U, "001", "400", 1U, 0, 0);
Add(0, 3U, "001", "400", 1000000000U, 0, 0);
Add(0, 31U, "001", "400", 1000000000U, 0, 0);
Add(0, 32U, "001", "400", 1000000000U, 0, 0);
Add(0, 33U, "001", "400", 1000000000U, 0, 0);
// L1 score more than 6.
Add(1, 4U, "050", "300", 1000000000U, 0, 0);
file_map_[4u].first->being_compacted = true;
Add(1, 5U, "301", "350", 1000000000U, 0, 0);
Add(1, 51U, "351", "400", 6000000000U, 0, 0);
// Output level overlaps with the beginning and the end of the chain
Add(2, 6U, "050", "100", 1U);
Add(2, 7U, "300", "400", 1U);
// If score in L1 is larger than L0, L1 compaction goes through despite
// there is pending L0 compaction.
UpdateVersionStorageInfo();
ASSERT_EQ(1, vstorage_->CompactionScoreLevel(0));
ASSERT_EQ(0, vstorage_->CompactionScoreLevel(1));
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
}
TEST_F(CompactionPickerTest, EstimateCompactionBytesNeeded1) {
int num_levels = ioptions_.num_levels;
ioptions_.level_compaction_dynamic_level_bytes = false;
@@ -867,22 +594,19 @@ TEST_F(CompactionPickerTest, EstimateCompactionBytesNeeded1) {
Add(1, 4U, "400", "500", 600);
Add(1, 5U, "600", "700", 600);
// Level 2 is less than target 10000 even added size of level 1
// Size ratio of L2/L1 is 9600 / 1200 = 8
Add(2, 6U, "150", "200", 2500);
Add(2, 7U, "201", "210", 2000);
Add(2, 8U, "300", "310", 2600);
Add(2, 8U, "300", "310", 2500);
Add(2, 9U, "400", "500", 2500);
// Level 3 exceeds target 100,000 of 1000
Add(3, 10U, "400", "500", 101000);
// Level 4 exceeds target 1,000,000 by 900 after adding size from level 3
// Size ratio L4/L3 is 9.9
// After merge from L3, L4 size is 1000900
Add(4, 11U, "400", "500", 999900);
Add(5, 11U, "400", "500", 8007200);
// Level 4 exceeds target 1,000,000 of 500 after adding size from level 3
Add(4, 11U, "400", "500", 999500);
Add(5, 11U, "400", "500", 8000000);
UpdateVersionStorageInfo();
ASSERT_EQ(200u * 9u + 10900u + 900u * 9,
ASSERT_EQ(2200u + 11000u + 5500u,
vstorage_->estimated_compaction_needed_bytes());
}
@@ -898,42 +622,17 @@ TEST_F(CompactionPickerTest, EstimateCompactionBytesNeeded2) {
Add(0, 4U, "150", "200", 200);
Add(0, 5U, "150", "200", 200);
Add(0, 6U, "150", "200", 200);
// Level 1 size will be 1400 after merging with L0
// Level 1 is over target by
Add(1, 7U, "400", "500", 200);
Add(1, 8U, "600", "700", 200);
// Level 2 is less than target 10000 even added size of level 1
Add(2, 9U, "150", "200", 9100);
// Level 3 over the target, but since level 4 is empty, we assume it will be
// a trivial move.
Add(2, 9U, "150", "200", 9500);
Add(3, 10U, "400", "500", 101000);
UpdateVersionStorageInfo();
// estimated L1->L2 merge: 400 * (9100.0 / 1400.0 + 1.0)
ASSERT_EQ(1400u + 3000u, vstorage_->estimated_compaction_needed_bytes());
}
TEST_F(CompactionPickerTest, EstimateCompactionBytesNeeded3) {
int num_levels = ioptions_.num_levels;
ioptions_.level_compaction_dynamic_level_bytes = false;
mutable_cf_options_.level0_file_num_compaction_trigger = 3;
mutable_cf_options_.max_bytes_for_level_base = 1000;
mutable_cf_options_.max_bytes_for_level_multiplier = 10;
NewVersionStorage(num_levels, kCompactionStyleLevel);
Add(0, 1U, "150", "200", 2000);
Add(0, 2U, "150", "200", 2000);
Add(0, 4U, "150", "200", 2000);
Add(0, 5U, "150", "200", 2000);
Add(0, 6U, "150", "200", 1000);
// Level 1 size will be 10000 after merging with L0
Add(1, 7U, "400", "500", 500);
Add(1, 8U, "600", "700", 500);
Add(2, 9U, "150", "200", 10000);
UpdateVersionStorageInfo();
ASSERT_EQ(10000u + 18000u, vstorage_->estimated_compaction_needed_bytes());
ASSERT_EQ(1400u + 4400u + 11000u,
vstorage_->estimated_compaction_needed_bytes());
}
TEST_F(CompactionPickerTest, EstimateCompactionBytesNeededDynamicLevel) {
@@ -957,14 +656,12 @@ TEST_F(CompactionPickerTest, EstimateCompactionBytesNeededDynamicLevel) {
// num_levels - 3 is over target by 100 + 1000
Add(num_levels - 3, 7U, "400", "500", 300);
Add(num_levels - 3, 8U, "600", "700", 300);
// num_levels - 2 is over target by 1100 + 200
Add(num_levels - 2, 9U, "150", "200", 5200);
// Level 2 is over target by 1100 + 100
Add(num_levels - 2, 9U, "150", "200", 5100);
UpdateVersionStorageInfo();
// Merging to the second last level: (5200 / 1600 + 1) * 1100
// Merging to the last level: (50000 / 6300 + 1) * 1300
ASSERT_EQ(1600u + 4675u + 11617u,
ASSERT_EQ(1600u + 12100u + 13200u,
vstorage_->estimated_compaction_needed_bytes());
}
@@ -1126,57 +823,6 @@ TEST_F(CompactionPickerTest, IsBottommostLevelTest) {
DeleteVersionStorage();
}
TEST_F(CompactionPickerTest, MaxCompactionBytesHit) {
mutable_cf_options_.max_bytes_for_level_base = 1000000u;
mutable_cf_options_.max_compaction_bytes = 800000u;
ioptions_.level_compaction_dynamic_level_bytes = false;
NewVersionStorage(6, kCompactionStyleLevel);
// A compaction should be triggered and pick file 2 and 5.
// It cannot expand because adding file 1 and 3, the compaction size will
// exceed mutable_cf_options_.max_bytes_for_level_base.
Add(1, 1U, "100", "150", 300000U);
Add(1, 2U, "151", "200", 300001U, 0, 0);
Add(1, 3U, "201", "250", 300000U, 0, 0);
Add(1, 4U, "251", "300", 300000U, 0, 0);
Add(2, 5U, "160", "256", 1U);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_EQ(1U, compaction->num_input_files(0));
ASSERT_EQ(1U, compaction->num_input_files(1));
ASSERT_EQ(2U, compaction->input(0, 0)->fd.GetNumber());
ASSERT_EQ(5U, compaction->input(1, 0)->fd.GetNumber());
}
TEST_F(CompactionPickerTest, MaxCompactionBytesNotHit) {
mutable_cf_options_.max_bytes_for_level_base = 1000000u;
mutable_cf_options_.max_compaction_bytes = 1000000u;
ioptions_.level_compaction_dynamic_level_bytes = false;
NewVersionStorage(6, kCompactionStyleLevel);
// A compaction should be triggered and pick file 2 and 5.
// and it expands to file 1 and 3 too.
Add(1, 1U, "100", "150", 300000U);
Add(1, 2U, "151", "200", 300001U, 0, 0);
Add(1, 3U, "201", "250", 300000U, 0, 0);
Add(1, 4U, "251", "300", 300000U, 0, 0);
Add(2, 5U, "000", "233", 1U);
UpdateVersionStorageInfo();
std::unique_ptr<Compaction> compaction(level_compaction_picker.PickCompaction(
cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
ASSERT_TRUE(compaction.get() != nullptr);
ASSERT_EQ(2U, compaction->num_input_levels());
ASSERT_EQ(3U, compaction->num_input_files(0));
ASSERT_EQ(1U, compaction->num_input_files(1));
ASSERT_EQ(1U, compaction->input(0, 0)->fd.GetNumber());
ASSERT_EQ(2U, compaction->input(0, 1)->fd.GetNumber());
ASSERT_EQ(3U, compaction->input(0, 2)->fd.GetNumber());
ASSERT_EQ(5U, compaction->input(1, 0)->fd.GetNumber());
}
} // namespace rocksdb
int main(int argc, char** argv) {
+1 -5
View File
@@ -1,6 +1,6 @@
// 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.
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -40,10 +40,6 @@ class KVIter : public Iterator {
virtual void Seek(const Slice& k) override {
iter_ = map_->lower_bound(k.ToString());
}
virtual void SeekForPrev(const Slice& k) override {
iter_ = map_->upper_bound(k.ToString());
Prev();
}
virtual void Next() override { ++iter_; }
virtual void Prev() override {
if (iter_ == map_->begin()) {
+1 -8
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -18,13 +18,6 @@ namespace rocksdb {
void CancelAllBackgroundWork(DB* db, bool wait) {
(dynamic_cast<DBImpl*>(db))->CancelAllBackgroundWork(wait);
}
Status DeleteFilesInRange(DB* db, ColumnFamilyHandle* column_family,
const Slice* begin, const Slice* end) {
return (dynamic_cast<DBImpl*>(db))
->DeleteFilesInRange(column_family, begin, end);
}
} // namespace rocksdb
#endif // ROCKSDB_LITE
+7 -10
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -11,7 +11,6 @@
#include "rocksdb/db.h"
#include <inttypes.h>
#include <errno.h>
#include <fcntl.h>
#include <sys/stat.h>
@@ -42,7 +41,6 @@ class CorruptionTest : public testing::Test {
CorruptionTest() {
tiny_cache_ = NewLRUCache(100);
options_.wal_recovery_mode = WALRecoveryMode::kTolerateCorruptedTailRecords;
options_.env = &env_;
dbname_ = test::TmpDir() + "/corruption_test";
DestroyDB(dbname_, options_);
@@ -106,8 +104,8 @@ class CorruptionTest : public testing::Test {
}
void Check(int min_expected, int max_expected) {
uint64_t next_expected = 0;
uint64_t missed = 0;
unsigned int next_expected = 0;
int missed = 0;
int bad_keys = 0;
int bad_values = 0;
int correct = 0;
@@ -128,7 +126,7 @@ class CorruptionTest : public testing::Test {
continue;
}
missed += (key - next_expected);
next_expected = key + 1;
next_expected = static_cast<unsigned int>(key + 1);
if (iter->value() != Value(static_cast<int>(key), &value_space)) {
bad_values++;
} else {
@@ -138,9 +136,8 @@ class CorruptionTest : public testing::Test {
delete iter;
fprintf(stderr,
"expected=%d..%d; got=%d; bad_keys=%d; bad_values=%d; missed=%llu\n",
min_expected, max_expected, correct, bad_keys, bad_values,
static_cast<unsigned long long>(missed));
"expected=%d..%d; got=%d; bad_keys=%d; bad_values=%d; missed=%d\n",
min_expected, max_expected, correct, bad_keys, bad_values, missed);
ASSERT_LE(min_expected, correct);
ASSERT_GE(max_expected, correct);
}
@@ -186,7 +183,7 @@ class CorruptionTest : public testing::Test {
FileType type;
std::string fname;
int picked_number = -1;
for (size_t i = 0; i < filenames.size(); i++) {
for (unsigned int i = 0; i < filenames.size(); i++) {
if (ParseFileName(filenames[i], &number, &type) &&
type == filetype &&
static_cast<int>(number) > picked_number) { // Pick latest file
+1 -2
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2014, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -41,7 +41,6 @@ class CuckooTableDBTest : public testing::Test {
options.memtable_factory.reset(NewHashLinkListRepFactory(4, 0, 3, true));
options.allow_mmap_reads = true;
options.create_if_missing = true;
options.allow_concurrent_memtable_write = false;
return options;
}
File diff suppressed because it is too large. Load diff
-579
View File
@@ -1,579 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same 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 <cstdlib>
#include "db/db_test_util.h"
#include "port/stack_trace.h"
#include "util/lru_cache.h"
namespace rocksdb {
class DBBlockCacheTest : public DBTestBase {
private:
size_t miss_count_ = 0;
size_t hit_count_ = 0;
size_t insert_count_ = 0;
size_t failure_count_ = 0;
size_t compressed_miss_count_ = 0;
size_t compressed_hit_count_ = 0;
size_t compressed_insert_count_ = 0;
size_t compressed_failure_count_ = 0;
public:
const size_t kNumBlocks = 10;
const size_t kValueSize = 100;
DBBlockCacheTest() : DBTestBase("/db_block_cache_test") {}
BlockBasedTableOptions GetTableOptions() {
BlockBasedTableOptions table_options;
// Set a small enough block size so that each key-value get its own block.
table_options.block_size = 1;
return table_options;
}
Options GetOptions(const BlockBasedTableOptions& table_options) {
Options options = CurrentOptions();
options.create_if_missing = true;
options.avoid_flush_during_recovery = false;
// options.compression = kNoCompression;
options.statistics = rocksdb::CreateDBStatistics();
options.table_factory.reset(new BlockBasedTableFactory(table_options));
return options;
}
void InitTable(const Options& options) {
std::string value(kValueSize, 'a');
for (size_t i = 0; i < kNumBlocks; i++) {
ASSERT_OK(Put(ToString(i), value.c_str()));
}
}
void RecordCacheCounters(const Options& options) {
miss_count_ = TestGetTickerCount(options, BLOCK_CACHE_MISS);
hit_count_ = TestGetTickerCount(options, BLOCK_CACHE_HIT);
insert_count_ = TestGetTickerCount(options, BLOCK_CACHE_ADD);
failure_count_ = TestGetTickerCount(options, BLOCK_CACHE_ADD_FAILURES);
compressed_miss_count_ =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_MISS);
compressed_hit_count_ =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_HIT);
compressed_insert_count_ =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_ADD);
compressed_failure_count_ =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_ADD_FAILURES);
}
void CheckCacheCounters(const Options& options, size_t expected_misses,
size_t expected_hits, size_t expected_inserts,
size_t expected_failures) {
size_t new_miss_count = TestGetTickerCount(options, BLOCK_CACHE_MISS);
size_t new_hit_count = TestGetTickerCount(options, BLOCK_CACHE_HIT);
size_t new_insert_count = TestGetTickerCount(options, BLOCK_CACHE_ADD);
size_t new_failure_count =
TestGetTickerCount(options, BLOCK_CACHE_ADD_FAILURES);
ASSERT_EQ(miss_count_ + expected_misses, new_miss_count);
ASSERT_EQ(hit_count_ + expected_hits, new_hit_count);
ASSERT_EQ(insert_count_ + expected_inserts, new_insert_count);
ASSERT_EQ(failure_count_ + expected_failures, new_failure_count);
miss_count_ = new_miss_count;
hit_count_ = new_hit_count;
insert_count_ = new_insert_count;
failure_count_ = new_failure_count;
}
void CheckCompressedCacheCounters(const Options& options,
size_t expected_misses,
size_t expected_hits,
size_t expected_inserts,
size_t expected_failures) {
size_t new_miss_count =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_MISS);
size_t new_hit_count =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_HIT);
size_t new_insert_count =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_ADD);
size_t new_failure_count =
TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_ADD_FAILURES);
ASSERT_EQ(compressed_miss_count_ + expected_misses, new_miss_count);
ASSERT_EQ(compressed_hit_count_ + expected_hits, new_hit_count);
ASSERT_EQ(compressed_insert_count_ + expected_inserts, new_insert_count);
ASSERT_EQ(compressed_failure_count_ + expected_failures, new_failure_count);
compressed_miss_count_ = new_miss_count;
compressed_hit_count_ = new_hit_count;
compressed_insert_count_ = new_insert_count;
compressed_failure_count_ = new_failure_count;
}
};
TEST_F(DBBlockCacheTest, TestWithoutCompressedBlockCache) {
ReadOptions read_options;
auto table_options = GetTableOptions();
auto options = GetOptions(table_options);
InitTable(options);
std::shared_ptr<Cache> cache = NewLRUCache(0, 0, false);
table_options.block_cache = cache;
options.table_factory.reset(new BlockBasedTableFactory(table_options));
Reopen(options);
RecordCacheCounters(options);
std::vector<std::unique_ptr<Iterator>> iterators(kNumBlocks - 1);
Iterator* iter = nullptr;
// Load blocks into cache.
for (size_t i = 0; i < kNumBlocks - 1; i++) {
iter = db_->NewIterator(read_options);
iter->Seek(ToString(i));
ASSERT_OK(iter->status());
CheckCacheCounters(options, 1, 0, 1, 0);
iterators[i].reset(iter);
}
size_t usage = cache->GetUsage();
ASSERT_LT(0, usage);
cache->SetCapacity(usage);
ASSERT_EQ(usage, cache->GetPinnedUsage());
// Test with strict capacity limit.
cache->SetStrictCapacityLimit(true);
iter = db_->NewIterator(read_options);
iter->Seek(ToString(kNumBlocks - 1));
ASSERT_TRUE(iter->status().IsIncomplete());
CheckCacheCounters(options, 1, 0, 0, 1);
delete iter;
iter = nullptr;
// Release interators and access cache again.
for (size_t i = 0; i < kNumBlocks - 1; i++) {
iterators[i].reset();
CheckCacheCounters(options, 0, 0, 0, 0);
}
ASSERT_EQ(0, cache->GetPinnedUsage());
for (size_t i = 0; i < kNumBlocks - 1; i++) {
iter = db_->NewIterator(read_options);
iter->Seek(ToString(i));
ASSERT_OK(iter->status());
CheckCacheCounters(options, 0, 1, 0, 0);
iterators[i].reset(iter);
}
}
#ifdef SNAPPY
TEST_F(DBBlockCacheTest, TestWithCompressedBlockCache) {
ReadOptions read_options;
auto table_options = GetTableOptions();
auto options = GetOptions(table_options);
options.compression = CompressionType::kSnappyCompression;
InitTable(options);
std::shared_ptr<Cache> cache = NewLRUCache(0, 0, false);
std::shared_ptr<Cache> compressed_cache = NewLRUCache(1 << 25, 0, false);
table_options.block_cache = cache;
table_options.block_cache_compressed = compressed_cache;
options.table_factory.reset(new BlockBasedTableFactory(table_options));
Reopen(options);
RecordCacheCounters(options);
std::vector<std::unique_ptr<Iterator>> iterators(kNumBlocks - 1);
Iterator* iter = nullptr;
// Load blocks into cache.
for (size_t i = 0; i < kNumBlocks - 1; i++) {
iter = db_->NewIterator(read_options);
iter->Seek(ToString(i));
ASSERT_OK(iter->status());
CheckCacheCounters(options, 1, 0, 1, 0);
CheckCompressedCacheCounters(options, 1, 0, 1, 0);
iterators[i].reset(iter);
}
size_t usage = cache->GetUsage();
ASSERT_LT(0, usage);
ASSERT_EQ(usage, cache->GetPinnedUsage());
size_t compressed_usage = compressed_cache->GetUsage();
ASSERT_LT(0, compressed_usage);
// Compressed block cache cannot be pinned.
ASSERT_EQ(0, compressed_cache->GetPinnedUsage());
// Set strict capacity limit flag. Now block will only load into compressed
// block cache.
cache->SetCapacity(usage);
cache->SetStrictCapacityLimit(true);
ASSERT_EQ(usage, cache->GetPinnedUsage());
iter = db_->NewIterator(read_options);
iter->Seek(ToString(kNumBlocks - 1));
ASSERT_TRUE(iter->status().IsIncomplete());
CheckCacheCounters(options, 1, 0, 0, 1);
CheckCompressedCacheCounters(options, 1, 0, 1, 0);
delete iter;
iter = nullptr;
// Clear strict capacity limit flag. This time we shall hit compressed block
// cache.
cache->SetStrictCapacityLimit(false);
iter = db_->NewIterator(read_options);
iter->Seek(ToString(kNumBlocks - 1));
ASSERT_OK(iter->status());
CheckCacheCounters(options, 1, 0, 1, 0);
CheckCompressedCacheCounters(options, 0, 1, 0, 0);
delete iter;
iter = nullptr;
}
#endif // SNAPPY
#ifndef ROCKSDB_LITE
// Make sure that when options.block_cache is set, after a new table is
// created its index/filter blocks are added to block cache.
TEST_F(DBBlockCacheTest, IndexAndFilterBlocksOfNewTableAddedToCache) {
Options options = CurrentOptions();
options.create_if_missing = true;
options.statistics = rocksdb::CreateDBStatistics();
BlockBasedTableOptions table_options;
table_options.cache_index_and_filter_blocks = true;
table_options.filter_policy.reset(NewBloomFilterPolicy(20));
options.table_factory.reset(new BlockBasedTableFactory(table_options));
CreateAndReopenWithCF({"pikachu"}, options);
ASSERT_OK(Put(1, "key", "val"));
// Create a new table.
ASSERT_OK(Flush(1));
// index/filter blocks added to block cache right after table creation.
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS));
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_MISS));
ASSERT_EQ(2, /* only index/filter were added */
TestGetTickerCount(options, BLOCK_CACHE_ADD));
ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_DATA_MISS));
uint64_t int_num;
ASSERT_TRUE(
dbfull()->GetIntProperty("rocksdb.estimate-table-readers-mem", &int_num));
ASSERT_EQ(int_num, 0U);
// Make sure filter block is in cache.
std::string value;
ReadOptions ropt;
db_->KeyMayExist(ReadOptions(), handles_[1], "key", &value);
// Miss count should remain the same.
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_MISS));
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_HIT));
db_->KeyMayExist(ReadOptions(), handles_[1], "key", &value);
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_MISS));
ASSERT_EQ(2, TestGetTickerCount(options, BLOCK_CACHE_FILTER_HIT));
// Make sure index block is in cache.
auto index_block_hit = TestGetTickerCount(options, BLOCK_CACHE_INDEX_HIT);
value = Get(1, "key");
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS));
ASSERT_EQ(index_block_hit + 1,
TestGetTickerCount(options, BLOCK_CACHE_INDEX_HIT));
value = Get(1, "key");
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS));
ASSERT_EQ(index_block_hit + 2,
TestGetTickerCount(options, BLOCK_CACHE_INDEX_HIT));
}
TEST_F(DBBlockCacheTest, IndexAndFilterBlocksStats) {
Options options = CurrentOptions();
options.create_if_missing = true;
options.statistics = rocksdb::CreateDBStatistics();
BlockBasedTableOptions table_options;
table_options.cache_index_and_filter_blocks = true;
// 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));
options.table_factory.reset(new BlockBasedTableFactory(table_options));
CreateAndReopenWithCF({"pikachu"}, options);
ASSERT_OK(Put(1, "key", "val"));
// Create a new table
ASSERT_OK(Flush(1));
size_t index_bytes_insert =
TestGetTickerCount(options, BLOCK_CACHE_INDEX_BYTES_INSERT);
size_t filter_bytes_insert =
TestGetTickerCount(options, BLOCK_CACHE_FILTER_BYTES_INSERT);
ASSERT_GT(index_bytes_insert, 0);
ASSERT_GT(filter_bytes_insert, 0);
ASSERT_EQ(cache->GetUsage(), index_bytes_insert + filter_bytes_insert);
// set the cache capacity to the current usage
cache->SetCapacity(index_bytes_insert + filter_bytes_insert);
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_INDEX_BYTES_EVICT), 0);
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_FILTER_BYTES_EVICT), 0);
ASSERT_OK(Put(1, "key2", "val"));
// Create a new table
ASSERT_OK(Flush(1));
// cache evicted old index and block entries
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_INDEX_BYTES_INSERT),
index_bytes_insert);
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_FILTER_BYTES_INSERT),
filter_bytes_insert);
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_INDEX_BYTES_EVICT),
index_bytes_insert);
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_FILTER_BYTES_EVICT),
filter_bytes_insert);
}
namespace {
// A mock cache wraps LRUCache, and record how many entries have been
// inserted for each priority.
class MockCache : public LRUCache {
public:
static uint32_t high_pri_insert_count;
static uint32_t low_pri_insert_count;
MockCache() : LRUCache(1 << 25, 0, false, 0.0) {}
virtual Status Insert(const Slice& key, void* value, size_t charge,
void (*deleter)(const Slice& key, void* value),
Handle** handle, Priority priority) override {
if (priority == Priority::LOW) {
low_pri_insert_count++;
} else {
high_pri_insert_count++;
}
return LRUCache::Insert(key, value, charge, deleter, handle, priority);
}
};
uint32_t MockCache::high_pri_insert_count = 0;
uint32_t MockCache::low_pri_insert_count = 0;
} // anonymous namespace
TEST_F(DBBlockCacheTest, IndexAndFilterBlocksCachePriority) {
for (auto priority : {Cache::Priority::LOW, Cache::Priority::HIGH}) {
Options options = CurrentOptions();
options.create_if_missing = true;
options.statistics = rocksdb::CreateDBStatistics();
BlockBasedTableOptions table_options;
table_options.cache_index_and_filter_blocks = true;
table_options.block_cache.reset(new MockCache());
table_options.filter_policy.reset(NewBloomFilterPolicy(20));
table_options.cache_index_and_filter_blocks_with_high_priority =
priority == Cache::Priority::HIGH ? true : false;
options.table_factory.reset(new BlockBasedTableFactory(table_options));
DestroyAndReopen(options);
MockCache::high_pri_insert_count = 0;
MockCache::low_pri_insert_count = 0;
// Create a new table.
ASSERT_OK(Put("foo", "value"));
ASSERT_OK(Put("bar", "value"));
ASSERT_OK(Flush());
ASSERT_EQ(1, NumTableFilesAtLevel(0));
// index/filter blocks added to block cache right after table creation.
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS));
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_MISS));
ASSERT_EQ(2, /* only index/filter were added */
TestGetTickerCount(options, BLOCK_CACHE_ADD));
ASSERT_EQ(0, TestGetTickerCount(options, BLOCK_CACHE_DATA_MISS));
if (priority == Cache::Priority::LOW) {
ASSERT_EQ(0, MockCache::high_pri_insert_count);
ASSERT_EQ(2, MockCache::low_pri_insert_count);
} else {
ASSERT_EQ(2, MockCache::high_pri_insert_count);
ASSERT_EQ(0, MockCache::low_pri_insert_count);
}
// Access data block.
ASSERT_EQ("value", Get("foo"));
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_INDEX_MISS));
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_FILTER_MISS));
ASSERT_EQ(3, /*adding data block*/
TestGetTickerCount(options, BLOCK_CACHE_ADD));
ASSERT_EQ(1, TestGetTickerCount(options, BLOCK_CACHE_DATA_MISS));
// Data block should be inserted with low priority.
if (priority == Cache::Priority::LOW) {
ASSERT_EQ(0, MockCache::high_pri_insert_count);
ASSERT_EQ(3, MockCache::low_pri_insert_count);
} else {
ASSERT_EQ(2, MockCache::high_pri_insert_count);
ASSERT_EQ(1, MockCache::low_pri_insert_count);
}
}
}
TEST_F(DBBlockCacheTest, ParanoidFileChecks) {
Options options = CurrentOptions();
options.create_if_missing = true;
options.statistics = rocksdb::CreateDBStatistics();
options.level0_file_num_compaction_trigger = 2;
options.paranoid_file_checks = true;
BlockBasedTableOptions table_options;
table_options.cache_index_and_filter_blocks = false;
table_options.filter_policy.reset(NewBloomFilterPolicy(20));
options.table_factory.reset(new BlockBasedTableFactory(table_options));
CreateAndReopenWithCF({"pikachu"}, options);
ASSERT_OK(Put(1, "1_key", "val"));
ASSERT_OK(Put(1, "9_key", "val"));
// Create a new table.
ASSERT_OK(Flush(1));
ASSERT_EQ(1, /* read and cache data block */
TestGetTickerCount(options, BLOCK_CACHE_ADD));
ASSERT_OK(Put(1, "1_key2", "val2"));
ASSERT_OK(Put(1, "9_key2", "val2"));
// Create a new SST file. This will further trigger a compaction
// and generate another file.
ASSERT_OK(Flush(1));
dbfull()->TEST_WaitForCompact();
ASSERT_EQ(3, /* Totally 3 files created up to now */
TestGetTickerCount(options, BLOCK_CACHE_ADD));
// After disabling options.paranoid_file_checks. NO further block
// is added after generating a new file.
ASSERT_OK(
dbfull()->SetOptions(handles_[1], {{"paranoid_file_checks", "false"}}));
ASSERT_OK(Put(1, "1_key3", "val3"));
ASSERT_OK(Put(1, "9_key3", "val3"));
ASSERT_OK(Flush(1));
ASSERT_OK(Put(1, "1_key4", "val4"));
ASSERT_OK(Put(1, "9_key4", "val4"));
ASSERT_OK(Flush(1));
dbfull()->TEST_WaitForCompact();
ASSERT_EQ(3, /* Totally 3 files created up to now */
TestGetTickerCount(options, BLOCK_CACHE_ADD));
}
TEST_F(DBBlockCacheTest, CompressedCache) {
if (!Snappy_Supported()) {
return;
}
int num_iter = 80;
// Run this test three iterations.
// Iteration 1: only a uncompressed block cache
// Iteration 2: only a compressed block cache
// Iteration 3: both block cache and compressed cache
// Iteration 4: both block cache and compressed cache, but DB is not
// compressed
for (int iter = 0; iter < 4; iter++) {
Options options = CurrentOptions();
options.write_buffer_size = 64 * 1024; // small write buffer
options.statistics = rocksdb::CreateDBStatistics();
BlockBasedTableOptions table_options;
switch (iter) {
case 0:
// only uncompressed block cache
table_options.block_cache = NewLRUCache(8 * 1024);
table_options.block_cache_compressed = nullptr;
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
break;
case 1:
// no block cache, only compressed cache
table_options.no_block_cache = true;
table_options.block_cache = nullptr;
table_options.block_cache_compressed = NewLRUCache(8 * 1024);
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
break;
case 2:
// both compressed and uncompressed block cache
table_options.block_cache = NewLRUCache(1024);
table_options.block_cache_compressed = NewLRUCache(8 * 1024);
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
break;
case 3:
// both block cache and compressed cache, but DB is not compressed
// also, make block cache sizes bigger, to trigger block cache hits
table_options.block_cache = NewLRUCache(1024 * 1024);
table_options.block_cache_compressed = NewLRUCache(8 * 1024 * 1024);
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
options.compression = kNoCompression;
break;
default:
ASSERT_TRUE(false);
}
CreateAndReopenWithCF({"pikachu"}, options);
// default column family doesn't have block cache
Options no_block_cache_opts;
no_block_cache_opts.statistics = options.statistics;
no_block_cache_opts = CurrentOptions(no_block_cache_opts);
BlockBasedTableOptions table_options_no_bc;
table_options_no_bc.no_block_cache = true;
no_block_cache_opts.table_factory.reset(
NewBlockBasedTableFactory(table_options_no_bc));
ReopenWithColumnFamilies(
{"default", "pikachu"},
std::vector<Options>({no_block_cache_opts, options}));
Random rnd(301);
// Write 8MB (80 values, each 100K)
ASSERT_EQ(NumTableFilesAtLevel(0, 1), 0);
std::vector<std::string> values;
std::string str;
for (int i = 0; i < num_iter; i++) {
if (i % 4 == 0) { // high compression ratio
str = RandomString(&rnd, 1000);
}
values.push_back(str);
ASSERT_OK(Put(1, Key(i), values[i]));
}
// flush all data from memtable so that reads are from block cache
ASSERT_OK(Flush(1));
for (int i = 0; i < num_iter; i++) {
ASSERT_EQ(Get(1, Key(i)), values[i]);
}
// check that we triggered the appropriate code paths in the cache
switch (iter) {
case 0:
// only uncompressed block cache
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_MISS), 0);
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_MISS), 0);
break;
case 1:
// no block cache, only compressed cache
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_MISS), 0);
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_MISS), 0);
break;
case 2:
// both compressed and uncompressed block cache
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_MISS), 0);
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_MISS), 0);
break;
case 3:
// both compressed and uncompressed block cache
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_MISS), 0);
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_HIT), 0);
ASSERT_GT(TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_MISS), 0);
// compressed doesn't have any hits since blocks are not compressed on
// storage
ASSERT_EQ(TestGetTickerCount(options, BLOCK_CACHE_COMPRESSED_HIT), 0);
break;
default:
ASSERT_TRUE(false);
}
options.create_if_missing = true;
DestroyAndReopen(options);
}
}
#endif // ROCKSDB_LITE
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
File diff suppressed because it is too large. Load diff
+15 -97
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -13,7 +13,6 @@
namespace rocksdb {
static int cfilter_count = 0;
static int cfilter_skips = 0;
// This is a static filter used for filtering
// kvs during the compaction process.
@@ -66,30 +65,6 @@ class DeleteISFilter : public CompactionFilter {
virtual const char* Name() const override { return "DeleteFilter"; }
};
// Skip x if floor(x/10) is even, use range skips. Requires that keys are
// zero-padded to length 10.
class SkipEvenFilter : public CompactionFilter {
public:
virtual Decision FilterV2(int level, const Slice& key, ValueType value_type,
const Slice& existing_value, std::string* new_value,
std::string* skip_until) const override {
cfilter_count++;
int i = std::stoi(key.ToString());
if (i / 10 % 2 == 0) {
char key_str[100];
snprintf(key_str, sizeof(key), "%010d", i / 10 * 10 + 10);
*skip_until = key_str;
++cfilter_skips;
return Decision::kRemoveAndSkipUntil;
}
return Decision::kKeep;
}
virtual bool IgnoreSnapshots() const override { return true; }
virtual const char* Name() const override { return "DeleteFilter"; }
};
class DelayFilter : public CompactionFilter {
public:
explicit DelayFilter(DBTestBase* d) : db_test(d) {}
@@ -199,20 +174,6 @@ class DeleteISFilterFactory : public CompactionFilterFactory {
virtual const char* Name() const override { return "DeleteFilterFactory"; }
};
class SkipEvenFilterFactory : public CompactionFilterFactory {
public:
virtual std::unique_ptr<CompactionFilter> CreateCompactionFilter(
const CompactionFilter::Context& context) override {
if (context.is_manual_compaction) {
return std::unique_ptr<CompactionFilter>(new SkipEvenFilter());
} else {
return std::unique_ptr<CompactionFilter>(nullptr);
}
}
virtual const char* Name() const override { return "SkipEvenFilterFactory"; }
};
class DelayFilterFactory : public CompactionFilterFactory {
public:
explicit DelayFilterFactory(DBTestBase* d) : db_test(d) {}
@@ -300,10 +261,8 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
int total = 0;
Arena arena;
{
InternalKeyComparator icmp(options.comparator);
RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
ScopedArenaIterator iter(
dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[1]));
dbfull()->NewInternalIterator(&arena, handles_[1]));
iter->SeekToFirst();
ASSERT_OK(iter->status());
while (iter->Valid()) {
@@ -390,10 +349,8 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
// level Lmax because this record is at the tip
count = 0;
{
InternalKeyComparator icmp(options.comparator);
RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
ScopedArenaIterator iter(
dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[1]));
dbfull()->NewInternalIterator(&arena, handles_[1]));
iter->SeekToFirst();
ASSERT_OK(iter->status());
while (iter->Valid()) {
@@ -411,10 +368,11 @@ TEST_F(DBTestCompactionFilter, CompactionFilter) {
// entries are deleted. The compaction should create bunch of 'DeleteFile'
// entries in VersionEdit, but none of the 'AddFile's.
TEST_F(DBTestCompactionFilter, CompactionFilterDeletesAll) {
Options options = CurrentOptions();
Options options;
options.compaction_filter_factory = std::make_shared<DeleteFilterFactory>();
options.disable_auto_compactions = true;
options.create_if_missing = true;
options = CurrentOptions(options);
DestroyAndReopen(options);
// put some data
@@ -442,10 +400,11 @@ TEST_F(DBTestCompactionFilter, CompactionFilterDeletesAll) {
TEST_F(DBTestCompactionFilter, CompactionFilterWithValueChange) {
do {
Options options = CurrentOptions();
Options options;
options.num_levels = 3;
options.compaction_filter_factory =
std::make_shared<ChangeFilterFactory>();
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Write 100K+1 keys, these are written to a few files
@@ -508,7 +467,8 @@ TEST_F(DBTestCompactionFilter, CompactionFilterWithMergeOperator) {
PutFixed64(&three, 3);
PutFixed64(&four, 4);
Options options = CurrentOptions();
Options options;
options = CurrentOptions(options);
options.create_if_missing = true;
options.merge_operator = MergeOperators::CreateUInt64AddOperator();
options.num_levels = 3;
@@ -609,10 +569,7 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
int count = 0;
int total = 0;
Arena arena;
InternalKeyComparator icmp(options.comparator);
RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
ScopedArenaIterator iter(
dbfull()->NewInternalIterator(&arena, &range_del_agg));
ScopedArenaIterator iter(dbfull()->NewInternalIterator(&arena));
iter->SeekToFirst();
ASSERT_OK(iter->status());
while (iter->Valid()) {
@@ -626,7 +583,7 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextManual) {
iter->Next();
}
ASSERT_EQ(total, 700);
ASSERT_EQ(count, 2);
ASSERT_EQ(count, 1);
}
}
#endif // ROCKSDB_LITE
@@ -664,10 +621,11 @@ TEST_F(DBTestCompactionFilter, CompactionFilterContextCfId) {
// Compaction filters should only be applied to records that are newer than the
// latest snapshot. This test inserts records and applies a delete filter.
TEST_F(DBTestCompactionFilter, CompactionFilterSnapshot) {
Options options = CurrentOptions();
Options options;
options.compaction_filter_factory = std::make_shared<DeleteFilterFactory>();
options.disable_auto_compactions = true;
options.create_if_missing = true;
options = CurrentOptions(options);
DestroyAndReopen(options);
// Put some data.
@@ -701,10 +659,11 @@ TEST_F(DBTestCompactionFilter, CompactionFilterSnapshot) {
// records newer than the snapshot will also be processed
TEST_F(DBTestCompactionFilter, CompactionFilterIgnoreSnapshot) {
std::string five = ToString(5);
Options options = CurrentOptions();
Options options;
options.compaction_filter_factory = std::make_shared<DeleteISFilterFactory>();
options.disable_auto_compactions = true;
options.create_if_missing = true;
options = CurrentOptions(options);
DestroyAndReopen(options);
// Put some data.
@@ -758,47 +717,6 @@ TEST_F(DBTestCompactionFilter, CompactionFilterIgnoreSnapshot) {
}
#endif // ROCKSDB_LITE
TEST_F(DBTestCompactionFilter, SkipUntil) {
Options options = CurrentOptions();
options.compaction_filter_factory = std::make_shared<SkipEvenFilterFactory>();
options.disable_auto_compactions = true;
options.create_if_missing = true;
DestroyAndReopen(options);
// Write 100K keys, these are written to a few files in L0.
for (int table = 0; table < 4; ++table) {
// Key ranges in tables are [0, 38], [106, 149], [212, 260], [318, 371].
for (int i = table * 6; i < 39 + table * 11; ++i) {
char key[100];
snprintf(key, sizeof(key), "%010d", table * 100 + i);
Put(key, std::to_string(table * 1000 + i));
}
Flush();
}
cfilter_skips = 0;
ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
// Numberof skips in tables: 2, 3, 3, 3.
ASSERT_EQ(11, cfilter_skips);
for (int table = 0; table < 4; ++table) {
for (int i = table * 6; i < 39 + table * 11; ++i) {
int k = table * 100 + i;
char key[100];
snprintf(key, sizeof(key), "%010d", table * 100 + i);
auto expected = std::to_string(table * 1000 + i);
std::string val;
Status s = db_->Get(ReadOptions(), key, &val);
if (k / 10 % 2 == 0) {
ASSERT_TRUE(s.IsNotFound());
} else {
ASSERT_OK(s);
ASSERT_EQ(expected, val);
}
}
}
}
} // namespace rocksdb
int main(int argc, char** argv) {
+69 -748
View File
File diff suppressed because it is too large. Load diff
+33 -44
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -125,19 +125,19 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
Options options = CurrentOptions();
options.create_if_missing = true;
options.db_write_buffer_size = 204800;
options.write_buffer_size = 20480;
options.db_write_buffer_size = 2048;
options.write_buffer_size = 2048;
options.max_write_buffer_number = 2;
options.level0_file_num_compaction_trigger = 2;
options.level0_slowdown_writes_trigger = 9999;
options.level0_stop_writes_trigger = 9999;
options.target_file_size_base = 9102;
options.target_file_size_base = 2048;
options.level_compaction_dynamic_level_bytes = true;
options.max_bytes_for_level_base = 40960;
options.max_bytes_for_level_base = 10240;
options.max_bytes_for_level_multiplier = 4;
options.max_background_compactions = 2;
options.num_levels = 5;
options.max_compaction_bytes = 0; // Force not expanding in compactions
options.expanded_compaction_factor = 0; // Force not expanding in compactions
BlockBasedTableOptions table_options;
table_options.block_size = 1024;
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
@@ -154,10 +154,10 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
ASSERT_EQ(4U, int_prop);
// Put about 28K to L0
// Put about 7K to L0
for (int i = 0; i < 70; i++) {
ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
RandomString(&rnd, 380)));
RandomString(&rnd, 80)));
}
ASSERT_OK(dbfull()->SetOptions({
{"disable_auto_compactions", "false"},
@@ -167,14 +167,14 @@ 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, base level
// Insert extra about 3.5K to L0. After they are compacted to L4, base level
// should be changed to L3.
ASSERT_OK(dbfull()->SetOptions({
{"disable_auto_compactions", "true"},
}));
for (int i = 0; i < 70; i++) {
ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
RandomString(&rnd, 380)));
RandomString(&rnd, 80)));
}
ASSERT_OK(dbfull()->SetOptions({
@@ -199,10 +199,10 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
ASSERT_OK(dbfull()->SetOptions({
{"disable_auto_compactions", "true"},
}));
// Write about 40K more
// Write about 10K more
for (int i = 0; i < 100; i++) {
ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
RandomString(&rnd, 380)));
RandomString(&rnd, 80)));
}
ASSERT_OK(dbfull()->SetOptions({
{"disable_auto_compactions", "false"},
@@ -218,16 +218,15 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
// 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
// We try to make last levels' targets to be 10K, 40K, 160K, add triggers
// another compaction from 40K->160K.
ASSERT_OK(dbfull()->SetOptions({
{"disable_auto_compactions", "true"},
}));
// Write about 650K more.
// Each file is about 11KB, with 9KB of data.
for (int i = 0; i < 1300; i++) {
// Write about 150K more
for (int i = 0; i < 1350; i++) {
ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
RandomString(&rnd, 380)));
RandomString(&rnd, 80)));
}
ASSERT_OK(dbfull()->SetOptions({
{"disable_auto_compactions", "false"},
@@ -237,40 +236,30 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesBase2) {
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();
rocksdb::SyncPoint::GetInstance()->LoadDependency({
{"CompactionJob::Run():Start", "DynamicLevelMaxBytesBase2:0"},
{"DynamicLevelMaxBytesBase2:1", "CompactionJob::Run():End"},
{"DynamicLevelMaxBytesBase2:compact_range_finish",
"FlushJob::WriteLevel0Table"},
});
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
for (int attempt = 0; attempt <= 20; attempt++) {
// Write about 5K more data with two flushes. It should be flush to level 2
// but when it is applied, base level is already 1.
for (int i = 0; i < 50; i++) {
ASSERT_OK(Put(Key(static_cast<int>(rnd.Uniform(kMaxKey))),
RandomString(&rnd, 80)));
}
Flush();
std::thread thread([this] {
TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:compact_range_start");
ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:compact_range_finish");
});
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)));
ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
if (int_prop == 2U) {
env_->SleepForMicroseconds(50000);
} else {
break;
}
}
TEST_SYNC_POINT("DynamicLevelMaxBytesBase2:1");
Flush();
thread.join();
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
rocksdb::SyncPoint::GetInstance()->ClearAllCallBacks();
env_->SleepForMicroseconds(200000);
ASSERT_TRUE(db_->GetIntProperty("rocksdb.base-level", &int_prop));
ASSERT_EQ(1U, int_prop);
}
@@ -295,7 +284,7 @@ TEST_F(DBTestDynamicLevel, DynamicLevelMaxBytesCompactRange) {
options.max_background_compactions = 1;
const int kNumLevels = 5;
options.num_levels = kNumLevels;
options.max_compaction_bytes = 1; // Force not expanding in compactions
options.expanded_compaction_factor = 0; // Force not expanding in compactions
BlockBasedTableOptions table_options;
table_options.block_size = 1024;
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
+12 -13
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -14,20 +14,19 @@
#endif
#include <inttypes.h>
#include <stdint.h>
#include <algorithm>
#include <string>
#include <stdint.h>
#include "db/db_impl.h"
#include "db/filename.h"
#include "db/job_context.h"
#include "db/version_set.h"
#include "port/port.h"
#include "rocksdb/db.h"
#include "rocksdb/env.h"
#include "util/file_util.h"
#include "port/port.h"
#include "util/mutexlock.h"
#include "util/sync_point.h"
#include "util/testharness.h"
#include "util/file_util.h"
namespace rocksdb {
@@ -35,10 +34,10 @@ Status DBImpl::DisableFileDeletions() {
InstrumentedMutexLock l(&mutex_);
++disable_delete_obsolete_files_;
if (disable_delete_obsolete_files_ == 1) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"File Deletions Disabled");
} else {
Log(InfoLogLevel::WARN_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::WARN_LEVEL, db_options_.info_log,
"File Deletions Disabled, but already disabled. Counter: %d",
disable_delete_obsolete_files_);
}
@@ -59,12 +58,12 @@ Status DBImpl::EnableFileDeletions(bool force) {
--disable_delete_obsolete_files_;
}
if (disable_delete_obsolete_files_ == 0) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"File Deletions Enabled");
should_purge_files = true;
FindObsoleteFiles(&job_context, true);
} else {
Log(InfoLogLevel::WARN_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::WARN_LEVEL, db_options_.info_log,
"File Deletions Enable, but not really enabled. Counter: %d",
disable_delete_obsolete_files_);
}
@@ -73,7 +72,7 @@ Status DBImpl::EnableFileDeletions(bool force) {
PurgeObsoleteFiles(job_context);
}
job_context.Clean();
LogFlush(immutable_db_options_.info_log);
LogFlush(db_options_.info_log);
return Status::OK();
}
@@ -84,6 +83,7 @@ int DBImpl::IsFileDeletionsEnabled() const {
Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
uint64_t* manifest_file_size,
bool flush_memtable) {
*manifest_file_size = 0;
mutex_.Lock();
@@ -110,7 +110,7 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
if (!status.ok()) {
mutex_.Unlock();
Log(InfoLogLevel::ERROR_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::ERROR_LEVEL, db_options_.info_log,
"Cannot Flush data %s\n", status.ToString().c_str());
return status;
}
@@ -126,7 +126,7 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
}
ret.clear();
ret.reserve(live.size() + 3); // *.sst + CURRENT + MANIFEST + OPTIONS
ret.reserve(live.size() + 2); //*.sst + CURRENT + MANIFEST
// create names of the live files. The names are not absolute
// paths, instead they are relative to dbname_;
@@ -136,7 +136,6 @@ Status DBImpl::GetLiveFiles(std::vector<std::string>& ret,
ret.push_back(CurrentFileName(""));
ret.push_back(DescriptorFileName("", versions_->manifest_file_number()));
ret.push_back(OptionsFileName("", versions_->options_file_number()));
// find length of manifest file while holding the mutex lock
*manifest_file_size = versions_->manifest_file_size();
-56
View File
@@ -1,56 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same 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/db_test_util.h"
#include "port/stack_trace.h"
#include "util/sync_point.h"
namespace rocksdb {
class DBFlushTest : public DBTestBase {
public:
DBFlushTest() : DBTestBase("/db_flush_test") {}
};
// 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) {
Options options;
options.disable_auto_compactions = true;
options.max_background_flushes = 2;
Reopen(options);
FlushOptions no_wait;
no_wait.wait = false;
SyncPoint::GetInstance()->LoadDependency(
{{"VersionSet::LogAndApply:WriteManifest",
"DBFlushTest::FlushWhileWritingManifest:1"},
{"MemTableList::InstallMemtableFlushResults:InProgress",
"VersionSet::LogAndApply:WriteManifestDone"}});
SyncPoint::GetInstance()->EnableProcessing();
ASSERT_OK(Put("foo", "v"));
ASSERT_OK(dbfull()->Flush(no_wait));
TEST_SYNC_POINT("DBFlushTest::FlushWhileWritingManifest:1");
ASSERT_OK(Put("bar", "v"));
ASSERT_OK(dbfull()->Flush(no_wait));
// If the issue is hit we will wait here forever.
dbfull()->TEST_WaitForFlushMemTable();
#ifndef ROCKSDB_LITE
ASSERT_EQ(2, TotalTableFiles());
#endif // ROCKSDB_LITE
}
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
+1061 -2287
View File
File diff suppressed because it is too large. Load diff
+68 -328
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -10,11 +10,8 @@
#include <atomic>
#include <deque>
#include <functional>
#include <limits>
#include <list>
#include <map>
#include <queue>
#include <set>
#include <string>
#include <utility>
@@ -23,7 +20,6 @@
#include "db/column_family.h"
#include "db/compaction_job.h"
#include "db/dbformat.h"
#include "db/external_sst_file_ingestion_job.h"
#include "db/flush_job.h"
#include "db/flush_scheduler.h"
#include "db/internal_stats.h"
@@ -33,16 +29,15 @@
#include "db/wal_manager.h"
#include "db/write_controller.h"
#include "db/write_thread.h"
#include "db/writebuffer.h"
#include "memtable_list.h"
#include "port/port.h"
#include "rocksdb/db.h"
#include "rocksdb/env.h"
#include "rocksdb/memtablerep.h"
#include "rocksdb/transaction_log.h"
#include "rocksdb/write_buffer_manager.h"
#include "table/scoped_arena_iterator.h"
#include "util/autovector.h"
#include "util/db_options.h"
#include "util/event_logger.h"
#include "util/hash.h"
#include "util/instrumented_mutex.h"
@@ -60,7 +55,6 @@ class Arena;
class WriteCallback;
struct JobContext;
struct ExternalSstFileInfo;
struct MemTableInfo;
class DBImpl : public DB {
public:
@@ -125,10 +119,6 @@ class DBImpl : public DB {
using DB::GetProperty;
virtual bool GetProperty(ColumnFamilyHandle* column_family,
const Slice& property, std::string* value) override;
using DB::GetMapProperty;
virtual bool GetMapProperty(ColumnFamilyHandle* column_family,
const Slice& property,
std::map<std::string, double>* value) override;
using DB::GetIntProperty;
virtual bool GetIntProperty(ColumnFamilyHandle* column_family,
const Slice& property, uint64_t* value) override;
@@ -162,9 +152,6 @@ class DBImpl : public DB {
ColumnFamilyHandle* column_family,
const std::unordered_map<std::string, std::string>& options_map) override;
virtual Status SetDBOptions(
const std::unordered_map<std::string, std::string>& options_map) override;
using DB::NumberLevels;
virtual int NumberLevels(ColumnFamilyHandle* column_family) override;
using DB::MaxMemCompactionLevel;
@@ -175,9 +162,10 @@ class DBImpl : public DB {
virtual const std::string& GetName() const override;
virtual Env* GetEnv() const override;
using DB::GetOptions;
virtual Options GetOptions(ColumnFamilyHandle* column_family) const override;
virtual const Options& GetOptions(
ColumnFamilyHandle* column_family) const override;
using DB::GetDBOptions;
virtual DBOptions GetDBOptions() const override;
virtual const DBOptions& GetDBOptions() const override;
using DB::Flush;
virtual Status Flush(const FlushOptions& options,
ColumnFamilyHandle* column_family) override;
@@ -200,8 +188,6 @@ class DBImpl : public DB {
const TransactionLogIterator::ReadOptions&
read_options = TransactionLogIterator::ReadOptions()) override;
virtual Status DeleteFile(std::string name) override;
Status DeleteFilesInRange(ColumnFamilyHandle* column_family,
const Slice* begin, const Slice* end);
virtual void GetLiveFilesMetaData(
std::vector<LiveFileMetaData>* metadata) override;
@@ -240,46 +226,23 @@ class DBImpl : public DB {
bool include_history);
// For a given key, check to see if there are any records for this key
// in the memtables, including memtable history. If cache_only is false,
// SST files will also be checked.
//
// If a key is found, *found_record_for_key will be set to true and
// *seq will be set to the stored sequence number for the latest
// operation on this key or kMaxSequenceNumber if unknown.
// If no key is found, *found_record_for_key will be set to false.
//
// Note: If cache_only=false, it is possible for *seq to be set to 0 if
// the sequence number has been cleared from the record. If the caller is
// holding an active db snapshot, we know the missing sequence must be less
// than the snapshot's sequence number (sequence numbers are only cleared
// when there are no earlier active snapshots).
//
// If NotFound is returned and found_record_for_key is set to false, then no
// record for this key was found. If the caller is holding an active db
// snapshot, we know that no key could have existing after this snapshot
// (since we do not compact keys that have an earlier snapshot).
//
// Returns OK or NotFound on success,
// other status on unexpected error.
// TODO(andrewkr): this API need to be aware of range deletion operations
Status GetLatestSequenceForKey(SuperVersion* sv, const Slice& key,
bool cache_only, SequenceNumber* seq,
bool* found_record_for_key);
// in the memtables, including memtable history.
using DB::IngestExternalFile;
virtual Status IngestExternalFile(
ColumnFamilyHandle* column_family,
const std::vector<std::string>& external_files,
const IngestExternalFileOptions& ingestion_options) override;
// On success, *seq will contain the sequence number for the
// latest such change or kMaxSequenceNumber if no records were present.
// Returns OK on success, other status on error reading memtables.
Status GetLatestSequenceForKeyFromMemtable(SuperVersion* sv, const Slice& key,
SequenceNumber* seq);
using DB::AddFile;
virtual Status AddFile(ColumnFamilyHandle* column_family,
const ExternalSstFileInfo* file_info,
bool move_file) override;
virtual Status AddFile(ColumnFamilyHandle* column_family,
const std::string& file_path, bool move_file) override;
#endif // ROCKSDB_LITE
// Similar to GetSnapshot(), but also lets the db know that this snapshot
// will be used for transaction write-conflict checking. The DB can then
// make sure not to compact any keys that would prevent a write-conflict from
// being detected.
const Snapshot* GetSnapshotForWriteConflictBoundary();
// checks if all live files exist on file system and that their file sizes
// match to our in-memory records
virtual Status CheckConsistency();
@@ -289,15 +252,13 @@ class DBImpl : public DB {
Status RunManualCompaction(ColumnFamilyData* cfd, int input_level,
int output_level, uint32_t output_path_id,
const Slice* begin, const Slice* end,
bool exclusive,
bool disallow_trivial_move = false);
// Return an internal iterator over the current state of the database.
// The keys of this iterator are internal keys (see format.h).
// The returned iterator should be deleted when no longer needed.
InternalIterator* NewInternalIterator(
Arena* arena, RangeDelAggregator* range_del_agg,
ColumnFamilyHandle* column_family = nullptr);
Arena* arena, ColumnFamilyHandle* column_family = nullptr);
#ifndef NDEBUG
// Extra methods (for testing) that are not in the public DB interface
@@ -308,19 +269,8 @@ class DBImpl : public DB {
ColumnFamilyHandle* column_family = nullptr,
bool disallow_trivial_move = false);
void TEST_MaybeFlushColumnFamilies();
bool TEST_UnableToFlushOldestLog() {
return unable_to_flush_oldest_log_;
}
bool TEST_IsLogGettingFlushed() {
return alive_log_files_.begin()->getting_flushed;
}
// Force current memtable contents to be flushed.
Status TEST_FlushMemTable(bool wait = true,
ColumnFamilyHandle* cfh = nullptr);
Status TEST_FlushMemTable(bool wait = true);
// Wait for memtable compaction
Status TEST_WaitForFlushMemTable(ColumnFamilyHandle* column_family = nullptr);
@@ -361,37 +311,18 @@ class DBImpl : public DB {
uint64_t TEST_LogfileNumber();
uint64_t TEST_total_log_size() const { return total_log_size_; }
// Returns column family name to ImmutableCFOptions map.
Status TEST_GetAllImmutableCFOptions(
std::unordered_map<std::string, const ImmutableCFOptions*>* iopts_map);
// Return the lastest MutableCFOptions of of a column family
Status TEST_GetLatestMutableCFOptions(ColumnFamilyHandle* column_family,
MutableCFOptions* mutable_cf_opitons);
Cache* TEST_table_cache() { return table_cache_.get(); }
WriteController& TEST_write_controler() { return write_controller_; }
uint64_t TEST_FindMinLogContainingOutstandingPrep();
uint64_t TEST_FindMinPrepLogReferencedByMemTable();
int TEST_BGCompactionsAllowed() const;
const WriteController& TEST_write_controler() const {
return write_controller_;
}
#endif // NDEBUG
// Return maximum background compaction allowed to be scheduled based on
// compaction status.
int BGCompactionsAllowed() const;
// move logs pending closing from job_context to the DB queue and
// schedule a purge
void ScheduleBgLogWriterClose(JobContext* job_context);
uint64_t MinLogNumberToKeep();
// 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
@@ -404,19 +335,12 @@ class DBImpl : public DB {
// belong to live files are posibly removed. Also, removes all the
// files in sst_delete_files and log_delete_files.
// It is not necessary to hold the mutex when invoking this method.
void PurgeObsoleteFiles(const JobContext& background_contet,
bool schedule_only = false);
void SchedulePurge();
void PurgeObsoleteFiles(const JobContext& background_contet);
ColumnFamilyHandle* DefaultColumnFamily() const override;
const SnapshotList& snapshots() const { return snapshots_; }
const ImmutableDBOptions& immutable_db_options() const {
return immutable_db_options_;
}
void CancelAllBackgroundWork(bool wait);
// Find Super version and reference it. Based on options, it might return
@@ -469,83 +393,17 @@ class DBImpl : public DB {
return num_running_compactions_;
}
// hollow transactions shell used for recovery.
// these will then be passed to TransactionDB so that
// locks can be reacquired before writing can resume.
struct RecoveredTransaction {
uint64_t log_number_;
std::string name_;
WriteBatch* batch_;
explicit RecoveredTransaction(const uint64_t log, const std::string& name,
WriteBatch* batch)
: log_number_(log), name_(name), batch_(batch) {}
~RecoveredTransaction() { delete batch_; }
};
bool allow_2pc() const { return immutable_db_options_.allow_2pc; }
std::unordered_map<std::string, RecoveredTransaction*>
recovered_transactions() {
return recovered_transactions_;
}
RecoveredTransaction* GetRecoveredTransaction(const std::string& name) {
auto it = recovered_transactions_.find(name);
if (it == recovered_transactions_.end()) {
return nullptr;
} else {
return it->second;
}
}
void InsertRecoveredTransaction(const uint64_t log, const std::string& name,
WriteBatch* batch) {
recovered_transactions_[name] = new RecoveredTransaction(log, name, batch);
MarkLogAsContainingPrepSection(log);
}
void DeleteRecoveredTransaction(const std::string& name) {
auto it = recovered_transactions_.find(name);
assert(it != recovered_transactions_.end());
auto* trx = it->second;
recovered_transactions_.erase(it);
MarkLogAsHavingPrepSectionFlushed(trx->log_number_);
delete trx;
}
void DeleteAllRecoveredTransactions() {
for (auto it = recovered_transactions_.begin();
it != recovered_transactions_.end(); it++) {
delete it->second;
}
recovered_transactions_.clear();
}
void MarkLogAsHavingPrepSectionFlushed(uint64_t log);
void MarkLogAsContainingPrepSection(uint64_t log);
void AddToLogsToFreeQueue(log::Writer* log_writer) {
logs_to_free_queue_.push_back(log_writer);
}
Status NewDB();
protected:
Env* const env_;
const std::string dbname_;
unique_ptr<VersionSet> versions_;
const DBOptions initial_db_options_;
const ImmutableDBOptions immutable_db_options_;
MutableDBOptions mutable_db_options_;
const DBOptions db_options_;
Statistics* stats_;
std::unordered_map<std::string, RecoveredTransaction*>
recovered_transactions_;
InternalIterator* NewInternalIterator(const ReadOptions&,
ColumnFamilyData* cfd,
SuperVersion* super_version,
Arena* arena,
RangeDelAggregator* range_del_agg);
Arena* arena);
// Except in DB::Open(), WriteOptionsFile can only be called when:
// 1. WriteThread::Writer::EnterUnbatched() is used.
@@ -566,11 +424,6 @@ class DBImpl : public DB {
Compaction *c, const Status &st,
const CompactionJobStats& job_stats,
int job_id);
void NotifyOnMemTableSealed(ColumnFamilyData* cfd,
const MemTableInfo& mem_table_info);
void NotifyOnExternalFileIngested(
ColumnFamilyData* cfd, const ExternalSstFileIngestionJob& ingestion_job);
void NewThreadStatusCfInfo(ColumnFamilyData* cfd) const;
@@ -579,17 +432,11 @@ class DBImpl : public DB {
void EraseThreadStatusDbInfo() const;
Status WriteImpl(const WriteOptions& options, WriteBatch* updates,
WriteCallback* callback = nullptr,
uint64_t* log_used = nullptr, uint64_t log_ref = 0,
bool disable_memtable = false);
uint64_t FindMinLogContainingOutstandingPrep();
uint64_t FindMinPrepLogReferencedByMemTable();
WriteCallback* callback);
private:
friend class DB;
friend class InternalStats;
friend class TransactionImpl;
#ifndef ROCKSDB_LITE
friend class ForwardIterator;
#endif
@@ -602,14 +449,13 @@ class DBImpl : public DB {
struct WriteContext;
struct PurgeFileInfo;
Status NewDB();
// Recover the descriptor from persistent storage. May do a significant
// amount of work to recover recently logged updates. Any changes to
// be made to the descriptor are added to *edit.
Status Recover(const std::vector<ColumnFamilyDescriptor>& column_families,
bool read_only = false, bool error_if_log_file_exist = false,
bool error_if_data_exists_in_logs = false);
bool read_only = false, bool error_if_log_file_exist = false);
void MaybeIgnoreError(Status* s) const;
@@ -617,18 +463,13 @@ class DBImpl : public DB {
// Delete any unneeded files and stale in-memory entries.
void DeleteObsoleteFiles();
// Delete obsolete files and log status and information of file deletion
void DeleteObsoleteFileImpl(Status file_deletion_status, int job_id,
const std::string& fname, FileType type,
uint64_t number, uint32_t path_id);
// Background process needs to call
// auto x = CaptureCurrentFileNumberInPendingOutputs()
// auto file_num = versions_->NewFileNumber();
// <do something>
// ReleaseFileNumberFromPendingOutputs(x)
// This will protect any file with number `file_num` or greater from being
// deleted while <do something> is running.
// This will protect any temporary files created while <do something> is
// executing from being deleted.
// -----------
// This function will capture current file number and append it to
// pending_outputs_. This will prevent any background process to delete any
@@ -641,8 +482,6 @@ class DBImpl : public DB {
// and blocked by any other pending_outputs_ calls)
void ReleaseFileNumberFromPendingOutputs(std::list<uint64_t>::iterator v);
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.
Status FlushMemTableToOutputFile(ColumnFamilyData* cfd,
@@ -652,10 +491,10 @@ class DBImpl : public DB {
// REQUIRES: log_numbers are sorted in ascending order
Status RecoverLogFiles(const std::vector<uint64_t>& log_numbers,
SequenceNumber* next_sequence, bool read_only);
SequenceNumber* max_sequence, bool read_only);
// The following two methods are used to flush a memtable to
// storage. The first one is used at database RecoveryTime (when the
// storage. The first one is used atdatabase RecoveryTime (when the
// database is opened) and is heavyweight because it holds the mutex
// for the entire period. The second method WriteLevel0Table supports
// concurrent flush memtables to storage.
@@ -664,35 +503,27 @@ class DBImpl : public DB {
// 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);
Status DelayWrite(uint64_t num_bytes);
Status ScheduleFlushes(WriteContext* context);
Status SwitchMemtable(ColumnFamilyData* cfd, WriteContext* context);
// Force current memtable contents to be flushed.
Status FlushMemTable(ColumnFamilyData* cfd, const FlushOptions& options,
bool writes_stopped = false);
Status FlushMemTable(ColumnFamilyData* cfd, const FlushOptions& options);
// Wait for memtable flushed
Status WaitForFlushMemTable(ColumnFamilyData* cfd);
void RecordFlushIOStats();
void RecordCompactionIOStats();
#ifndef ROCKSDB_LITE
Status CompactFilesImpl(const CompactionOptions& compact_options,
ColumnFamilyData* cfd, Version* version,
const std::vector<std::string>& input_file_names,
const int output_level, int output_path_id,
JobContext* job_context, LogBuffer* log_buffer);
// Wait for current IngestExternalFile() calls to finish.
// REQUIRES: mutex_ held
void WaitForIngestFile();
#else
// IngestExternalFile is not supported in ROCKSDB_LITE so this function
// will be no-op
void WaitForIngestFile() {}
Status CompactFilesImpl(
const CompactionOptions& compact_options, ColumnFamilyData* cfd,
Version* version, const std::vector<std::string>& input_file_names,
const int output_level, int output_path_id, JobContext* job_context,
LogBuffer* log_buffer);
#endif // ROCKSDB_LITE
ColumnFamilyData* GetColumnFamilyDataByName(const std::string& cf_name);
@@ -700,17 +531,12 @@ class DBImpl : public DB {
void MaybeScheduleFlushOrCompaction();
void SchedulePendingFlush(ColumnFamilyData* cfd);
void SchedulePendingCompaction(ColumnFamilyData* cfd);
void SchedulePendingPurge(std::string fname, FileType type, uint64_t number,
uint32_t path_id, int job_id);
static void BGWorkCompaction(void* arg);
static void BGWorkCompaction(void* db);
static void BGWorkFlush(void* db);
static void BGWorkPurge(void* arg);
static void UnscheduleCallback(void* arg);
void BackgroundCallCompaction(void* arg);
void BackgroundCallCompaction();
void BackgroundCallFlush();
void BackgroundCallPurge();
Status BackgroundCompaction(bool* madeProgress, JobContext* job_context,
LogBuffer* log_buffer, void* m = 0);
LogBuffer* log_buffer);
Status BackgroundFlush(bool* madeProgress, JobContext* job_context,
LogBuffer* log_buffer);
@@ -738,16 +564,6 @@ class DBImpl : public DB {
// helper function to call after some of the logs_ were synced
void MarkLogsSynced(uint64_t up_to, bool synced_dir, const Status& status);
const Snapshot* GetSnapshotImpl(bool is_write_conflict_boundary);
// Persist RocksDB options under the single write thread
// REQUIRES: mutex locked
Status PersistOptions();
void MaybeFlushColumnFamilies();
uint64_t GetMaxTotalWalSize() const;
// table_cache_ provides its own synchronization
std::shared_ptr<Cache> table_cache_;
@@ -759,18 +575,17 @@ class DBImpl : public DB {
// same time.
InstrumentedMutex options_files_mutex_;
// State below is protected by mutex_
mutable InstrumentedMutex mutex_;
InstrumentedMutex mutex_;
std::atomic<bool> shutting_down_;
// This condition variable is signaled on these conditions:
// * whenever bg_compaction_scheduled_ goes down to 0
// * if AnyManualCompaction, whenever a compaction finishes, even if it hasn't
// * if bg_manual_only_ > 0, whenever a compaction finishes, even if it hasn't
// made any progress
// * whenever a compaction made any progress
// * whenever bg_flush_scheduled_ or bg_purge_scheduled_ value decreases
// (i.e. whenever a flush is done, even if it didn't make any progress)
// * whenever there is an error in background purge, flush or compaction
// * whenever num_running_ingest_file_ goes to 0.
// * whenever bg_flush_scheduled_ value decreases (i.e. whenever a flush is
// done, even if it didn't make any progress)
// * whenever there is an error in background flush or compaction
InstrumentedCondVar bg_cv_;
uint64_t logfile_number_;
std::deque<uint64_t>
@@ -863,7 +678,7 @@ class DBImpl : public DB {
Directories directories_;
WriteBufferManager* write_buffer_manager_;
WriteBuffer write_buffer_;
WriteThread write_thread_;
@@ -890,19 +705,6 @@ class DBImpl : public DB {
// State is protected with db mutex.
std::list<uint64_t> pending_outputs_;
// PurgeFileInfo is a structure to hold information of files to be deleted in
// purge_queue_
struct PurgeFileInfo {
std::string fname;
FileType type;
uint64_t number;
uint32_t path_id;
int job_id;
PurgeFileInfo(std::string fn, FileType t, uint64_t num, uint32_t pid,
int jid)
: fname(fn), type(t), number(num), path_id(pid), job_id(jid) {}
};
// flush_queue_ and compaction_queue_ hold column families that we need to
// flush and compact, respectively.
// A column family is inserted into flush_queue_ when it satisfies condition
@@ -927,12 +729,6 @@ class DBImpl : public DB {
// invariant(column family present in compaction_queue_ <==>
// ColumnFamilyData::pending_compaction_ == true)
std::deque<ColumnFamilyData*> compaction_queue_;
// A queue to store filenames of the files to be purged
std::deque<PurgeFileInfo> purge_queue_;
// A queue to store log writers to close
std::deque<log::Writer*> logs_to_free_queue_;
int unscheduled_flushes_;
int unscheduled_compactions_;
@@ -942,40 +738,32 @@ class DBImpl : public DB {
// stores the number of compactions are currently running
int num_running_compactions_;
// If non-zero, MaybeScheduleFlushOrCompaction() will only schedule manual
// compactions (if manual_compaction_ is not null). This mechanism enables
// manual compactions to wait until all other compactions are finished.
int bg_manual_only_;
// number of background memtable flush jobs, submitted to the HIGH pool
int bg_flush_scheduled_;
// stores the number of flushes are currently running
int num_running_flushes_;
// number of background obsolete file purge jobs, submitted to the HIGH pool
int bg_purge_scheduled_;
// Information for a manual compaction
struct ManualCompaction {
ColumnFamilyData* cfd;
int input_level;
int output_level;
uint32_t output_path_id;
Status status;
bool done;
Status status;
bool in_progress; // compaction request being processed?
bool incomplete; // only part of requested range compacted
bool exclusive; // current behavior of only one manual
bool disallow_trivial_move; // Force actual compaction to run
const InternalKey* begin; // nullptr means beginning of key range
const InternalKey* end; // nullptr means end of key range
InternalKey* manual_end; // how far we are compacting
InternalKey tmp_storage; // Used to keep track of compaction progress
InternalKey tmp_storage1; // Used to keep track of compaction progress
Compaction* compaction;
};
std::deque<ManualCompaction*> manual_compaction_dequeue_;
struct CompactionArg {
DBImpl* db;
ManualCompaction* m;
bool disallow_trivial_move; // Force actual compaction to run
};
ManualCompaction* manual_compaction_;
// Have we encountered a background error in paranoid mode?
Status bg_error_;
@@ -998,19 +786,7 @@ class DBImpl : public DB {
// they're unique
std::atomic<int> next_job_id_;
// A flag indicating whether the current rocksdb database has any
// data that is not yet persisted into either WAL or SST file.
// Used when disableWAL is true.
bool has_unpersisted_data_;
// if an attempt was made to flush all column families that
// the oldest log depends on but uncommited data in the oldest
// log prevents the log from being released.
// We must attempt to free the dependent memtables again
// at a later time after the transaction in the oldest
// log is fully commited.
bool unable_to_flush_oldest_log_;
bool flush_on_destroy_; // Used when disableWAL is true.
static const int KEEP_LOG_FILE_NUM = 1000;
// MSVC version 1800 still does not have constexpr for ::max()
@@ -1021,10 +797,6 @@ class DBImpl : public DB {
// The options to access storage files
const EnvOptions env_options_;
// Number of running IngestExternalFile() calls.
// REQUIRES: mutex held
int num_running_ingest_file_;
#ifndef ROCKSDB_LITE
WalManager wal_manager_;
#endif // ROCKSDB_LITE
@@ -1032,40 +804,15 @@ class DBImpl : public DB {
// Unified interface for logging events
EventLogger event_logger_;
// A value of > 0 temporarily disables scheduling of background work
// A value of >0 temporarily disables scheduling of background work
int bg_work_paused_;
// A value of > 0 temporarily disables scheduling of background compaction
int bg_compaction_paused_;
// Guard against multiple concurrent refitting
bool refitting_level_;
// Indicate DB was opened successfully
bool opened_successfully_;
// minmum log number still containing prepared data.
// this is used by FindObsoleteFiles to determine which
// flushed logs we must keep around because they still
// contain prepared data which has not been flushed or rolled back
std::priority_queue<uint64_t, std::vector<uint64_t>, std::greater<uint64_t>>
min_log_with_prep_;
// to be used in conjunction with min_log_with_prep_.
// once a transaction with data in log L is committed or rolled back
// rather than removing the value from the heap we add that value
// to prepared_section_completed_ which maps LOG -> instance_count
// since a log could contain multiple prepared sections
//
// when trying to determine the minmum log still active we first
// consult min_log_with_prep_. while that root value maps to
// a value > 0 in prepared_section_completed_ we decrement the
// instance_count for that log and pop the root value in
// min_log_with_prep_. This will work the same as a min_heap
// where we are deleteing arbitrary elements and the up heaping.
std::unordered_map<uint64_t, uint64_t> prepared_section_completed_;
std::mutex prep_heap_mutex_;
// No copying allowed
DBImpl(const DBImpl&);
void operator=(const DBImpl&);
@@ -1110,23 +857,16 @@ class DBImpl : public DB {
bool* value_found = nullptr);
bool GetIntPropertyInternal(ColumnFamilyData* cfd,
const DBPropertyInfo& property_info,
bool is_locked, uint64_t* value);
bool HasPendingManualCompaction();
bool HasExclusiveManualCompaction();
void AddManualCompaction(ManualCompaction* m);
void RemoveManualCompaction(ManualCompaction* m);
bool ShouldntRunManualCompaction(ManualCompaction* m);
bool HaveManualCompaction(ColumnFamilyData* cfd);
bool MCOverlap(ManualCompaction* m, ManualCompaction* m1);
size_t GetWalPreallocateBlockSize(uint64_t write_buffer_size) const;
DBPropertyType property_type,
bool need_out_of_mutex, bool is_locked,
uint64_t* value);
};
// Sanitize db options. The caller should delete result.info_log if
// it is not equal to src.info_log.
extern Options SanitizeOptions(const std::string& db,
const InternalKeyComparator* icmp,
const Options& src);
extern DBOptions SanitizeOptions(const std::string& db, const DBOptions& src);
// Fix user-supplied options to be reasonable
+4 -38
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -19,11 +19,6 @@ uint64_t DBImpl::TEST_GetLevel0TotalSize() {
return default_cf_handle_->cfd()->current()->storage_info()->NumLevelBytes(0);
}
void DBImpl::TEST_MaybeFlushColumnFamilies() {
InstrumentedMutexLock l(&mutex_);
MaybeFlushColumnFamilies();
}
int64_t DBImpl::TEST_MaxNextLevelOverlappingBytes(
ColumnFamilyHandle* column_family) {
ColumnFamilyData* cfd;
@@ -75,21 +70,14 @@ Status DBImpl::TEST_CompactRange(int level, const Slice* begin,
cfd->ioptions()->compaction_style == kCompactionStyleFIFO)
? level
: level + 1;
return RunManualCompaction(cfd, level, output_level, 0, begin, end, true,
return RunManualCompaction(cfd, level, output_level, 0, begin, end,
disallow_trivial_move);
}
Status DBImpl::TEST_FlushMemTable(bool wait, ColumnFamilyHandle* cfh) {
Status DBImpl::TEST_FlushMemTable(bool wait) {
FlushOptions fo;
fo.wait = wait;
ColumnFamilyData* cfd;
if (cfh == nullptr) {
cfd = default_cf_handle_->cfd();
} else {
auto cfhi = reinterpret_cast<ColumnFamilyHandleImpl*>(cfh);
cfd = cfhi->cfd();
}
return FlushMemTable(cfd, fo);
return FlushMemTable(default_cf_handle_->cfd(), fo);
}
Status DBImpl::TEST_WaitForFlushMemTable(ColumnFamilyHandle* column_family) {
@@ -166,27 +154,5 @@ Status DBImpl::TEST_GetAllImmutableCFOptions(
return Status::OK();
}
uint64_t DBImpl::TEST_FindMinLogContainingOutstandingPrep() {
return FindMinLogContainingOutstandingPrep();
}
uint64_t DBImpl::TEST_FindMinPrepLogReferencedByMemTable() {
return FindMinPrepLogReferencedByMemTable();
}
Status DBImpl::TEST_GetLatestMutableCFOptions(
ColumnFamilyHandle* column_family, MutableCFOptions* mutable_cf_options) {
InstrumentedMutexLock l(&mutex_);
auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
*mutable_cf_options = *cfh->cfd()->GetLatestMutableCFOptions();
return Status::OK();
}
int DBImpl::TEST_BGCompactionsAllowed() const {
InstrumentedMutexLock l(&mutex_);
return BGCompactionsAllowed();
}
} // namespace rocksdb
#endif // NDEBUG
+9 -9
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -49,8 +49,8 @@ Status DBImpl::SuggestCompactRange(ColumnFamilyHandle* column_family,
}
// Since we have some more files to compact, we should also recompute
// compaction score
vstorage->ComputeCompactionScore(*cfd->ioptions(),
*cfd->GetLatestMutableCFOptions());
vstorage->ComputeCompactionScore(*cfd->GetLatestMutableCFOptions(),
CompactionOptionsFIFO());
SchedulePendingCompaction(cfd);
MaybeScheduleFlushOrCompaction();
}
@@ -61,7 +61,7 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
assert(column_family);
if (target_level < 1) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"PromoteL0 FAILED. Invalid target level %d\n", target_level);
return Status::InvalidArgument("Invalid target level");
}
@@ -75,7 +75,7 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
const auto* vstorage = cfd->current()->storage_info();
if (target_level >= vstorage->num_levels()) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"PromoteL0 FAILED. Target level %d does not exist\n", target_level);
job_context.Clean();
return Status::InvalidArgument("Target level does not exist");
@@ -94,7 +94,7 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
for (size_t i = 0; i < l0_files.size(); ++i) {
auto f = l0_files[i];
if (f->being_compacted) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"PromoteL0 FAILED. File %" PRIu64 " being compacted\n",
f->fd.GetNumber());
job_context.Clean();
@@ -104,7 +104,7 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
if (i == 0) continue;
auto prev_f = l0_files[i - 1];
if (icmp->Compare(prev_f->largest, f->smallest) >= 0) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"PromoteL0 FAILED. Files %" PRIu64 " and %" PRIu64
" have overlapping ranges\n",
prev_f->fd.GetNumber(), f->fd.GetNumber());
@@ -116,7 +116,7 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
// Check that all levels up to target_level are empty.
for (int level = 1; level <= target_level; ++level) {
if (vstorage->NumLevelFiles(level) > 0) {
Log(InfoLogLevel::INFO_LEVEL, immutable_db_options_.info_log,
Log(InfoLogLevel::INFO_LEVEL, db_options_.info_log,
"PromoteL0 FAILED. Level %d not empty\n", level);
job_context.Clean();
return Status::InvalidArgument(
@@ -141,7 +141,7 @@ Status DBImpl::PromoteL0(ColumnFamilyHandle* column_family, int target_level) {
cfd, &job_context, *cfd->GetLatestMutableCFOptions());
}
} // lock released here
LogFlush(immutable_db_options_.info_log);
LogFlush(db_options_.info_log);
job_context.Clean();
return status;
+17 -26
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -8,9 +8,8 @@
#include "db/compacted_db_impl.h"
#include "db/db_impl.h"
#include "db/db_iter.h"
#include "db/merge_context.h"
#include "db/range_del_aggregator.h"
#include "db/db_iter.h"
#include "util/perf_context_imp.h"
namespace rocksdb {
@@ -20,9 +19,8 @@ namespace rocksdb {
DBImplReadOnly::DBImplReadOnly(const DBOptions& db_options,
const std::string& dbname)
: DBImpl(db_options, dbname) {
Log(INFO_LEVEL, immutable_db_options_.info_log,
"Opening the db in read only mode");
LogFlush(immutable_db_options_.info_log);
Log(INFO_LEVEL, db_options_.info_log, "Opening the db in read only mode");
LogFlush(db_options_.info_log);
}
DBImplReadOnly::~DBImplReadOnly() {
@@ -38,14 +36,11 @@ Status DBImplReadOnly::Get(const ReadOptions& read_options,
auto cfd = cfh->cfd();
SuperVersion* super_version = cfd->GetSuperVersion();
MergeContext merge_context;
RangeDelAggregator range_del_agg(cfd->internal_comparator(), snapshot);
LookupKey lkey(key, snapshot);
if (super_version->mem->Get(lkey, value, &s, &merge_context, &range_del_agg,
read_options)) {
if (super_version->mem->Get(lkey, value, &s, &merge_context)) {
} else {
PERF_TIMER_GUARD(get_from_output_files_time);
super_version->current->Get(read_options, lkey, value, &s, &merge_context,
&range_del_agg);
super_version->current->Get(read_options, lkey, value, &s, &merge_context);
}
return s;
}
@@ -59,14 +54,12 @@ Iterator* DBImplReadOnly::NewIterator(const ReadOptions& read_options,
auto db_iter = NewArenaWrappedDbIterator(
env_, *cfd->ioptions(), cfd->user_comparator(),
(read_options.snapshot != nullptr
? reinterpret_cast<const SnapshotImpl*>(read_options.snapshot)
->number_
? reinterpret_cast<const SnapshotImpl*>(
read_options.snapshot)->number_
: latest_snapshot),
super_version->mutable_cf_options.max_sequential_skip_in_iterations,
super_version->version_number);
auto internal_iter =
NewInternalIterator(read_options, cfd, super_version, db_iter->GetArena(),
db_iter->GetRangeDelAggregator());
super_version->mutable_cf_options.max_sequential_skip_in_iterations);
auto internal_iter = NewInternalIterator(
read_options, cfd, super_version, db_iter->GetArena());
db_iter->SetIterUnderDBIter(internal_iter);
return db_iter;
}
@@ -88,14 +81,12 @@ Status DBImplReadOnly::NewIterators(
auto* db_iter = NewArenaWrappedDbIterator(
env_, *cfd->ioptions(), cfd->user_comparator(),
(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);
auto* internal_iter =
NewInternalIterator(read_options, cfd, sv, db_iter->GetArena(),
db_iter->GetRangeDelAggregator());
? reinterpret_cast<const SnapshotImpl*>(
read_options.snapshot)->number_
: latest_snapshot),
sv->mutable_cf_options.max_sequential_skip_in_iterations);
auto* internal_iter = NewInternalIterator(
read_options, cfd, sv, db_iter->GetArena());
db_iter->SetIterUnderDBIter(internal_iter);
iterators->push_back(db_iter);
}
+1 -1
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
+14 -20
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -18,13 +18,12 @@ class DBTestInPlaceUpdate : public DBTestBase {
TEST_F(DBTestInPlaceUpdate, InPlaceUpdate) {
do {
Options options = CurrentOptions();
Options options;
options.create_if_missing = true;
options.inplace_update_support = true;
options.env = env_;
options.write_buffer_size = 100000;
options.allow_concurrent_memtable_write = false;
Reopen(options);
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Update key with values of smaller size
@@ -42,13 +41,12 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdate) {
TEST_F(DBTestInPlaceUpdate, InPlaceUpdateLargeNewValue) {
do {
Options options = CurrentOptions();
Options options;
options.create_if_missing = true;
options.inplace_update_support = true;
options.env = env_;
options.write_buffer_size = 100000;
options.allow_concurrent_memtable_write = false;
Reopen(options);
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Update key with values of larger size
@@ -66,7 +64,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateLargeNewValue) {
TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerSize) {
do {
Options options = CurrentOptions();
Options options;
options.create_if_missing = true;
options.inplace_update_support = true;
@@ -74,8 +72,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerSize) {
options.write_buffer_size = 100000;
options.inplace_callback =
rocksdb::DBTestInPlaceUpdate::updateInPlaceSmallerSize;
options.allow_concurrent_memtable_write = false;
Reopen(options);
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Update key with values of smaller size
@@ -95,7 +92,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerSize) {
TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerVarintSize) {
do {
Options options = CurrentOptions();
Options options;
options.create_if_missing = true;
options.inplace_update_support = true;
@@ -103,8 +100,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerVarintSize) {
options.write_buffer_size = 100000;
options.inplace_callback =
rocksdb::DBTestInPlaceUpdate::updateInPlaceSmallerVarintSize;
options.allow_concurrent_memtable_write = false;
Reopen(options);
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Update key with values of smaller varint size
@@ -124,7 +120,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackSmallerVarintSize) {
TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackLargeNewValue) {
do {
Options options = CurrentOptions();
Options options;
options.create_if_missing = true;
options.inplace_update_support = true;
@@ -132,8 +128,7 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackLargeNewValue) {
options.write_buffer_size = 100000;
options.inplace_callback =
rocksdb::DBTestInPlaceUpdate::updateInPlaceLargerSize;
options.allow_concurrent_memtable_write = false;
Reopen(options);
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Update key with values of larger size
@@ -151,16 +146,15 @@ TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackLargeNewValue) {
TEST_F(DBTestInPlaceUpdate, InPlaceUpdateCallbackNoAction) {
do {
Options options = CurrentOptions();
Options options;
options.create_if_missing = true;
options.inplace_update_support = true;
options.env = env_;
options.write_buffer_size = 100000;
options.inplace_callback =
rocksdb::DBTestInPlaceUpdate::updateInPlaceNoAction;
options.allow_concurrent_memtable_write = false;
Reopen(options);
rocksdb::DBTestInPlaceUpdate::updateInPlaceNoAction;
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Callback function requests no actions from db
-260
View File
@@ -1,260 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same 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/db_test_util.h"
#include "port/stack_trace.h"
namespace rocksdb {
class DBIOFailureTest : public DBTestBase {
public:
DBIOFailureTest() : DBTestBase("/db_io_failure_test") {}
};
#ifndef ROCKSDB_LITE
// Check that number of files does not grow when writes are dropped
TEST_F(DBIOFailureTest, DropWrites) {
do {
Options options = CurrentOptions();
options.env = env_;
options.paranoid_checks = false;
Reopen(options);
ASSERT_OK(Put("foo", "v1"));
ASSERT_EQ("v1", Get("foo"));
Compact("a", "z");
const size_t num_files = CountFiles();
// Force out-of-space errors
env_->drop_writes_.store(true, std::memory_order_release);
env_->sleep_counter_.Reset();
env_->no_slowdown_ = true;
for (int i = 0; i < 5; i++) {
if (option_config_ != kUniversalCompactionMultiLevel &&
option_config_ != kUniversalSubcompactions) {
for (int level = 0; level < dbfull()->NumberLevels(); level++) {
if (level > 0 && level == dbfull()->NumberLevels() - 1) {
break;
}
dbfull()->TEST_CompactRange(level, nullptr, nullptr, nullptr,
true /* disallow trivial move */);
}
} else {
dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr);
}
}
std::string property_value;
ASSERT_TRUE(db_->GetProperty("rocksdb.background-errors", &property_value));
ASSERT_EQ("5", property_value);
env_->drop_writes_.store(false, std::memory_order_release);
ASSERT_LT(CountFiles(), num_files + 3);
// Check that compaction attempts slept after errors
// TODO @krad: Figure out why ASSERT_EQ 5 keeps failing in certain compiler
// versions
ASSERT_GE(env_->sleep_counter_.Read(), 4);
} while (ChangeCompactOptions());
}
// Check background error counter bumped on flush failures.
TEST_F(DBIOFailureTest, DropWritesFlush) {
do {
Options options = CurrentOptions();
options.env = env_;
options.max_background_flushes = 1;
Reopen(options);
ASSERT_OK(Put("foo", "v1"));
// Force out-of-space errors
env_->drop_writes_.store(true, std::memory_order_release);
std::string property_value;
// Background error count is 0 now.
ASSERT_TRUE(db_->GetProperty("rocksdb.background-errors", &property_value));
ASSERT_EQ("0", property_value);
dbfull()->TEST_FlushMemTable(true);
ASSERT_TRUE(db_->GetProperty("rocksdb.background-errors", &property_value));
ASSERT_EQ("1", property_value);
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
TEST_F(DBIOFailureTest, NoSpaceCompactRange) {
do {
Options options = CurrentOptions();
options.env = env_;
options.disable_auto_compactions = true;
Reopen(options);
// generate 5 tables
for (int i = 0; i < 5; ++i) {
ASSERT_OK(Put(Key(i), Key(i) + "v"));
ASSERT_OK(Flush());
}
// Force out-of-space errors
env_->no_space_.store(true, std::memory_order_release);
Status s = dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
true /* disallow trivial move */);
ASSERT_TRUE(s.IsIOError());
ASSERT_TRUE(s.IsNoSpace());
env_->no_space_.store(false, std::memory_order_release);
} while (ChangeCompactOptions());
}
TEST_F(DBIOFailureTest, NonWritableFileSystem) {
do {
Options options = CurrentOptions();
options.write_buffer_size = 4096;
options.arena_block_size = 4096;
options.env = env_;
Reopen(options);
ASSERT_OK(Put("foo", "v1"));
env_->non_writeable_rate_.store(100);
std::string big(100000, 'x');
int errors = 0;
for (int i = 0; i < 20; i++) {
if (!Put("foo", big).ok()) {
errors++;
env_->SleepForMicroseconds(100000);
}
}
ASSERT_GT(errors, 0);
env_->non_writeable_rate_.store(0);
} while (ChangeCompactOptions());
}
#ifndef ROCKSDB_LITE
TEST_F(DBIOFailureTest, ManifestWriteError) {
// Test for the following problem:
// (a) Compaction produces file F
// (b) Log record containing F is written to MANIFEST file, but Sync() fails
// (c) GC deletes F
// (d) After reopening DB, reads fail since deleted F is named in log record
// We iterate twice. In the second iteration, everything is the
// same except the log record never makes it to the MANIFEST file.
for (int iter = 0; iter < 2; iter++) {
std::atomic<bool>* error_type = (iter == 0) ? &env_->manifest_sync_error_
: &env_->manifest_write_error_;
// Insert foo=>bar mapping
Options options = CurrentOptions();
options.env = env_;
options.create_if_missing = true;
options.error_if_exists = false;
options.paranoid_checks = true;
DestroyAndReopen(options);
ASSERT_OK(Put("foo", "bar"));
ASSERT_EQ("bar", Get("foo"));
// Memtable compaction (will succeed)
Flush();
ASSERT_EQ("bar", Get("foo"));
const int last = 2;
MoveFilesToLevel(2);
ASSERT_EQ(NumTableFilesAtLevel(last), 1); // foo=>bar is now in last level
// Merging compaction (will fail)
error_type->store(true, std::memory_order_release);
dbfull()->TEST_CompactRange(last, nullptr, nullptr); // Should fail
ASSERT_EQ("bar", Get("foo"));
error_type->store(false, std::memory_order_release);
// Since paranoid_checks=true, writes should fail
ASSERT_NOK(Put("foo2", "bar2"));
// Recovery: should not lose data
ASSERT_EQ("bar", Get("foo"));
// Try again with paranoid_checks=false
Close();
options.paranoid_checks = false;
Reopen(options);
// Merging compaction (will fail)
error_type->store(true, std::memory_order_release);
dbfull()->TEST_CompactRange(last, nullptr, nullptr); // Should fail
ASSERT_EQ("bar", Get("foo"));
// Recovery: should not lose data
error_type->store(false, std::memory_order_release);
Reopen(options);
ASSERT_EQ("bar", Get("foo"));
// Since paranoid_checks=false, writes should succeed
ASSERT_OK(Put("foo2", "bar2"));
ASSERT_EQ("bar", Get("foo"));
ASSERT_EQ("bar2", Get("foo2"));
}
}
#endif // ROCKSDB_LITE
TEST_F(DBIOFailureTest, PutFailsParanoid) {
// Test the following:
// (a) A random put fails in paranoid mode (simulate by sync fail)
// (b) All other puts have to fail, even if writes would succeed
// (c) All of that should happen ONLY if paranoid_checks = true
Options options = CurrentOptions();
options.env = env_;
options.create_if_missing = true;
options.error_if_exists = false;
options.paranoid_checks = true;
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
Status s;
ASSERT_OK(Put(1, "foo", "bar"));
ASSERT_OK(Put(1, "foo1", "bar1"));
// simulate error
env_->log_write_error_.store(true, std::memory_order_release);
s = Put(1, "foo2", "bar2");
ASSERT_TRUE(!s.ok());
env_->log_write_error_.store(false, std::memory_order_release);
s = Put(1, "foo3", "bar3");
// the next put should fail, too
ASSERT_TRUE(!s.ok());
// but we're still able to read
ASSERT_EQ("bar", Get(1, "foo"));
// do the same thing with paranoid checks off
options.paranoid_checks = false;
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
ASSERT_OK(Put(1, "foo", "bar"));
ASSERT_OK(Put(1, "foo1", "bar1"));
// simulate error
env_->log_write_error_.store(true, std::memory_order_release);
s = Put(1, "foo2", "bar2");
ASSERT_TRUE(!s.ok());
env_->log_write_error_.store(false, std::memory_order_release);
s = Put(1, "foo3", "bar3");
// the next put should NOT fail
ASSERT_TRUE(s.ok());
}
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
+189 -501
View File
File diff suppressed because it is too large. Load diff
+11 -18
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -9,14 +9,11 @@
#pragma once
#include <stdint.h>
#include <string>
#include "db/dbformat.h"
#include "db/range_del_aggregator.h"
#include "rocksdb/db.h"
#include "rocksdb/iterator.h"
#include "db/dbformat.h"
#include "util/arena.h"
#include "util/autovector.h"
#include "util/cf_options.h"
namespace rocksdb {
@@ -27,13 +24,13 @@ 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 ImmutableCFOptions& options,
const Comparator* user_key_comparator, InternalIterator* internal_iter,
const SequenceNumber& sequence, uint64_t max_sequential_skip_in_iterations,
uint64_t version_number, const Slice* iterate_upper_bound = nullptr,
bool prefix_same_as_start = false, bool pin_data = false,
bool total_order_seek = false);
extern Iterator* NewDBIterator(Env* env, const ImmutableCFOptions& options,
const Comparator* user_key_comparator,
InternalIterator* internal_iter,
const SequenceNumber& sequence,
uint64_t max_sequential_skip_in_iterations,
const Slice* iterate_upper_bound = nullptr,
bool prefix_same_as_start = 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
@@ -47,7 +44,6 @@ 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 RangeDelAggregator* GetRangeDelAggregator();
// Set the DB Iterator to be wrapped
@@ -60,7 +56,6 @@ class ArenaWrappedDBIter : public Iterator {
virtual void SeekToFirst() override;
virtual void SeekToLast() override;
virtual void Seek(const Slice& target) override;
virtual void SeekForPrev(const Slice& target) override;
virtual void Next() override;
virtual void Prev() override;
virtual Slice key() const override;
@@ -68,7 +63,6 @@ class ArenaWrappedDBIter : public Iterator {
virtual Status status() const override;
void RegisterCleanup(CleanupFunction function, void* arg1, void* arg2);
virtual Status GetProperty(std::string prop_name, std::string* prop) override;
private:
DBIter* db_iter_;
@@ -79,9 +73,8 @@ class ArenaWrappedDBIter : public Iterator {
extern ArenaWrappedDBIter* NewArenaWrappedDbIterator(
Env* env, const ImmutableCFOptions& options,
const Comparator* user_key_comparator, const SequenceNumber& sequence,
uint64_t max_sequential_skip_in_iterations, uint64_t version_number,
uint64_t max_sequential_skip_in_iterations,
const Slice* iterate_upper_bound = nullptr,
bool prefix_same_as_start = false, bool pin_data = false,
bool total_order_seek = false);
bool prefix_same_as_start = false);
} // namespace rocksdb
+100 -152
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2014, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -117,11 +117,6 @@ class TestIterator : public InternalIterator {
}
}
virtual void SeekForPrev(const Slice& target) override {
assert(initialized_);
SeekForPrevImpl(target, &cmp);
}
virtual void Next() override {
assert(initialized_);
if (data_.empty() || (iter_ == data_.size() - 1)) {
@@ -155,9 +150,6 @@ class TestIterator : public InternalIterator {
return Status::OK();
}
virtual bool IsKeyPinned() const override { return true; }
virtual bool IsValuePinned() const override { return true; }
private:
bool initialized_;
bool valid_;
@@ -189,9 +181,10 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
internal_iter->AddPut("b", "val_b");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 10,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -222,7 +215,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
10, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -248,8 +241,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
10, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -281,8 +273,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
10, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -317,8 +308,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
10, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
db_iter->SeekToLast();
ASSERT_TRUE(!db_iter->Valid());
@@ -347,8 +337,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
7, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
7, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
SetPerfLevel(kEnableCount);
ASSERT_TRUE(GetPerfLevel() == kEnableCount);
@@ -357,7 +346,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(static_cast<int>(perf_context.internal_key_skipped_count), 7);
ASSERT_EQ(static_cast<int>(perf_context.internal_key_skipped_count), 1);
ASSERT_EQ(db_iter->key().ToString(), "b");
SetPerfLevel(kDisable);
@@ -385,8 +374,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
4, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
4, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -411,8 +399,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
10, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
db_iter->SeekToLast();
ASSERT_TRUE(!db_iter->Valid());
@@ -434,8 +421,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
10, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -470,8 +456,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
7, options.max_sequential_skip_in_iterations, 0,
ro.iterate_upper_bound));
7, options.max_sequential_skip_in_iterations, ro.iterate_upper_bound));
SetPerfLevel(kEnableCount);
ASSERT_TRUE(GetPerfLevel() == kEnableCount);
@@ -480,7 +465,7 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(static_cast<int>(perf_context.internal_delete_skipped_count), 1);
ASSERT_EQ(static_cast<int>(perf_context.internal_delete_skipped_count), 0);
ASSERT_EQ(db_iter->key().ToString(), "b");
SetPerfLevel(kDisable);
@@ -497,9 +482,10 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
internal_iter->AddPut("b", "val_b");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 10,
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
@@ -538,9 +524,10 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
internal_iter->AddPut("b", "val_b");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
2, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 2,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "b");
@@ -568,9 +555,10 @@ TEST_F(DBIteratorTest, DBIteratorPrevNext) {
internal_iter->AddPut("c", "val_c");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 10,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -595,9 +583,10 @@ TEST_F(DBIteratorTest, DBIteratorEmpty) {
TestIterator* internal_iter = new TestIterator(BytewiseComparator());
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
0, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 0,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(!db_iter->Valid());
}
@@ -606,9 +595,10 @@ TEST_F(DBIteratorTest, DBIteratorEmpty) {
TestIterator* internal_iter = new TestIterator(BytewiseComparator());
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
0, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 0,
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(!db_iter->Valid());
}
@@ -627,9 +617,10 @@ TEST_F(DBIteratorTest, DBIteratorUseSkipCountSkips) {
}
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 2,
options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 2,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -668,8 +659,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
options.statistics = rocksdb::CreateDBStatistics();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, i + 2, options.max_sequential_skip_in_iterations, 0));
env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, i + 2,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -703,8 +695,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, i + 2, options.max_sequential_skip_in_iterations, 0));
env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, i + 2,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -731,8 +724,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, 202, options.max_sequential_skip_in_iterations, 0));
env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 202,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -762,9 +756,10 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
}
internal_iter->AddPut("c", "200");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, i, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, i,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(!db_iter->Valid());
@@ -778,9 +773,10 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
}
internal_iter->AddPut("c", "200");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
200, options.max_sequential_skip_in_iterations, 0));
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, 200,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -813,8 +809,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, i + 2, options.max_sequential_skip_in_iterations, 0));
env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, i + 2,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -847,8 +844,9 @@ TEST_F(DBIteratorTest, DBIteratorUseSkip) {
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, i + 2, options.max_sequential_skip_in_iterations, 0));
env_, ImmutableCFOptions(options),
BytewiseComparator(), internal_iter, i + 2,
options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -889,7 +887,7 @@ TEST_F(DBIteratorTest, DBIterator1) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 1,
options.max_sequential_skip_in_iterations, 0));
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -915,7 +913,7 @@ TEST_F(DBIteratorTest, DBIterator2) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 0,
options.max_sequential_skip_in_iterations, 0));
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -938,7 +936,7 @@ TEST_F(DBIteratorTest, DBIterator3) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 2,
options.max_sequential_skip_in_iterations, 0));
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -960,7 +958,7 @@ TEST_F(DBIteratorTest, DBIterator4) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 4,
options.max_sequential_skip_in_iterations, 0));
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -989,7 +987,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
0, options.max_sequential_skip_in_iterations, 0));
0, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1011,7 +1009,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
1, options.max_sequential_skip_in_iterations, 0));
1, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1033,7 +1031,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
2, options.max_sequential_skip_in_iterations, 0));
2, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1055,7 +1053,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
3, options.max_sequential_skip_in_iterations, 0));
3, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1077,7 +1075,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
4, options.max_sequential_skip_in_iterations, 0));
4, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1099,7 +1097,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
5, options.max_sequential_skip_in_iterations, 0));
5, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1121,7 +1119,7 @@ TEST_F(DBIteratorTest, DBIterator5) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
6, options.max_sequential_skip_in_iterations, 0));
6, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1129,26 +1127,6 @@ TEST_F(DBIteratorTest, DBIterator5) {
db_iter->Prev();
ASSERT_TRUE(!db_iter->Valid());
}
{
// put, singledelete, merge
TestIterator* internal_iter = new TestIterator(BytewiseComparator());
internal_iter->AddPut("a", "val_a");
internal_iter->AddSingleDeletion("a");
internal_iter->AddMerge("a", "merge_1");
internal_iter->AddMerge("a", "merge_2");
internal_iter->AddPut("b", "val_b");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
db_iter->Seek("b");
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "b");
db_iter->Prev();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
}
}
TEST_F(DBIteratorTest, DBIterator6) {
@@ -1167,7 +1145,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
0, options.max_sequential_skip_in_iterations, 0));
0, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1189,7 +1167,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
1, options.max_sequential_skip_in_iterations, 0));
1, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1211,7 +1189,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
2, options.max_sequential_skip_in_iterations, 0));
2, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1233,7 +1211,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
3, options.max_sequential_skip_in_iterations, 0));
3, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(!db_iter->Valid());
}
@@ -1251,7 +1229,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
4, options.max_sequential_skip_in_iterations, 0));
4, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1273,7 +1251,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
5, options.max_sequential_skip_in_iterations, 0));
5, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1295,7 +1273,7 @@ TEST_F(DBIteratorTest, DBIterator6) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
6, options.max_sequential_skip_in_iterations, 0));
6, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1333,7 +1311,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
0, options.max_sequential_skip_in_iterations, 0));
0, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1367,7 +1345,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
2, options.max_sequential_skip_in_iterations, 0));
2, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1407,7 +1385,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
4, options.max_sequential_skip_in_iterations, 0));
4, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1447,7 +1425,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
5, options.max_sequential_skip_in_iterations, 0));
5, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1492,7 +1470,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
6, options.max_sequential_skip_in_iterations, 0));
6, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1538,7 +1516,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
7, options.max_sequential_skip_in_iterations, 0));
7, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1578,7 +1556,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
9, options.max_sequential_skip_in_iterations, 0));
9, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1624,7 +1602,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
13, options.max_sequential_skip_in_iterations, 0));
13, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1671,7 +1649,7 @@ TEST_F(DBIteratorTest, DBIterator7) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
14, options.max_sequential_skip_in_iterations, 0));
14, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1700,7 +1678,7 @@ TEST_F(DBIteratorTest, DBIterator8) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
10, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "b");
@@ -1729,7 +1707,7 @@ TEST_F(DBIteratorTest, DBIterator9) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
10, options.max_sequential_skip_in_iterations));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
@@ -1751,15 +1729,6 @@ TEST_F(DBIteratorTest, DBIterator9) {
ASSERT_EQ(db_iter->key().ToString(), "a");
ASSERT_EQ(db_iter->value().ToString(), "merge_1,merge_2");
db_iter->SeekForPrev("b");
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "b");
ASSERT_EQ(db_iter->value().ToString(), "merge_3,merge_4");
db_iter->Next();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "d");
ASSERT_EQ(db_iter->value().ToString(), "merge_5,merge_6");
db_iter->Seek("c");
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "d");
@@ -1768,15 +1737,6 @@ TEST_F(DBIteratorTest, DBIterator9) {
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "b");
ASSERT_EQ(db_iter->value().ToString(), "merge_3,merge_4");
db_iter->SeekForPrev("c");
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "b");
ASSERT_EQ(db_iter->value().ToString(), "merge_3,merge_4");
db_iter->Next();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "d");
ASSERT_EQ(db_iter->value().ToString(), "merge_5,merge_6");
}
}
@@ -1794,7 +1754,7 @@ TEST_F(DBIteratorTest, DBIterator10) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, options.max_sequential_skip_in_iterations, 0));
10, options.max_sequential_skip_in_iterations));
db_iter->Seek("c");
ASSERT_TRUE(db_iter->Valid());
@@ -1807,18 +1767,6 @@ TEST_F(DBIteratorTest, DBIterator10) {
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "c");
ASSERT_EQ(db_iter->value().ToString(), "3");
db_iter->SeekForPrev("c");
ASSERT_TRUE(db_iter->Valid());
db_iter->Next();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "d");
ASSERT_EQ(db_iter->value().ToString(), "4");
db_iter->Prev();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "c");
ASSERT_EQ(db_iter->value().ToString(), "3");
}
TEST_F(DBIteratorTest, SeekToLastOccurrenceSeq0) {
@@ -1830,9 +1778,9 @@ TEST_F(DBIteratorTest, SeekToLastOccurrenceSeq0) {
internal_iter->AddPut("b", "2");
internal_iter->Finish();
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, 10, 0 /* force seek */, 0));
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter,
10, 0 /* force seek */));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1859,7 +1807,7 @@ TEST_F(DBIteratorTest, DBIterator11) {
std::unique_ptr<Iterator> db_iter(NewDBIterator(
env_, ImmutableCFOptions(options), BytewiseComparator(), internal_iter, 1,
options.max_sequential_skip_in_iterations, 0));
options.max_sequential_skip_in_iterations));
db_iter->SeekToFirst();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "a");
@@ -1884,7 +1832,7 @@ TEST_F(DBIteratorTest, DBIterator12) {
std::unique_ptr<Iterator> db_iter(
NewDBIterator(env_, ImmutableCFOptions(options), BytewiseComparator(),
internal_iter, 10, 0, 0));
internal_iter, 10, 0));
db_iter->SeekToLast();
ASSERT_TRUE(db_iter->Valid());
ASSERT_EQ(db_iter->key().ToString(), "c");
@@ -1926,7 +1874,7 @@ class DBIterWithMergeIterTest : public testing::Test {
db_iter_.reset(NewDBIterator(env_, ImmutableCFOptions(options_),
BytewiseComparator(), merge_iter,
8 /* read data earlier than seqId 8 */,
3 /* max iterators before reseek */, 0));
3 /* max iterators before reseek */));
}
Env* env_;
@@ -1969,7 +1917,7 @@ TEST_F(DBIterWithMergeIterTest, InnerMergeIterator1) {
TEST_F(DBIterWithMergeIterTest, InnerMergeIterator2) {
// Test Prev() when one child iterator is at its end.
db_iter_->SeekForPrev("g");
db_iter_->Seek("g");
ASSERT_TRUE(db_iter_->Valid());
ASSERT_EQ(db_iter_->key().ToString(), "g");
ASSERT_EQ(db_iter_->value().ToString(), "3");
File diff suppressed because it is too large. Load diff
+2 -2
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -27,7 +27,7 @@ class DBTestXactLogIterator : public DBTestBase {
Status status = dbfull()->GetUpdatesSince(seq, &iter);
EXPECT_OK(status);
EXPECT_TRUE(iter->Valid());
return iter;
return std::move(iter);
}
};
-163
View File
@@ -1,163 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
#include <memory>
#include <string>
#include "db/db_test_util.h"
#include "db/memtable.h"
#include "port/stack_trace.h"
#include "rocksdb/memtablerep.h"
#include "rocksdb/slice_transform.h"
namespace rocksdb {
class DBMemTableTest : public DBTestBase {
public:
DBMemTableTest() : DBTestBase("/db_memtable_test") {}
};
class MockMemTableRep : public MemTableRep {
public:
explicit MockMemTableRep(MemTableAllocator* allocator, MemTableRep* rep)
: MemTableRep(allocator), rep_(rep), num_insert_with_hint_(0) {}
virtual KeyHandle Allocate(const size_t len, char** buf) override {
return rep_->Allocate(len, buf);
}
virtual void Insert(KeyHandle handle) override {
return rep_->Insert(handle);
}
virtual void InsertWithHint(KeyHandle handle, void** hint) override {
num_insert_with_hint_++;
ASSERT_NE(nullptr, hint);
last_hint_in_ = *hint;
rep_->InsertWithHint(handle, hint);
last_hint_out_ = *hint;
}
virtual bool Contains(const char* key) const override {
return rep_->Contains(key);
}
virtual void Get(const LookupKey& k, void* callback_args,
bool (*callback_func)(void* arg,
const char* entry)) override {
rep_->Get(k, callback_args, callback_func);
}
virtual size_t ApproximateMemoryUsage() override {
return rep_->ApproximateMemoryUsage();
}
virtual Iterator* GetIterator(Arena* arena) override {
return rep_->GetIterator(arena);
}
void* last_hint_in() { return last_hint_in_; }
void* last_hint_out() { return last_hint_out_; }
int num_insert_with_hint() { return num_insert_with_hint_; }
private:
std::unique_ptr<MemTableRep> rep_;
void* last_hint_in_;
void* last_hint_out_;
int num_insert_with_hint_;
};
class MockMemTableRepFactory : public MemTableRepFactory {
public:
virtual MemTableRep* CreateMemTableRep(const MemTableRep::KeyComparator& cmp,
MemTableAllocator* allocator,
const SliceTransform* transform,
Logger* logger) override {
SkipListFactory factory;
MemTableRep* skiplist_rep =
factory.CreateMemTableRep(cmp, allocator, transform, logger);
mock_rep_ = new MockMemTableRep(allocator, skiplist_rep);
return mock_rep_;
}
virtual const char* Name() const override { return "MockMemTableRepFactory"; }
MockMemTableRep* rep() { return mock_rep_; }
bool IsInsertConcurrentlySupported() const override { return false; }
private:
MockMemTableRep* mock_rep_;
};
class TestPrefixExtractor : public SliceTransform {
public:
virtual const char* Name() const override { return "TestPrefixExtractor"; }
virtual Slice Transform(const Slice& key) const override {
const char* p = separator(key);
if (p == nullptr) {
return Slice();
}
return Slice(key.data(), p - key.data() + 1);
}
virtual bool InDomain(const Slice& key) const override {
return separator(key) != nullptr;
}
virtual bool InRange(const Slice& key) const override { return false; }
private:
const char* separator(const Slice& key) const {
return reinterpret_cast<const char*>(memchr(key.data(), '_', key.size()));
}
};
TEST_F(DBMemTableTest, InsertWithHint) {
Options options;
options.allow_concurrent_memtable_write = false;
options.create_if_missing = true;
options.memtable_factory.reset(new MockMemTableRepFactory());
options.memtable_insert_with_hint_prefix_extractor.reset(
new TestPrefixExtractor());
Reopen(options);
MockMemTableRep* rep =
reinterpret_cast<MockMemTableRepFactory*>(options.memtable_factory.get())
->rep();
ASSERT_OK(Put("foo_k1", "foo_v1"));
ASSERT_EQ(nullptr, rep->last_hint_in());
void* hint_foo = rep->last_hint_out();
ASSERT_OK(Put("foo_k2", "foo_v2"));
ASSERT_EQ(hint_foo, rep->last_hint_in());
ASSERT_EQ(hint_foo, rep->last_hint_out());
ASSERT_OK(Put("foo_k3", "foo_v3"));
ASSERT_EQ(hint_foo, rep->last_hint_in());
ASSERT_EQ(hint_foo, rep->last_hint_out());
ASSERT_OK(Put("bar_k1", "bar_v1"));
ASSERT_EQ(nullptr, rep->last_hint_in());
void* hint_bar = rep->last_hint_out();
ASSERT_NE(hint_foo, hint_bar);
ASSERT_OK(Put("bar_k2", "bar_v2"));
ASSERT_EQ(hint_bar, rep->last_hint_in());
ASSERT_EQ(hint_bar, rep->last_hint_out());
ASSERT_EQ(5, rep->num_insert_with_hint());
ASSERT_OK(Put("whitelisted", "vvv"));
ASSERT_EQ(5, rep->num_insert_with_hint());
ASSERT_EQ("foo_v1", Get("foo_k1"));
ASSERT_EQ("foo_v2", Get("foo_k2"));
ASSERT_EQ("foo_v3", Get("foo_k3"));
ASSERT_EQ("bar_v1", Get("bar_k1"));
ASSERT_EQ("bar_v2", Get("bar_k2"));
ASSERT_EQ("vvv", Get("whitelisted"));
}
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
-98
View File
@@ -1,98 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
#include <string>
#include <vector>
#include "db/db_test_util.h"
#include "db/forward_iterator.h"
#include "port/stack_trace.h"
#include "utilities/merge_operators.h"
namespace rocksdb {
// Test merge operator functionality.
class DBMergeOperatorTest : public DBTestBase {
public:
DBMergeOperatorTest() : DBTestBase("/db_merge_operator_test") {}
};
// A test merge operator mimics put but also fails if one of merge operands is
// "corrupted".
class TestPutOperator : public MergeOperator {
public:
virtual bool FullMergeV2(const MergeOperationInput& merge_in,
MergeOperationOutput* merge_out) const override {
if (merge_in.existing_value != nullptr &&
*(merge_in.existing_value) == "corrupted") {
return false;
}
for (auto value : merge_in.operand_list) {
if (value == "corrupted") {
return false;
}
}
merge_out->existing_operand = merge_in.operand_list.back();
return true;
}
virtual const char* Name() const override { return "TestPutOperator"; }
};
TEST_F(DBMergeOperatorTest, MergeErrorOnRead) {
Options options;
options.create_if_missing = true;
options.merge_operator.reset(new TestPutOperator());
Reopen(options);
ASSERT_OK(Merge("k1", "v1"));
ASSERT_OK(Merge("k1", "corrupted"));
std::string value;
ASSERT_TRUE(db_->Get(ReadOptions(), "k1", &value).IsCorruption());
VerifyDBInternal({{"k1", "corrupted"}, {"k1", "v1"}});
}
TEST_F(DBMergeOperatorTest, MergeErrorOnWrite) {
Options options;
options.create_if_missing = true;
options.merge_operator.reset(new TestPutOperator());
options.max_successive_merges = 3;
Reopen(options);
ASSERT_OK(Merge("k1", "v1"));
ASSERT_OK(Merge("k1", "v2"));
// Will trigger a merge when hitting max_successive_merges and the merge
// will fail. The delta will be inserted nevertheless.
ASSERT_OK(Merge("k1", "corrupted"));
// Data should stay unmerged after the error.
VerifyDBInternal({{"k1", "corrupted"}, {"k1", "v2"}, {"k1", "v1"}});
}
TEST_F(DBMergeOperatorTest, MergeErrorOnIteration) {
Options options;
options.create_if_missing = true;
options.merge_operator.reset(new TestPutOperator());
DestroyAndReopen(options);
ASSERT_OK(Merge("k1", "v1"));
ASSERT_OK(Merge("k1", "corrupted"));
ASSERT_OK(Put("k2", "v2"));
VerifyDBFromMap({{"k1", ""}, {"k2", "v2"}}, nullptr, false,
{{"k1", Status::Corruption()}});
VerifyDBInternal({{"k1", "corrupted"}, {"k1", "v1"}, {"k2", "v2"}});
DestroyAndReopen(options);
ASSERT_OK(Merge("k1", "v1"));
ASSERT_OK(Put("k2", "v2"));
ASSERT_OK(Merge("k2", "corrupted"));
VerifyDBFromMap({{"k1", "v1"}, {"k2", ""}}, nullptr, false,
{{"k2", Status::Corruption()}});
VerifyDBInternal({{"k1", "v1"}, {"k2", "corrupted"}, {"k2", "v2"}});
}
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
-327
View File
@@ -1,327 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same 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 <limits>
#include <string>
#include <unordered_map>
#include "db/column_family.h"
#include "db/db_impl.h"
#include "db/db_test_util.h"
#include "port/stack_trace.h"
#include "rocksdb/convenience.h"
#include "util/options_helper.h"
#include "util/random.h"
#include "util/sync_point.h"
#include "util/testutil.h"
namespace rocksdb {
class DBOptionsTest : public DBTestBase {
public:
DBOptionsTest() : DBTestBase("/db_options_test") {}
#ifndef ROCKSDB_LITE
std::unordered_map<std::string, std::string> GetMutableDBOptionsMap(
const DBOptions& options) {
std::string options_str;
GetStringFromDBOptions(&options_str, options);
std::unordered_map<std::string, std::string> options_map;
StringToMap(options_str, &options_map);
std::unordered_map<std::string, std::string> mutable_map;
for (const auto opt : db_options_type_info) {
if (opt.second.is_mutable &&
opt.second.verification != OptionVerificationType::kDeprecated) {
mutable_map[opt.first] = options_map[opt.first];
}
}
return mutable_map;
}
std::unordered_map<std::string, std::string> GetMutableCFOptionsMap(
const ColumnFamilyOptions& options) {
std::string options_str;
GetStringFromColumnFamilyOptions(&options_str, options);
std::unordered_map<std::string, std::string> options_map;
StringToMap(options_str, &options_map);
std::unordered_map<std::string, std::string> mutable_map;
for (const auto opt : cf_options_type_info) {
if (opt.second.is_mutable &&
opt.second.verification != OptionVerificationType::kDeprecated) {
mutable_map[opt.first] = options_map[opt.first];
}
}
return mutable_map;
}
std::unordered_map<std::string, std::string> GetRandomizedMutableCFOptionsMap(
Random* rnd) {
Options options;
ImmutableDBOptions db_options(options);
test::RandomInitCFOptions(&options, rnd);
auto sanitized_options = SanitizeOptions(db_options, options);
auto opt_map = GetMutableCFOptionsMap(sanitized_options);
delete options.compaction_filter;
return opt_map;
}
std::unordered_map<std::string, std::string> GetRandomizedMutableDBOptionsMap(
Random* rnd) {
DBOptions db_options;
test::RandomInitDBOptions(&db_options, rnd);
auto sanitized_options = SanitizeOptions(dbname_, db_options);
return GetMutableDBOptionsMap(sanitized_options);
}
#endif // ROCKSDB_LITE
};
// RocksDB lite don't support dynamic options.
#ifndef ROCKSDB_LITE
TEST_F(DBOptionsTest, GetLatestDBOptions) {
// GetOptions should be able to get latest option changed by SetOptions.
Options options;
options.create_if_missing = true;
Random rnd(228);
Reopen(options);
auto new_options = GetRandomizedMutableDBOptionsMap(&rnd);
ASSERT_OK(dbfull()->SetDBOptions(new_options));
ASSERT_EQ(new_options, GetMutableDBOptionsMap(dbfull()->GetDBOptions()));
}
TEST_F(DBOptionsTest, GetLatestCFOptions) {
// GetOptions should be able to get latest option changed by SetOptions.
Options options;
options.create_if_missing = true;
Random rnd(228);
Reopen(options);
CreateColumnFamilies({"foo"}, options);
ReopenWithColumnFamilies({"default", "foo"}, options);
auto options_default = GetRandomizedMutableCFOptionsMap(&rnd);
auto options_foo = GetRandomizedMutableCFOptionsMap(&rnd);
ASSERT_OK(dbfull()->SetOptions(handles_[0], options_default));
ASSERT_OK(dbfull()->SetOptions(handles_[1], options_foo));
ASSERT_EQ(options_default,
GetMutableCFOptionsMap(dbfull()->GetOptions(handles_[0])));
ASSERT_EQ(options_foo,
GetMutableCFOptionsMap(dbfull()->GetOptions(handles_[1])));
}
TEST_F(DBOptionsTest, SetOptionsAndReopen) {
Random rnd(1044);
auto rand_opts = GetRandomizedMutableCFOptionsMap(&rnd);
ASSERT_OK(dbfull()->SetOptions(rand_opts));
// Verify if DB can be reopen after setting options.
Options options;
ASSERT_OK(TryReopen(options));
}
TEST_F(DBOptionsTest, EnableAutoCompactionAndTriggerStall) {
const std::string kValue(1024, 'v');
for (int method_type = 0; method_type < 2; method_type++) {
for (int option_type = 0; option_type < 4; option_type++) {
Options options;
options.create_if_missing = true;
options.disable_auto_compactions = true;
options.write_buffer_size = 1024 * 1024 * 10;
options.compression = CompressionType::kNoCompression;
options.level0_file_num_compaction_trigger = 1;
options.level0_stop_writes_trigger = std::numeric_limits<int>::max();
options.level0_slowdown_writes_trigger = std::numeric_limits<int>::max();
options.hard_pending_compaction_bytes_limit =
std::numeric_limits<uint64_t>::max();
options.soft_pending_compaction_bytes_limit =
std::numeric_limits<uint64_t>::max();
DestroyAndReopen(options);
int i = 0;
for (; i < 1024; i++) {
Put(Key(i), kValue);
}
Flush();
for (; i < 1024 * 2; i++) {
Put(Key(i), kValue);
}
Flush();
dbfull()->TEST_WaitForFlushMemTable();
ASSERT_EQ(2, NumTableFilesAtLevel(0));
uint64_t l0_size = SizeAtLevel(0);
switch (option_type) {
case 0:
// test with level0_stop_writes_trigger
options.level0_stop_writes_trigger = 2;
options.level0_slowdown_writes_trigger = 2;
break;
case 1:
options.level0_slowdown_writes_trigger = 2;
break;
case 2:
options.hard_pending_compaction_bytes_limit = l0_size;
options.soft_pending_compaction_bytes_limit = l0_size;
break;
case 3:
options.soft_pending_compaction_bytes_limit = l0_size;
break;
}
Reopen(options);
dbfull()->TEST_WaitForCompact();
ASSERT_FALSE(dbfull()->TEST_write_controler().IsStopped());
ASSERT_FALSE(dbfull()->TEST_write_controler().NeedsDelay());
SyncPoint::GetInstance()->LoadDependency(
{{"DBOptionsTest::EnableAutoCompactionAndTriggerStall:1",
"BackgroundCallCompaction:0"},
{"DBImpl::BackgroundCompaction():BeforePickCompaction",
"DBOptionsTest::EnableAutoCompactionAndTriggerStall:2"},
{"DBOptionsTest::EnableAutoCompactionAndTriggerStall:3",
"DBImpl::BackgroundCompaction():AfterPickCompaction"}});
// Block background compaction.
SyncPoint::GetInstance()->EnableProcessing();
switch (method_type) {
case 0:
ASSERT_OK(
dbfull()->SetOptions({{"disable_auto_compactions", "false"}}));
break;
case 1:
ASSERT_OK(dbfull()->EnableAutoCompaction(
{dbfull()->DefaultColumnFamily()}));
break;
}
TEST_SYNC_POINT("DBOptionsTest::EnableAutoCompactionAndTriggerStall:1");
// Wait for stall condition recalculate.
TEST_SYNC_POINT("DBOptionsTest::EnableAutoCompactionAndTriggerStall:2");
switch (option_type) {
case 0:
ASSERT_TRUE(dbfull()->TEST_write_controler().IsStopped());
break;
case 1:
ASSERT_FALSE(dbfull()->TEST_write_controler().IsStopped());
ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
break;
case 2:
ASSERT_TRUE(dbfull()->TEST_write_controler().IsStopped());
break;
case 3:
ASSERT_FALSE(dbfull()->TEST_write_controler().IsStopped());
ASSERT_TRUE(dbfull()->TEST_write_controler().NeedsDelay());
break;
}
TEST_SYNC_POINT("DBOptionsTest::EnableAutoCompactionAndTriggerStall:3");
// Background compaction executed.
dbfull()->TEST_WaitForCompact();
ASSERT_FALSE(dbfull()->TEST_write_controler().IsStopped());
ASSERT_FALSE(dbfull()->TEST_write_controler().NeedsDelay());
}
}
}
TEST_F(DBOptionsTest, SetOptionsMayTriggerCompaction) {
Options options;
options.create_if_missing = true;
options.level0_file_num_compaction_trigger = 1000;
Reopen(options);
for (int i = 0; i < 3; i++) {
// Need to insert two keys to avoid trivial move.
ASSERT_OK(Put("foo", ToString(i)));
ASSERT_OK(Put("bar", ToString(i)));
Flush();
}
ASSERT_EQ("3", FilesPerLevel());
ASSERT_OK(
dbfull()->SetOptions({{"level0_file_num_compaction_trigger", "3"}}));
dbfull()->TEST_WaitForCompact();
ASSERT_EQ("0,1", FilesPerLevel());
}
TEST_F(DBOptionsTest, SetBackgroundCompactionThreads) {
Options options;
options.create_if_missing = true;
options.base_background_compactions = 1; // default value
options.max_background_compactions = 1; // default value
Reopen(options);
ASSERT_EQ(1, dbfull()->TEST_BGCompactionsAllowed());
ASSERT_OK(dbfull()->SetDBOptions({{"base_background_compactions", "2"},
{"max_background_compactions", "3"}}));
ASSERT_EQ(2, dbfull()->TEST_BGCompactionsAllowed());
auto stop_token = dbfull()->TEST_write_controler().GetStopToken();
ASSERT_EQ(3, dbfull()->TEST_BGCompactionsAllowed());
}
TEST_F(DBOptionsTest, AvoidFlushDuringShutdown) {
Options options;
options.create_if_missing = true;
options.disable_auto_compactions = true;
WriteOptions write_without_wal;
write_without_wal.disableWAL = true;
ASSERT_FALSE(options.avoid_flush_during_shutdown);
DestroyAndReopen(options);
ASSERT_OK(Put("foo", "v1", write_without_wal));
Reopen(options);
ASSERT_EQ("v1", Get("foo"));
ASSERT_EQ("1", FilesPerLevel());
DestroyAndReopen(options);
ASSERT_OK(Put("foo", "v2", write_without_wal));
ASSERT_OK(dbfull()->SetDBOptions({{"avoid_flush_during_shutdown", "true"}}));
Reopen(options);
ASSERT_EQ("NOT_FOUND", Get("foo"));
ASSERT_EQ("", FilesPerLevel());
}
TEST_F(DBOptionsTest, SetDelayedWriteRateOption) {
Options options;
options.create_if_missing = true;
options.delayed_write_rate = 2 * 1024U * 1024U;
Reopen(options);
ASSERT_EQ(2 * 1024U * 1024U, dbfull()->TEST_write_controler().max_delayed_write_rate());
ASSERT_OK(dbfull()->SetDBOptions({{"delayed_write_rate", "20000"}}));
ASSERT_EQ(20000, dbfull()->TEST_write_controler().max_delayed_write_rate());
}
TEST_F(DBOptionsTest, MaxTotalWalSizeChange) {
Random rnd(1044);
const auto value_size = size_t(1024);
std::string value;
test::RandomString(&rnd, value_size, &value);
Options options;
options.create_if_missing = true;
CreateColumnFamilies({"1", "2", "3"}, options);
ReopenWithColumnFamilies({"default", "1", "2", "3"}, options);
WriteOptions write_options;
const int key_count = 100;
for (int i = 0; i < key_count; ++i) {
for (size_t cf = 0; cf < handles_.size(); ++cf) {
ASSERT_OK(Put(static_cast<int>(cf), Key(i), value));
}
}
ASSERT_OK(dbfull()->SetDBOptions({{"max_total_wal_size", "10"}}));
for (size_t cf = 0; cf < handles_.size(); ++cf) {
dbfull()->TEST_WaitForFlushMemTable(handles_[cf]);
ASSERT_EQ("1", FilesPerLevel(static_cast<int>(cf)));
}
}
#endif // ROCKSDB_LITE
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
File diff suppressed because it is too large. Load diff
-492
View File
@@ -1,492 +0,0 @@
// Copyright (c) 2016-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
#include "db/db_test_util.h"
#include "port/stack_trace.h"
#include "util/testutil.h"
#include "utilities/merge_operators.h"
namespace rocksdb {
class DBRangeDelTest : public DBTestBase {
public:
DBRangeDelTest() : DBTestBase("/db_range_del_test") {}
std::string GetNumericStr(int key) {
uint64_t uint64_key = static_cast<uint64_t>(key);
std::string str;
str.resize(8);
memcpy(&str[0], static_cast<void*>(&uint64_key), 8);
return str;
}
};
// PlainTableFactory and NumTableFilesAtLevel() are not supported in
// ROCKSDB_LITE
#ifndef ROCKSDB_LITE
TEST_F(DBRangeDelTest, NonBlockBasedTableNotSupported) {
Options opts = CurrentOptions();
opts.table_factory.reset(new PlainTableFactory());
opts.prefix_extractor.reset(NewNoopTransform());
opts.allow_mmap_reads = true;
opts.max_sequential_skip_in_iterations = 999999;
Reopen(opts);
ASSERT_TRUE(
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr1", "dr1")
.IsNotSupported());
}
TEST_F(DBRangeDelTest, FlushOutputHasOnlyRangeTombstones) {
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr1",
"dr1"));
ASSERT_OK(db_->Flush(FlushOptions()));
ASSERT_EQ(1, NumTableFilesAtLevel(0));
}
TEST_F(DBRangeDelTest, CompactionOutputHasOnlyRangeTombstone) {
Options opts = CurrentOptions();
opts.disable_auto_compactions = true;
opts.statistics = CreateDBStatistics();
Reopen(opts);
// snapshot protects range tombstone from dropping due to becoming obsolete.
const Snapshot* snapshot = db_->GetSnapshot();
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "z");
db_->Flush(FlushOptions());
ASSERT_EQ(1, NumTableFilesAtLevel(0));
ASSERT_EQ(0, NumTableFilesAtLevel(1));
dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
true /* disallow_trivial_move */);
ASSERT_EQ(0, NumTableFilesAtLevel(0));
ASSERT_EQ(1, NumTableFilesAtLevel(1));
ASSERT_EQ(0, TestGetTickerCount(opts, COMPACTION_RANGE_DEL_DROP_OBSOLETE));
db_->ReleaseSnapshot(snapshot);
}
TEST_F(DBRangeDelTest, CompactionOutputFilesExactlyFilled) {
// regression test for exactly filled compaction output files. Previously
// another file would be generated containing all range deletions, which
// could invalidate the non-overlapping file boundary invariant.
const int kNumPerFile = 4, kNumFiles = 2, kFileBytes = 9 << 10;
Options options = CurrentOptions();
options.disable_auto_compactions = true;
options.level0_file_num_compaction_trigger = kNumFiles;
options.memtable_factory.reset(new SpecialSkipListFactory(kNumPerFile));
options.num_levels = 2;
options.target_file_size_base = kFileBytes;
BlockBasedTableOptions table_options;
table_options.block_size_deviation = 50; // each block holds two keys
options.table_factory.reset(NewBlockBasedTableFactory(table_options));
Reopen(options);
// snapshot protects range tombstone from dropping due to becoming obsolete.
const Snapshot* snapshot = db_->GetSnapshot();
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), Key(0), Key(1));
Random rnd(301);
for (int i = 0; i < kNumFiles; ++i) {
std::vector<std::string> values;
// Write 12K (4 values, each 3K)
for (int j = 0; j < kNumPerFile; j++) {
values.push_back(RandomString(&rnd, 3 << 10));
ASSERT_OK(Put(Key(i * kNumPerFile + j), values[j]));
if (j == 0 && i > 0) {
dbfull()->TEST_WaitForFlushMemTable();
}
}
}
// put extra key to trigger final flush
ASSERT_OK(Put("", ""));
dbfull()->TEST_WaitForFlushMemTable();
ASSERT_EQ(kNumFiles, NumTableFilesAtLevel(0));
ASSERT_EQ(0, NumTableFilesAtLevel(1));
dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
true /* disallow_trivial_move */);
ASSERT_EQ(0, NumTableFilesAtLevel(0));
ASSERT_EQ(2, NumTableFilesAtLevel(1));
db_->ReleaseSnapshot(snapshot);
}
TEST_F(DBRangeDelTest, MaxCompactionBytesCutsOutputFiles) {
// Ensures range deletion spanning multiple compaction output files that are
// cut by max_compaction_bytes will have non-overlapping key-ranges.
// https://github.com/facebook/rocksdb/issues/1778
const int kNumFiles = 2, kNumPerFile = 1 << 8, kBytesPerVal = 1 << 12;
Options opts = CurrentOptions();
opts.comparator = test::Uint64Comparator();
opts.disable_auto_compactions = true;
opts.level0_file_num_compaction_trigger = kNumFiles;
opts.max_compaction_bytes = kNumPerFile * kBytesPerVal;
opts.memtable_factory.reset(new SpecialSkipListFactory(kNumPerFile));
// Want max_compaction_bytes to trigger the end of compaction output file, not
// target_file_size_base, so make the latter much bigger
opts.target_file_size_base = 100 * opts.max_compaction_bytes;
Reopen(opts);
// snapshot protects range tombstone from dropping due to becoming obsolete.
const Snapshot* snapshot = db_->GetSnapshot();
// It spans the whole key-range, thus will be included in all output files
ASSERT_OK(db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
GetNumericStr(0),
GetNumericStr(kNumFiles * kNumPerFile - 1)));
Random rnd(301);
for (int i = 0; i < kNumFiles; ++i) {
std::vector<std::string> values;
// Write 1MB (256 values, each 4K)
for (int j = 0; j < kNumPerFile; j++) {
values.push_back(RandomString(&rnd, kBytesPerVal));
ASSERT_OK(Put(GetNumericStr(kNumPerFile * i + j), values[j]));
}
// extra entry to trigger SpecialSkipListFactory's flush
ASSERT_OK(Put(GetNumericStr(kNumPerFile), ""));
dbfull()->TEST_WaitForFlushMemTable();
ASSERT_EQ(i + 1, NumTableFilesAtLevel(0));
}
dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
true /* disallow_trivial_move */);
ASSERT_EQ(0, NumTableFilesAtLevel(0));
ASSERT_GE(NumTableFilesAtLevel(1), 2);
std::vector<std::vector<FileMetaData>> files;
dbfull()->TEST_GetFilesMetaData(db_->DefaultColumnFamily(), &files);
for (size_t i = 0; i < files[1].size() - 1; ++i) {
ASSERT_TRUE(InternalKeyComparator(opts.comparator)
.Compare(files[1][i].largest, files[1][i + 1].smallest) <
0);
}
db_->ReleaseSnapshot(snapshot);
}
TEST_F(DBRangeDelTest, FlushRangeDelsSameStartKey) {
db_->Put(WriteOptions(), "b1", "val");
ASSERT_OK(
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "c"));
db_->Put(WriteOptions(), "b2", "val");
ASSERT_OK(
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "b"));
// first iteration verifies query correctness in memtable, second verifies
// query correctness for a single SST file
for (int i = 0; i < 2; ++i) {
if (i > 0) {
ASSERT_OK(db_->Flush(FlushOptions()));
ASSERT_EQ(1, NumTableFilesAtLevel(0));
}
std::string value;
ASSERT_TRUE(db_->Get(ReadOptions(), "b1", &value).IsNotFound());
ASSERT_OK(db_->Get(ReadOptions(), "b2", &value));
}
}
TEST_F(DBRangeDelTest, CompactRangeDelsSameStartKey) {
db_->Put(WriteOptions(), "unused", "val"); // prevents empty after compaction
db_->Put(WriteOptions(), "b1", "val");
ASSERT_OK(db_->Flush(FlushOptions()));
ASSERT_OK(
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "c"));
ASSERT_OK(db_->Flush(FlushOptions()));
ASSERT_OK(
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "a", "b"));
ASSERT_OK(db_->Flush(FlushOptions()));
ASSERT_EQ(3, NumTableFilesAtLevel(0));
for (int i = 0; i < 2; ++i) {
if (i > 0) {
dbfull()->TEST_CompactRange(0, nullptr, nullptr, nullptr,
true /* disallow_trivial_move */);
ASSERT_EQ(0, NumTableFilesAtLevel(0));
ASSERT_EQ(1, NumTableFilesAtLevel(1));
}
std::string value;
ASSERT_TRUE(db_->Get(ReadOptions(), "b1", &value).IsNotFound());
}
}
#endif // ROCKSDB_LITE
TEST_F(DBRangeDelTest, FlushRemovesCoveredKeys) {
const int kNum = 300, kRangeBegin = 50, kRangeEnd = 250;
Options opts = CurrentOptions();
opts.comparator = test::Uint64Comparator();
Reopen(opts);
// Write a third before snapshot, a third between snapshot and tombstone, and
// a third after the tombstone. Keys older than snapshot or newer than the
// tombstone should be preserved.
const Snapshot* snapshot = nullptr;
for (int i = 0; i < kNum; ++i) {
if (i == kNum / 3) {
snapshot = db_->GetSnapshot();
} else if (i == 2 * kNum / 3) {
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
GetNumericStr(kRangeBegin), GetNumericStr(kRangeEnd));
}
db_->Put(WriteOptions(), GetNumericStr(i), "val");
}
db_->Flush(FlushOptions());
for (int i = 0; i < kNum; ++i) {
ReadOptions read_opts;
read_opts.ignore_range_deletions = true;
std::string value;
if (i < kRangeBegin || i > kRangeEnd || i < kNum / 3 || i >= 2 * kNum / 3) {
ASSERT_OK(db_->Get(read_opts, GetNumericStr(i), &value));
} else {
ASSERT_TRUE(db_->Get(read_opts, GetNumericStr(i), &value).IsNotFound());
}
}
db_->ReleaseSnapshot(snapshot);
}
// NumTableFilesAtLevel() is not supported in ROCKSDB_LITE
#ifndef ROCKSDB_LITE
TEST_F(DBRangeDelTest, CompactionRemovesCoveredKeys) {
const int kNumPerFile = 100, kNumFiles = 4;
Options opts = CurrentOptions();
opts.comparator = test::Uint64Comparator();
opts.disable_auto_compactions = true;
opts.memtable_factory.reset(new SpecialSkipListFactory(kNumPerFile));
opts.num_levels = 2;
opts.statistics = CreateDBStatistics();
Reopen(opts);
for (int i = 0; i < kNumFiles; ++i) {
if (i > 0) {
// range tombstone covers first half of the previous file
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
GetNumericStr((i - 1) * kNumPerFile),
GetNumericStr((i - 1) * kNumPerFile + kNumPerFile / 2));
}
// Make sure a given key appears in each file so compaction won't be able to
// use trivial move, which would happen if the ranges were non-overlapping.
// Also, we need an extra element since flush is only triggered when the
// number of keys is one greater than SpecialSkipListFactory's limit.
// We choose a key outside the key-range used by the test to avoid conflict.
db_->Put(WriteOptions(), GetNumericStr(kNumPerFile * kNumFiles), "val");
for (int j = 0; j < kNumPerFile; ++j) {
db_->Put(WriteOptions(), GetNumericStr(i * kNumPerFile + j), "val");
}
dbfull()->TEST_WaitForFlushMemTable();
ASSERT_EQ(i + 1, NumTableFilesAtLevel(0));
}
db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
ASSERT_EQ(0, NumTableFilesAtLevel(0));
ASSERT_GT(NumTableFilesAtLevel(1), 0);
ASSERT_EQ((kNumFiles - 1) * kNumPerFile / 2,
TestGetTickerCount(opts, COMPACTION_KEY_DROP_RANGE_DEL));
for (int i = 0; i < kNumFiles; ++i) {
for (int j = 0; j < kNumPerFile; ++j) {
ReadOptions read_opts;
read_opts.ignore_range_deletions = true;
std::string value;
if (i == kNumFiles - 1 || j >= kNumPerFile / 2) {
ASSERT_OK(
db_->Get(read_opts, GetNumericStr(i * kNumPerFile + j), &value));
} else {
ASSERT_TRUE(
db_->Get(read_opts, GetNumericStr(i * kNumPerFile + j), &value)
.IsNotFound());
}
}
}
}
TEST_F(DBRangeDelTest, ValidLevelSubcompactionBoundaries) {
const int kNumPerFile = 100, kNumFiles = 4, kFileBytes = 100 << 10;
Options options = CurrentOptions();
options.level0_file_num_compaction_trigger = kNumFiles;
options.max_bytes_for_level_base = 2 * kFileBytes;
options.max_subcompactions = 4;
options.memtable_factory.reset(new SpecialSkipListFactory(kNumPerFile));
options.num_levels = 3;
options.target_file_size_base = kFileBytes;
options.target_file_size_multiplier = 1;
Reopen(options);
Random rnd(301);
for (int i = 0; i < 2; ++i) {
for (int j = 0; j < kNumFiles; ++j) {
if (i > 0) {
// delete [95,105) in two files, [295,305) in next two
int mid = (j + (1 - j % 2)) * kNumPerFile;
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(mid - 5), Key(mid + 5));
}
std::vector<std::string> values;
// Write 100KB (100 values, each 1K)
for (int k = 0; k < kNumPerFile; k++) {
values.push_back(RandomString(&rnd, 990));
ASSERT_OK(Put(Key(j * kNumPerFile + k), values[k]));
}
// put extra key to trigger flush
ASSERT_OK(Put("", ""));
dbfull()->TEST_WaitForFlushMemTable();
if (j < kNumFiles - 1) {
// background compaction may happen early for kNumFiles'th file
ASSERT_EQ(NumTableFilesAtLevel(0), j + 1);
}
if (j == options.level0_file_num_compaction_trigger - 1) {
// When i == 1, compaction will output some files to L1, at which point
// L1 is not bottommost so range deletions cannot be compacted away. The
// new L1 files must be generated with non-overlapping key ranges even
// though multiple subcompactions see the same ranges deleted, else an
// assertion will fail.
dbfull()->TEST_WaitForCompact();
ASSERT_EQ(NumTableFilesAtLevel(0), 0);
ASSERT_GT(NumTableFilesAtLevel(1), 0);
ASSERT_GT(NumTableFilesAtLevel(2), 0);
}
}
}
}
TEST_F(DBRangeDelTest, ValidUniversalSubcompactionBoundaries) {
const int kNumPerFile = 100, kFilesPerLevel = 4, kNumLevels = 4;
Options options = CurrentOptions();
options.compaction_options_universal.min_merge_width = kFilesPerLevel;
options.compaction_options_universal.max_merge_width = kFilesPerLevel;
options.compaction_options_universal.size_ratio = 10;
options.compaction_style = kCompactionStyleUniversal;
options.level0_file_num_compaction_trigger = kFilesPerLevel;
options.max_subcompactions = 4;
options.memtable_factory.reset(new SpecialSkipListFactory(kNumPerFile));
options.num_levels = kNumLevels;
options.target_file_size_base = kNumPerFile << 10;
options.target_file_size_multiplier = 1;
Reopen(options);
Random rnd(301);
for (int i = 0; i < kNumLevels - 1; ++i) {
for (int j = 0; j < kFilesPerLevel; ++j) {
if (i == kNumLevels - 2) {
// insert range deletions [95,105) in two files, [295,305) in next two
// to prepare L1 for later manual compaction.
int mid = (j + (1 - j % 2)) * kNumPerFile;
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(),
Key(mid - 5), Key(mid + 5));
}
std::vector<std::string> values;
// Write 100KB (100 values, each 1K)
for (int k = 0; k < kNumPerFile; k++) {
values.push_back(RandomString(&rnd, 990));
ASSERT_OK(Put(Key(j * kNumPerFile + k), values[k]));
}
// put extra key to trigger flush
ASSERT_OK(Put("", ""));
dbfull()->TEST_WaitForFlushMemTable();
if (j < kFilesPerLevel - 1) {
// background compaction may happen early for kFilesPerLevel'th file
ASSERT_EQ(NumTableFilesAtLevel(0), j + 1);
}
}
dbfull()->TEST_WaitForCompact();
ASSERT_EQ(NumTableFilesAtLevel(0), 0);
ASSERT_GT(NumTableFilesAtLevel(kNumLevels - 1 - i), kFilesPerLevel - 1);
}
// Now L1-L3 are full, when we compact L1->L2 we should see (1) subcompactions
// happen since input level > 0; (2) range deletions are not dropped since
// output level is not bottommost. If no file boundary assertion fails, that
// probably means universal compaction + subcompaction + range deletion are
// compatible.
ASSERT_OK(dbfull()->RunManualCompaction(
reinterpret_cast<ColumnFamilyHandleImpl*>(db_->DefaultColumnFamily())
->cfd(),
1 /* input_level */, 2 /* output_level */, 0 /* output_path_id */,
nullptr /* begin */, nullptr /* end */, true /* exclusive */,
true /* disallow_trivial_move */));
}
#endif // ROCKSDB_LITE
TEST_F(DBRangeDelTest, CompactionRemovesCoveredMergeOperands) {
const int kNumPerFile = 3, kNumFiles = 3;
Options opts = CurrentOptions();
opts.disable_auto_compactions = true;
opts.memtable_factory.reset(new SpecialSkipListFactory(2 * kNumPerFile));
opts.merge_operator = MergeOperators::CreateUInt64AddOperator();
opts.num_levels = 2;
Reopen(opts);
// Iterates kNumFiles * kNumPerFile + 1 times since flushing the last file
// requires an extra entry.
for (int i = 0; i <= kNumFiles * kNumPerFile; ++i) {
if (i % kNumPerFile == 0 && i / kNumPerFile == kNumFiles - 1) {
// Delete merge operands from all but the last file
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "key",
"key_");
}
std::string val;
PutFixed64(&val, i);
db_->Merge(WriteOptions(), "key", val);
// we need to prevent trivial move using Puts so compaction will actually
// process the merge operands.
db_->Put(WriteOptions(), "prevent_trivial_move", "");
if (i > 0 && i % kNumPerFile == 0) {
dbfull()->TEST_WaitForFlushMemTable();
}
}
ReadOptions read_opts;
read_opts.ignore_range_deletions = true;
std::string expected, actual;
ASSERT_OK(db_->Get(read_opts, "key", &actual));
PutFixed64(&expected, 45); // 1+2+...+9
ASSERT_EQ(expected, actual);
db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
expected.clear();
ASSERT_OK(db_->Get(read_opts, "key", &actual));
uint64_t tmp;
Slice tmp2(actual);
GetFixed64(&tmp2, &tmp);
PutFixed64(&expected, 30); // 6+7+8+9 (earlier operands covered by tombstone)
ASSERT_EQ(expected, actual);
}
// NumTableFilesAtLevel() is not supported in ROCKSDB_LITE
#ifndef ROCKSDB_LITE
TEST_F(DBRangeDelTest, ObsoleteTombstoneCleanup) {
// During compaction to bottommost level, verify range tombstones older than
// the oldest snapshot are removed, while others are preserved.
Options opts = CurrentOptions();
opts.disable_auto_compactions = true;
opts.num_levels = 2;
opts.statistics = CreateDBStatistics();
Reopen(opts);
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr1",
"dr1"); // obsolete after compaction
db_->Put(WriteOptions(), "key", "val");
db_->Flush(FlushOptions());
const Snapshot* snapshot = db_->GetSnapshot();
db_->DeleteRange(WriteOptions(), db_->DefaultColumnFamily(), "dr2",
"dr2"); // protected by snapshot
db_->Put(WriteOptions(), "key", "val");
db_->Flush(FlushOptions());
ASSERT_EQ(2, NumTableFilesAtLevel(0));
ASSERT_EQ(0, NumTableFilesAtLevel(1));
db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
ASSERT_EQ(0, NumTableFilesAtLevel(0));
ASSERT_EQ(1, NumTableFilesAtLevel(1));
ASSERT_EQ(1, TestGetTickerCount(opts, COMPACTION_RANGE_DEL_DROP_OBSOLETE));
db_->ReleaseSnapshot(snapshot);
}
#endif // ROCKSDB_LITE
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
-825
View File
@@ -1,825 +0,0 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same 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/db_test_util.h"
#include "port/port.h"
#include "port/stack_trace.h"
#include "rocksdb/sst_file_manager.h"
#include "util/sst_file_manager_impl.h"
namespace rocksdb {
class DBSSTTest : public DBTestBase {
public:
DBSSTTest() : DBTestBase("/db_sst_test") {}
};
TEST_F(DBSSTTest, DontDeletePendingOutputs) {
Options options;
options.env = env_;
options.create_if_missing = true;
DestroyAndReopen(options);
// Every time we write to a table file, call FOF/POF with full DB scan. This
// will make sure our pending_outputs_ protection work correctly
std::function<void()> purge_obsolete_files_function = [&]() {
JobContext job_context(0);
dbfull()->TEST_LockMutex();
dbfull()->FindObsoleteFiles(&job_context, true /*force*/);
dbfull()->TEST_UnlockMutex();
dbfull()->PurgeObsoleteFiles(job_context);
job_context.Clean();
};
env_->table_write_callback_ = &purge_obsolete_files_function;
for (int i = 0; i < 2; ++i) {
ASSERT_OK(Put("a", "begin"));
ASSERT_OK(Put("z", "end"));
ASSERT_OK(Flush());
}
// If pending output guard does not work correctly, PurgeObsoleteFiles() will
// delete the file that Compaction is trying to create, causing this: error
// db/db_test.cc:975: IO error:
// /tmp/rocksdbtest-1552237650/db_test/000009.sst: No such file or directory
Compact("a", "b");
}
// 1 Create some SST files by inserting K-V pairs into DB
// 2 Close DB and change suffix from ".sst" to ".ldb" for every other SST file
// 3 Open DB and check if all key can be read
TEST_F(DBSSTTest, SSTsWithLdbSuffixHandling) {
Options options = CurrentOptions();
options.write_buffer_size = 110 << 10; // 110KB
options.num_levels = 4;
DestroyAndReopen(options);
Random rnd(301);
int key_id = 0;
for (int i = 0; i < 10; ++i) {
GenerateNewFile(&rnd, &key_id, false);
}
Flush();
Close();
int const num_files = GetSstFileCount(dbname_);
ASSERT_GT(num_files, 0);
std::vector<std::string> filenames;
GetSstFiles(dbname_, &filenames);
int num_ldb_files = 0;
for (size_t i = 0; i < filenames.size(); ++i) {
if (i & 1) {
continue;
}
std::string const rdb_name = dbname_ + "/" + filenames[i];
std::string const ldb_name = Rocks2LevelTableFileName(rdb_name);
ASSERT_TRUE(env_->RenameFile(rdb_name, ldb_name).ok());
++num_ldb_files;
}
ASSERT_GT(num_ldb_files, 0);
ASSERT_EQ(num_files, GetSstFileCount(dbname_));
Reopen(options);
for (int k = 0; k < key_id; ++k) {
ASSERT_NE("NOT_FOUND", Get(Key(k)));
}
Destroy(options);
}
#ifndef ROCKSDB_LITE
TEST_F(DBSSTTest, DontDeleteMovedFile) {
// This test triggers move compaction and verifies that the file is not
// deleted when it's part of move compaction
Options options = CurrentOptions();
options.env = env_;
options.create_if_missing = true;
options.max_bytes_for_level_base = 1024 * 1024; // 1 MB
options.level0_file_num_compaction_trigger =
2; // trigger compaction when we have 2 files
DestroyAndReopen(options);
Random rnd(301);
// Create two 1MB sst files
for (int i = 0; i < 2; ++i) {
// Create 1MB sst file
for (int j = 0; j < 100; ++j) {
ASSERT_OK(Put(Key(i * 50 + j), RandomString(&rnd, 10 * 1024)));
}
ASSERT_OK(Flush());
}
// this should execute both L0->L1 and L1->(move)->L2 compactions
dbfull()->TEST_WaitForCompact();
ASSERT_EQ("0,0,1", FilesPerLevel(0));
// If the moved file is actually deleted (the move-safeguard in
// ~Version::Version() is not there), we get this failure:
// Corruption: Can't access /000009.sst
Reopen(options);
}
// This reproduces a bug where we don't delete a file because when it was
// supposed to be deleted, it was blocked by pending_outputs
// Consider:
// 1. current file_number is 13
// 2. compaction (1) starts, blocks deletion of all files starting with 13
// (pending outputs)
// 3. file 13 is created by compaction (2)
// 4. file 13 is consumed by compaction (3) and file 15 was created. Since file
// 13 has no references, it is put into VersionSet::obsolete_files_
// 5. FindObsoleteFiles() gets file 13 from VersionSet::obsolete_files_. File 13
// is deleted from obsolete_files_ set.
// 6. PurgeObsoleteFiles() tries to delete file 13, but this file is blocked by
// pending outputs since compaction (1) is still running. It is not deleted and
// it is not present in obsolete_files_ anymore. Therefore, we never delete it.
TEST_F(DBSSTTest, DeleteObsoleteFilesPendingOutputs) {
Options options = CurrentOptions();
options.env = env_;
options.write_buffer_size = 2 * 1024 * 1024; // 2 MB
options.max_bytes_for_level_base = 1024 * 1024; // 1 MB
options.level0_file_num_compaction_trigger =
2; // trigger compaction when we have 2 files
options.max_background_flushes = 2;
options.max_background_compactions = 2;
OnFileDeletionListener* listener = new OnFileDeletionListener();
options.listeners.emplace_back(listener);
Reopen(options);
Random rnd(301);
// Create two 1MB sst files
for (int i = 0; i < 2; ++i) {
// Create 1MB sst file
for (int j = 0; j < 100; ++j) {
ASSERT_OK(Put(Key(i * 50 + j), RandomString(&rnd, 10 * 1024)));
}
ASSERT_OK(Flush());
}
// this should execute both L0->L1 and L1->(move)->L2 compactions
dbfull()->TEST_WaitForCompact();
ASSERT_EQ("0,0,1", FilesPerLevel(0));
test::SleepingBackgroundTask blocking_thread;
port::Mutex mutex_;
bool already_blocked(false);
// block the flush
std::function<void()> block_first_time = [&]() {
bool blocking = false;
{
MutexLock l(&mutex_);
if (!already_blocked) {
blocking = true;
already_blocked = true;
}
}
if (blocking) {
blocking_thread.DoSleep();
}
};
env_->table_write_callback_ = &block_first_time;
// Insert 2.5MB data, which should trigger a flush because we exceed
// write_buffer_size. The flush will be blocked with block_first_time
// pending_file is protecting all the files created after
for (int j = 0; j < 256; ++j) {
ASSERT_OK(Put(Key(j), RandomString(&rnd, 10 * 1024)));
}
blocking_thread.WaitUntilSleeping();
ASSERT_OK(dbfull()->TEST_CompactRange(2, nullptr, nullptr));
ASSERT_EQ("0,0,0,1", FilesPerLevel(0));
std::vector<LiveFileMetaData> metadata;
db_->GetLiveFilesMetaData(&metadata);
ASSERT_EQ(metadata.size(), 1U);
auto file_on_L2 = metadata[0].name;
listener->SetExpectedFileName(dbname_ + file_on_L2);
ASSERT_OK(dbfull()->TEST_CompactRange(3, nullptr, nullptr, nullptr,
true /* disallow trivial move */));
ASSERT_EQ("0,0,0,0,1", FilesPerLevel(0));
// finish the flush!
blocking_thread.WakeUp();
blocking_thread.WaitUntilDone();
dbfull()->TEST_WaitForFlushMemTable();
ASSERT_EQ("1,0,0,0,1", FilesPerLevel(0));
metadata.clear();
db_->GetLiveFilesMetaData(&metadata);
ASSERT_EQ(metadata.size(), 2U);
// This file should have been deleted during last compaction
ASSERT_EQ(Status::NotFound(), env_->FileExists(dbname_ + file_on_L2));
listener->VerifyMatchedCount(1);
}
#endif // ROCKSDB_LITE
TEST_F(DBSSTTest, DBWithSstFileManager) {
std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
int files_added = 0;
int files_deleted = 0;
int files_moved = 0;
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"SstFileManagerImpl::OnAddFile", [&](void* arg) { files_added++; });
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"SstFileManagerImpl::OnDeleteFile", [&](void* arg) { files_deleted++; });
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"SstFileManagerImpl::OnMoveFile", [&](void* arg) { files_moved++; });
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
Options options = CurrentOptions();
options.sst_file_manager = sst_file_manager;
DestroyAndReopen(options);
Random rnd(301);
for (int i = 0; i < 25; i++) {
GenerateNewRandomFile(&rnd);
ASSERT_OK(Flush());
dbfull()->TEST_WaitForFlushMemTable();
dbfull()->TEST_WaitForCompact();
// Verify that we are tracking all sst files in dbname_
ASSERT_EQ(sfm->GetTrackedFiles(), GetAllSSTFiles());
}
ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
auto files_in_db = GetAllSSTFiles();
// Verify that we are tracking all sst files in dbname_
ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
// Verify the total files size
uint64_t total_files_size = 0;
for (auto& file_to_size : files_in_db) {
total_files_size += file_to_size.second;
}
ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
// We flushed at least 25 files
ASSERT_GE(files_added, 25);
// Compaction must have deleted some files
ASSERT_GT(files_deleted, 0);
// No files were moved
ASSERT_EQ(files_moved, 0);
Close();
Reopen(options);
ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
// Verify that we track all the files again after the DB is closed and opened
Close();
sst_file_manager.reset(NewSstFileManager(env_));
options.sst_file_manager = sst_file_manager;
sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
Reopen(options);
ASSERT_EQ(sfm->GetTrackedFiles(), files_in_db);
ASSERT_EQ(sfm->GetTotalSize(), total_files_size);
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
}
#ifndef ROCKSDB_LITE
TEST_F(DBSSTTest, RateLimitedDelete) {
Destroy(last_options_);
rocksdb::SyncPoint::GetInstance()->LoadDependency({
{"DBSSTTest::RateLimitedDelete:1",
"DeleteScheduler::BackgroundEmptyTrash"},
});
std::vector<uint64_t> penalties;
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"DeleteScheduler::BackgroundEmptyTrash:Wait",
[&](void* arg) { penalties.push_back(*(static_cast<int*>(arg))); });
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"InstrumentedCondVar::TimedWaitInternal", [&](void* arg) {
// Turn timed wait into a simulated sleep
uint64_t* abs_time_us = static_cast<uint64_t*>(arg);
int64_t cur_time = 0;
env_->GetCurrentTime(&cur_time);
if (*abs_time_us > static_cast<uint64_t>(cur_time)) {
env_->addon_time_.fetch_add(*abs_time_us -
static_cast<uint64_t>(cur_time));
}
// Randomly sleep shortly
env_->addon_time_.fetch_add(
static_cast<uint64_t>(Random::GetTLSInstance()->Uniform(10)));
// Set wait until time to before current to force not to sleep.
int64_t real_cur_time = 0;
Env::Default()->GetCurrentTime(&real_cur_time);
*abs_time_us = static_cast<uint64_t>(real_cur_time);
});
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
env_->no_slowdown_ = true;
env_->time_elapse_only_sleep_ = true;
Options options = CurrentOptions();
options.disable_auto_compactions = true;
options.env = env_;
std::string trash_dir = test::TmpDir(env_) + "/trash";
int64_t rate_bytes_per_sec = 1024 * 10; // 10 Kbs / Sec
Status s;
options.sst_file_manager.reset(NewSstFileManager(
env_, nullptr, trash_dir, rate_bytes_per_sec, false, &s));
ASSERT_OK(s);
auto sfm = static_cast<SstFileManagerImpl*>(options.sst_file_manager.get());
ASSERT_OK(TryReopen(options));
// Create 4 files in L0
for (char v = 'a'; v <= 'd'; v++) {
ASSERT_OK(Put("Key2", DummyString(1024, v)));
ASSERT_OK(Put("Key3", DummyString(1024, v)));
ASSERT_OK(Put("Key4", DummyString(1024, v)));
ASSERT_OK(Put("Key1", DummyString(1024, v)));
ASSERT_OK(Put("Key4", DummyString(1024, v)));
ASSERT_OK(Flush());
}
// We created 4 sst files in L0
ASSERT_EQ("4", FilesPerLevel(0));
std::vector<LiveFileMetaData> metadata;
db_->GetLiveFilesMetaData(&metadata);
// Compaction will move the 4 files in L0 to trash and create 1 L1 file
ASSERT_OK(db_->CompactRange(CompactRangeOptions(), nullptr, nullptr));
ASSERT_EQ("0,1", FilesPerLevel(0));
uint64_t delete_start_time = env_->NowMicros();
// Hold BackgroundEmptyTrash
TEST_SYNC_POINT("DBSSTTest::RateLimitedDelete:1");
sfm->WaitForEmptyTrash();
uint64_t time_spent_deleting = env_->NowMicros() - delete_start_time;
uint64_t total_files_size = 0;
uint64_t expected_penlty = 0;
ASSERT_EQ(penalties.size(), metadata.size());
for (size_t i = 0; i < metadata.size(); i++) {
total_files_size += metadata[i].size;
expected_penlty = ((total_files_size * 1000000) / rate_bytes_per_sec);
ASSERT_EQ(expected_penlty, penalties[i]);
}
ASSERT_GT(time_spent_deleting, expected_penlty * 0.9);
ASSERT_LT(time_spent_deleting, expected_penlty * 1.1);
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
}
// Create a DB with 2 db_paths, and generate multiple files in the 2
// db_paths using CompactRangeOptions, make sure that files that were
// deleted from first db_path were deleted using DeleteScheduler and
// files in the second path were not.
TEST_F(DBSSTTest, DeleteSchedulerMultipleDBPaths) {
int bg_delete_file = 0;
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"DeleteScheduler::DeleteTrashFile:DeleteFile",
[&](void* arg) { bg_delete_file++; });
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
Options options = CurrentOptions();
options.disable_auto_compactions = true;
options.db_paths.emplace_back(dbname_, 1024 * 100);
options.db_paths.emplace_back(dbname_ + "_2", 1024 * 100);
options.env = env_;
std::string trash_dir = test::TmpDir(env_) + "/trash";
int64_t rate_bytes_per_sec = 1024 * 1024; // 1 Mb / Sec
Status s;
options.sst_file_manager.reset(NewSstFileManager(
env_, nullptr, trash_dir, rate_bytes_per_sec, false, &s));
ASSERT_OK(s);
auto sfm = static_cast<SstFileManagerImpl*>(options.sst_file_manager.get());
DestroyAndReopen(options);
// Create 4 files in L0
for (int i = 0; i < 4; i++) {
ASSERT_OK(Put("Key" + ToString(i), DummyString(1024, 'A')));
ASSERT_OK(Flush());
}
// We created 4 sst files in L0
ASSERT_EQ("4", FilesPerLevel(0));
// Compaction will delete files from L0 in first db path and generate a new
// file in L1 in second db path
CompactRangeOptions compact_options;
compact_options.target_path_id = 1;
Slice begin("Key0");
Slice end("Key3");
ASSERT_OK(db_->CompactRange(compact_options, &begin, &end));
ASSERT_EQ("0,1", FilesPerLevel(0));
// Create 4 files in L0
for (int i = 4; i < 8; i++) {
ASSERT_OK(Put("Key" + ToString(i), DummyString(1024, 'B')));
ASSERT_OK(Flush());
}
ASSERT_EQ("4,1", FilesPerLevel(0));
// Compaction will delete files from L0 in first db path and generate a new
// file in L1 in second db path
begin = "Key4";
end = "Key7";
ASSERT_OK(db_->CompactRange(compact_options, &begin, &end));
ASSERT_EQ("0,2", FilesPerLevel(0));
sfm->WaitForEmptyTrash();
ASSERT_EQ(bg_delete_file, 8);
compact_options.bottommost_level_compaction =
BottommostLevelCompaction::kForce;
ASSERT_OK(db_->CompactRange(compact_options, nullptr, nullptr));
ASSERT_EQ("0,1", FilesPerLevel(0));
sfm->WaitForEmptyTrash();
ASSERT_EQ(bg_delete_file, 8);
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
}
TEST_F(DBSSTTest, DestroyDBWithRateLimitedDelete) {
int bg_delete_file = 0;
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"DeleteScheduler::DeleteTrashFile:DeleteFile",
[&](void* arg) { bg_delete_file++; });
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
Options options = CurrentOptions();
options.disable_auto_compactions = true;
options.env = env_;
DestroyAndReopen(options);
// Create 4 files in L0
for (int i = 0; i < 4; i++) {
ASSERT_OK(Put("Key" + ToString(i), DummyString(1024, 'A')));
ASSERT_OK(Flush());
}
// We created 4 sst files in L0
ASSERT_EQ("4", FilesPerLevel(0));
// Close DB and destroy it using DeleteScheduler
Close();
std::string trash_dir = test::TmpDir(env_) + "/trash";
int64_t rate_bytes_per_sec = 1024 * 1024; // 1 Mb / Sec
Status s;
options.sst_file_manager.reset(NewSstFileManager(
env_, nullptr, trash_dir, rate_bytes_per_sec, false, &s));
ASSERT_OK(s);
ASSERT_OK(DestroyDB(dbname_, options));
auto sfm = static_cast<SstFileManagerImpl*>(options.sst_file_manager.get());
sfm->WaitForEmptyTrash();
// We have deleted the 4 sst files in the delete_scheduler
ASSERT_EQ(bg_delete_file, 4);
}
#endif // ROCKSDB_LITE
TEST_F(DBSSTTest, DBWithMaxSpaceAllowed) {
std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
Options options = CurrentOptions();
options.sst_file_manager = sst_file_manager;
options.disable_auto_compactions = true;
DestroyAndReopen(options);
Random rnd(301);
// Generate a file containing 100 keys.
for (int i = 0; i < 100; i++) {
ASSERT_OK(Put(Key(i), RandomString(&rnd, 50)));
}
ASSERT_OK(Flush());
uint64_t first_file_size = 0;
auto files_in_db = GetAllSSTFiles(&first_file_size);
ASSERT_EQ(sfm->GetTotalSize(), first_file_size);
// Set the maximum allowed space usage to the current total size
sfm->SetMaxAllowedSpaceUsage(first_file_size + 1);
ASSERT_OK(Put("key1", "val1"));
// This flush will cause bg_error_ and will fail
ASSERT_NOK(Flush());
}
TEST_F(DBSSTTest, DBWithMaxSpaceAllowedRandomized) {
// This test will set a maximum allowed space for the DB, then it will
// keep filling the DB until the limit is reached and bg_error_ is set.
// 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, 2, 4, 8, 10};
bool bg_error_set = false;
uint64_t total_sst_files_size = 0;
int reached_max_space_on_flush = 0;
int reached_max_space_on_compaction = 0;
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"DBImpl::FlushMemTableToOutputFile:MaxAllowedSpaceReached",
[&](void* arg) {
bg_error_set = true;
GetAllSSTFiles(&total_sst_files_size);
reached_max_space_on_flush++;
});
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"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;
rocksdb::SyncPoint::GetInstance()->ClearTrace();
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
std::shared_ptr<SstFileManager> sst_file_manager(NewSstFileManager(env_));
auto sfm = static_cast<SstFileManagerImpl*>(sst_file_manager.get());
Options options = CurrentOptions();
options.sst_file_manager = sst_file_manager;
options.write_buffer_size = 1024 * 512; // 512 Kb
DestroyAndReopen(options);
Random rnd(301);
sfm->SetMaxAllowedSpaceUsage(limit_mb * 1024 * 1024);
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, limit_mb * 1024 * 1024 * 2);
}
ASSERT_TRUE(bg_error_set);
ASSERT_GE(total_sst_files_size, limit_mb * 1024 * 1024);
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
}
ASSERT_GT(reached_max_space_on_flush, 0);
ASSERT_GT(reached_max_space_on_compaction, 0);
}
#ifndef ROCKSDB_LITE
TEST_F(DBSSTTest, OpenDBWithInfiniteMaxOpenFiles) {
// Open DB with infinite max open files
// - First iteration use 1 thread to open files
// - Second iteration use 5 threads to open files
for (int iter = 0; iter < 2; iter++) {
Options options;
options.create_if_missing = true;
options.write_buffer_size = 100000;
options.disable_auto_compactions = true;
options.max_open_files = -1;
if (iter == 0) {
options.max_file_opening_threads = 1;
} else {
options.max_file_opening_threads = 5;
}
options = CurrentOptions(options);
DestroyAndReopen(options);
// Create 12 Files in L0 (then move then to L2)
for (int i = 0; i < 12; i++) {
std::string k = "L2_" + Key(i);
ASSERT_OK(Put(k, k + std::string(1000, 'a')));
ASSERT_OK(Flush());
}
CompactRangeOptions compact_options;
compact_options.change_level = true;
compact_options.target_level = 2;
db_->CompactRange(compact_options, nullptr, nullptr);
// Create 12 Files in L0
for (int i = 0; i < 12; i++) {
std::string k = "L0_" + Key(i);
ASSERT_OK(Put(k, k + std::string(1000, 'a')));
ASSERT_OK(Flush());
}
Close();
// Reopening the DB will load all exisitng files
Reopen(options);
ASSERT_EQ("12,0,12", FilesPerLevel(0));
std::vector<std::vector<FileMetaData>> files;
dbfull()->TEST_GetFilesMetaData(db_->DefaultColumnFamily(), &files);
for (const auto& level : files) {
for (const auto& file : level) {
ASSERT_TRUE(file.table_reader_handle != nullptr);
}
}
for (int i = 0; i < 12; i++) {
ASSERT_EQ(Get("L0_" + Key(i)), "L0_" + Key(i) + std::string(1000, 'a'));
ASSERT_EQ(Get("L2_" + Key(i)), "L2_" + Key(i) + std::string(1000, 'a'));
}
}
}
TEST_F(DBSSTTest, GetTotalSstFilesSize) {
Options options = CurrentOptions();
options.disable_auto_compactions = true;
options.compression = kNoCompression;
DestroyAndReopen(options);
// Generate 5 files in L0
for (int i = 0; i < 5; i++) {
for (int j = 0; j < 10; j++) {
std::string val = "val_file_" + ToString(i);
ASSERT_OK(Put(Key(j), val));
}
Flush();
}
ASSERT_EQ("5", FilesPerLevel(0));
std::vector<LiveFileMetaData> live_files_meta;
dbfull()->GetLiveFilesMetaData(&live_files_meta);
ASSERT_EQ(live_files_meta.size(), 5);
uint64_t single_file_size = live_files_meta[0].size;
uint64_t live_sst_files_size = 0;
uint64_t total_sst_files_size = 0;
for (const auto& file_meta : live_files_meta) {
live_sst_files_size += file_meta.size;
}
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 5
// Total SST files = 5
ASSERT_EQ(live_sst_files_size, 5 * single_file_size);
ASSERT_EQ(total_sst_files_size, 5 * single_file_size);
// hold current version
std::unique_ptr<Iterator> iter1(dbfull()->NewIterator(ReadOptions()));
// Compact 5 files into 1 file in L0
ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
ASSERT_EQ("0,1", FilesPerLevel(0));
live_files_meta.clear();
dbfull()->GetLiveFilesMetaData(&live_files_meta);
ASSERT_EQ(live_files_meta.size(), 1);
live_sst_files_size = 0;
total_sst_files_size = 0;
for (const auto& file_meta : live_files_meta) {
live_sst_files_size += file_meta.size;
}
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 1 (compacted file)
// Total SST files = 6 (5 original files + compacted file)
ASSERT_EQ(live_sst_files_size, 1 * single_file_size);
ASSERT_EQ(total_sst_files_size, 6 * single_file_size);
// hold current version
std::unique_ptr<Iterator> iter2(dbfull()->NewIterator(ReadOptions()));
// Delete all keys and compact, this will delete all live files
for (int i = 0; i < 10; i++) {
ASSERT_OK(Delete(Key(i)));
}
Flush();
ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
ASSERT_EQ("", FilesPerLevel(0));
live_files_meta.clear();
dbfull()->GetLiveFilesMetaData(&live_files_meta);
ASSERT_EQ(live_files_meta.size(), 0);
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 0
// Total SST files = 6 (5 original files + compacted file)
ASSERT_EQ(total_sst_files_size, 6 * single_file_size);
iter1.reset();
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 0
// Total SST files = 1 (compacted file)
ASSERT_EQ(total_sst_files_size, 1 * single_file_size);
iter2.reset();
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 0
// Total SST files = 0
ASSERT_EQ(total_sst_files_size, 0);
}
TEST_F(DBSSTTest, GetTotalSstFilesSizeVersionsFilesShared) {
Options options = CurrentOptions();
options.disable_auto_compactions = true;
options.compression = kNoCompression;
DestroyAndReopen(options);
// Generate 5 files in L0
for (int i = 0; i < 5; i++) {
ASSERT_OK(Put(Key(i), "val"));
Flush();
}
ASSERT_EQ("5", FilesPerLevel(0));
std::vector<LiveFileMetaData> live_files_meta;
dbfull()->GetLiveFilesMetaData(&live_files_meta);
ASSERT_EQ(live_files_meta.size(), 5);
uint64_t single_file_size = live_files_meta[0].size;
uint64_t live_sst_files_size = 0;
uint64_t total_sst_files_size = 0;
for (const auto& file_meta : live_files_meta) {
live_sst_files_size += file_meta.size;
}
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 5
// Total SST files = 5
ASSERT_EQ(live_sst_files_size, 5 * single_file_size);
ASSERT_EQ(total_sst_files_size, 5 * single_file_size);
// hold current version
std::unique_ptr<Iterator> iter1(dbfull()->NewIterator(ReadOptions()));
// Compaction will do trivial move from L0 to L1
ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
ASSERT_EQ("0,5", FilesPerLevel(0));
live_files_meta.clear();
dbfull()->GetLiveFilesMetaData(&live_files_meta);
ASSERT_EQ(live_files_meta.size(), 5);
live_sst_files_size = 0;
total_sst_files_size = 0;
for (const auto& file_meta : live_files_meta) {
live_sst_files_size += file_meta.size;
}
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 5
// Total SST files = 5 (used in 2 version)
ASSERT_EQ(live_sst_files_size, 5 * single_file_size);
ASSERT_EQ(total_sst_files_size, 5 * single_file_size);
// hold current version
std::unique_ptr<Iterator> iter2(dbfull()->NewIterator(ReadOptions()));
// Delete all keys and compact, this will delete all live files
for (int i = 0; i < 5; i++) {
ASSERT_OK(Delete(Key(i)));
}
Flush();
ASSERT_OK(dbfull()->CompactRange(CompactRangeOptions(), nullptr, nullptr));
ASSERT_EQ("", FilesPerLevel(0));
live_files_meta.clear();
dbfull()->GetLiveFilesMetaData(&live_files_meta);
ASSERT_EQ(live_files_meta.size(), 0);
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 0
// Total SST files = 5 (used in 2 version)
ASSERT_EQ(total_sst_files_size, 5 * single_file_size);
iter1.reset();
iter2.reset();
ASSERT_TRUE(dbfull()->GetIntProperty("rocksdb.total-sst-files-size",
&total_sst_files_size));
// Live SST files = 0
// Total SST files = 0
ASSERT_EQ(total_sst_files_size, 0);
}
#endif // ROCKSDB_LITE
} // namespace rocksdb
int main(int argc, char** argv) {
rocksdb::port::InstallStackTraceHandler();
::testing::InitGoogleTest(&argc, argv);
return RUN_ALL_TESTS();
}
+3 -37
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -138,14 +138,14 @@ TEST_F(DBTablePropertiesTest, GetPropertiesOfTablesInRange) {
Options options;
options.create_if_missing = true;
options.write_buffer_size = 4096;
options.max_write_buffer_number = 3;
options.max_write_buffer_number = 8;
options.level0_file_num_compaction_trigger = 2;
options.level0_slowdown_writes_trigger = 2;
options.level0_stop_writes_trigger = 4;
options.target_file_size_base = 2048;
options.max_bytes_for_level_base = 10240;
options.max_bytes_for_level_multiplier = 4;
options.hard_pending_compaction_bytes_limit = 16 * 1024;
options.soft_rate_limit = 1.1;
options.num_levels = 8;
DestroyAndReopen(options);
@@ -155,12 +155,6 @@ TEST_F(DBTablePropertiesTest, GetPropertiesOfTablesInRange) {
ASSERT_OK(Put(test::RandomKey(&rnd, 5), RandomString(&rnd, 102)));
}
Flush();
dbfull()->TEST_WaitForCompact();
if (NumTableFilesAtLevel(0) == 0) {
ASSERT_OK(Put(test::RandomKey(&rnd, 5), RandomString(&rnd, 102)));
Flush();
}
db_->PauseBackgroundWork();
// Ensure that we have at least L0, L1 and L2
@@ -221,34 +215,6 @@ TEST_F(DBTablePropertiesTest, GetPropertiesOfTablesInRange) {
TestGetPropertiesOfTablesInRange(std::move(ranges));
}
}
TEST_F(DBTablePropertiesTest, GetColumnFamilyNameProperty) {
std::string kExtraCfName = "pikachu";
CreateAndReopenWithCF({kExtraCfName}, CurrentOptions());
// Create one table per CF, then verify it was created with the column family
// name property.
for (int cf = 0; cf < 2; ++cf) {
Put(cf, "key", "val");
Flush(cf);
TablePropertiesCollection fname_to_props;
ASSERT_OK(db_->GetPropertiesOfAllTables(handles_[cf], &fname_to_props));
ASSERT_EQ(1U, fname_to_props.size());
std::string expected_cf_name;
if (cf > 0) {
expected_cf_name = kExtraCfName;
} else {
expected_cf_name = kDefaultColumnFamilyName;
}
ASSERT_EQ(expected_cf_name,
fname_to_props.begin()->second->column_family_name);
ASSERT_EQ(cf, static_cast<uint32_t>(
fname_to_props.begin()->second->column_family_id));
}
}
} // namespace rocksdb
#endif // ROCKSDB_LITE
+3 -155
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -168,7 +168,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
char buf3[32];
char buf4[32];
snprintf(buf1, sizeof(buf1), "00a0%016d", i * 5);
snprintf(buf3, sizeof(buf3), "00b0%016d", i * 5);
snprintf(buf3, sizeof(buf1), "00b0%016d", i * 5);
Slice key(buf1, 20);
ASSERT_OK(Put(1, key, value));
@@ -181,7 +181,7 @@ TEST_F(DBTestTailingIterator, TailingIteratorTrimSeekToNext) {
if (i == 299) {
file_iters_deleted = true;
}
snprintf(buf4, sizeof(buf4), "00a0%016d", i * 5 / 2);
snprintf(buf4, sizeof(buf1), "00a0%016d", i * 5 / 2);
Slice target(buf4, 20);
iterh->Seek(target);
ASSERT_TRUE(iter->Valid());
@@ -305,7 +305,6 @@ TEST_F(DBTestTailingIterator, TailingIteratorPrefixSeek) {
options.disable_auto_compactions = true;
options.prefix_extractor.reset(NewFixedPrefixTransform(2));
options.memtable_factory.reset(NewHashSkipListRepFactory(16));
options.allow_concurrent_memtable_write = false;
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -429,59 +428,6 @@ TEST_F(DBTestTailingIterator, TailingIteratorUpperBound) {
ASSERT_EQ(0, immutable_seeks);
}
TEST_F(DBTestTailingIterator, TailingIteratorGap) {
// level 1: [20, 25] [35, 40]
// level 2: [10 - 15] [45 - 50]
// level 3: [20, 30, 40]
// Previously there is a bug in tailing_iterator that if there is a gap in
// lower level, the key will be skipped if it is within the range between
// the largest key of index n file and the smallest key of index n+1 file
// if both file fit in that gap. In this example, 25 < key < 35
// https://github.com/facebook/rocksdb/issues/1372
CreateAndReopenWithCF({"pikachu"}, CurrentOptions());
ReadOptions read_options;
read_options.tailing = true;
ASSERT_OK(Put(1, "20", "20"));
ASSERT_OK(Put(1, "30", "30"));
ASSERT_OK(Put(1, "40", "40"));
ASSERT_OK(Flush(1));
MoveFilesToLevel(3, 1);
ASSERT_OK(Put(1, "10", "10"));
ASSERT_OK(Put(1, "15", "15"));
ASSERT_OK(Flush(1));
ASSERT_OK(Put(1, "45", "45"));
ASSERT_OK(Put(1, "50", "50"));
ASSERT_OK(Flush(1));
MoveFilesToLevel(2, 1);
ASSERT_OK(Put(1, "20", "20"));
ASSERT_OK(Put(1, "25", "25"));
ASSERT_OK(Flush(1));
ASSERT_OK(Put(1, "35", "35"));
ASSERT_OK(Put(1, "40", "40"));
ASSERT_OK(Flush(1));
MoveFilesToLevel(1, 1);
ColumnFamilyMetaData meta;
db_->GetColumnFamilyMetaData(handles_[1], &meta);
std::unique_ptr<Iterator> it(db_->NewIterator(read_options, handles_[1]));
it->Seek("30");
ASSERT_TRUE(it->Valid());
ASSERT_EQ("30", it->key().ToString());
it->Next();
ASSERT_TRUE(it->Valid());
ASSERT_EQ("35", it->key().ToString());
it->Next();
ASSERT_TRUE(it->Valid());
ASSERT_EQ("40", it->key().ToString());
}
TEST_F(DBTestTailingIterator, ManagedTailingIteratorSingle) {
ReadOptions read_options;
read_options.tailing = true;
@@ -626,7 +572,6 @@ TEST_F(DBTestTailingIterator, ManagedTailingIteratorPrefixSeek) {
options.disable_auto_compactions = true;
options.prefix_extractor.reset(NewFixedPrefixTransform(2));
options.memtable_factory.reset(NewHashSkipListRepFactory(16));
options.allow_concurrent_memtable_write = false;
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -709,103 +654,6 @@ TEST_F(DBTestTailingIterator, ManagedTailingIteratorSeekToSame) {
ASSERT_EQ(found, iter->key().ToString());
}
TEST_F(DBTestTailingIterator, ForwardIteratorVersionProperty) {
Options options = CurrentOptions();
options.write_buffer_size = 1000;
ReadOptions read_options;
read_options.tailing = true;
Put("foo", "bar");
uint64_t v1, v2, v3, v4;
{
std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
iter->Seek("foo");
std::string prop_value;
ASSERT_OK(iter->GetProperty("rocksdb.iterator.super-version-number",
&prop_value));
v1 = static_cast<uint64_t>(std::atoi(prop_value.c_str()));
Put("foo1", "bar1");
Flush();
ASSERT_OK(iter->GetProperty("rocksdb.iterator.super-version-number",
&prop_value));
v2 = static_cast<uint64_t>(std::atoi(prop_value.c_str()));
iter->Seek("f");
ASSERT_OK(iter->GetProperty("rocksdb.iterator.super-version-number",
&prop_value));
v3 = static_cast<uint64_t>(std::atoi(prop_value.c_str()));
ASSERT_EQ(v1, v2);
ASSERT_GT(v3, v2);
}
{
std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
iter->Seek("foo");
std::string prop_value;
ASSERT_OK(iter->GetProperty("rocksdb.iterator.super-version-number",
&prop_value));
v4 = static_cast<uint64_t>(std::atoi(prop_value.c_str()));
}
ASSERT_EQ(v3, v4);
}
TEST_F(DBTestTailingIterator, SeekWithUpperBoundBug) {
ReadOptions read_options;
read_options.tailing = true;
const Slice upper_bound("cc", 3);
read_options.iterate_upper_bound = &upper_bound;
// 1st L0 file
ASSERT_OK(db_->Put(WriteOptions(), "aa", "SEEN"));
ASSERT_OK(Flush());
// 2nd L0 file
ASSERT_OK(db_->Put(WriteOptions(), "zz", "NOT-SEEN"));
ASSERT_OK(Flush());
std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
iter->Seek("aa");
ASSERT_TRUE(iter->Valid());
ASSERT_EQ(iter->key().ToString(), "aa");
}
TEST_F(DBTestTailingIterator, SeekToFirstWithUpperBoundBug) {
ReadOptions read_options;
read_options.tailing = true;
const Slice upper_bound("cc", 3);
read_options.iterate_upper_bound = &upper_bound;
// 1st L0 file
ASSERT_OK(db_->Put(WriteOptions(), "aa", "SEEN"));
ASSERT_OK(Flush());
// 2nd L0 file
ASSERT_OK(db_->Put(WriteOptions(), "zz", "NOT-SEEN"));
ASSERT_OK(Flush());
std::unique_ptr<Iterator> iter(db_->NewIterator(read_options));
iter->SeekToFirst();
ASSERT_TRUE(iter->Valid());
ASSERT_EQ(iter->key().ToString(), "aa");
iter->Next();
ASSERT_FALSE(iter->Valid());
iter->SeekToFirst();
ASSERT_TRUE(iter->Valid());
ASSERT_EQ(iter->key().ToString(), "aa");
}
} // namespace rocksdb
#endif // !defined(ROCKSDB_LITE)
+6024 -1036
View File
File diff suppressed because it is too large. Load diff
-2235
View File
File diff suppressed because it is too large. Load diff
+66 -310
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -8,7 +8,6 @@
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include "db/db_test_util.h"
#include "db/forward_iterator.h"
namespace rocksdb {
@@ -20,7 +19,7 @@ SpecialEnv::SpecialEnv(Env* base)
sleep_counter_(this),
addon_time_(0),
time_elapse_only_sleep_(false),
no_slowdown_(false) {
no_sleep_(false) {
delay_sstable_sync_.store(false, std::memory_order_release);
drop_writes_.store(false, std::memory_order_release);
no_space_.store(false, std::memory_order_release);
@@ -31,8 +30,6 @@ SpecialEnv::SpecialEnv(Env* base)
manifest_write_error_.store(false, std::memory_order_release);
log_write_error_.store(false, std::memory_order_release);
random_file_open_counter_.store(0, std::memory_order_relaxed);
delete_count_.store(0, std::memory_order_relaxed);
num_open_wal_file_.store(0);
log_write_slowdown_ = 0;
bytes_written_ = 0;
sync_counter_ = 0;
@@ -76,65 +73,63 @@ DBTestBase::~DBTestBase() {
delete env_;
}
bool DBTestBase::ShouldSkipOptions(int option_config, int skip_mask) {
#ifdef ROCKSDB_LITE
// These options are not supported in ROCKSDB_LITE
if (option_config == kHashSkipList ||
option_config == kPlainTableFirstBytePrefix ||
option_config == kPlainTableCappedPrefix ||
option_config == kPlainTableCappedPrefixNonMmap ||
option_config == kPlainTableAllBytesPrefix ||
option_config == kVectorRep || option_config == kHashLinkList ||
option_config == kHashCuckoo || option_config == kUniversalCompaction ||
option_config == kUniversalCompactionMultiLevel ||
option_config == kUniversalSubcompactions ||
option_config == kFIFOCompaction ||
option_config == kConcurrentSkipList) {
return true;
}
#endif
if ((skip_mask & kSkipUniversalCompaction) &&
(option_config == kUniversalCompaction ||
option_config == kUniversalCompactionMultiLevel)) {
return true;
}
if ((skip_mask & kSkipMergePut) && option_config == kMergePut) {
return true;
}
if ((skip_mask & kSkipNoSeekToLast) &&
(option_config == kHashLinkList || option_config == kHashSkipList)) {
return true;
}
if ((skip_mask & kSkipPlainTable) &&
(option_config == kPlainTableAllBytesPrefix ||
option_config == kPlainTableFirstBytePrefix ||
option_config == kPlainTableCappedPrefix ||
option_config == kPlainTableCappedPrefixNonMmap)) {
return true;
}
if ((skip_mask & kSkipHashIndex) &&
(option_config == kBlockBasedTableWithPrefixHashIndex ||
option_config == kBlockBasedTableWithWholeKeyHashIndex)) {
return true;
}
if ((skip_mask & kSkipHashCuckoo) && (option_config == kHashCuckoo)) {
return true;
}
if ((skip_mask & kSkipFIFOCompaction) && option_config == kFIFOCompaction) {
return true;
}
if ((skip_mask & kSkipMmapReads) && option_config == kWalDirAndMmapReads) {
return true;
}
return false;
}
// Switch to a fresh database with the next option configuration to
// test. Return false if there are no more configurations to test.
bool DBTestBase::ChangeOptions(int skip_mask) {
for (option_config_++; option_config_ < kEnd; option_config_++) {
if (ShouldSkipOptions(option_config_, skip_mask)) {
#ifdef ROCKSDB_LITE
// These options are not supported in ROCKSDB_LITE
if (option_config_ == kHashSkipList ||
option_config_ == kPlainTableFirstBytePrefix ||
option_config_ == kPlainTableCappedPrefix ||
option_config_ == kPlainTableCappedPrefixNonMmap ||
option_config_ == kPlainTableAllBytesPrefix ||
option_config_ == kVectorRep || option_config_ == kHashLinkList ||
option_config_ == kHashCuckoo ||
option_config_ == kUniversalCompaction ||
option_config_ == kUniversalCompactionMultiLevel ||
option_config_ == kUniversalSubcompactions ||
option_config_ == kFIFOCompaction) {
continue;
}
#endif
if ((skip_mask & kSkipDeletesFilterFirst) &&
option_config_ == kDeletesFilterFirst) {
continue;
}
if ((skip_mask & kSkipUniversalCompaction) &&
(option_config_ == kUniversalCompaction ||
option_config_ == kUniversalCompactionMultiLevel)) {
continue;
}
if ((skip_mask & kSkipMergePut) && option_config_ == kMergePut) {
continue;
}
if ((skip_mask & kSkipNoSeekToLast) &&
(option_config_ == kHashLinkList || option_config_ == kHashSkipList)) {
continue;
}
if ((skip_mask & kSkipPlainTable) &&
(option_config_ == kPlainTableAllBytesPrefix ||
option_config_ == kPlainTableFirstBytePrefix ||
option_config_ == kPlainTableCappedPrefix ||
option_config_ == kPlainTableCappedPrefixNonMmap)) {
continue;
}
if ((skip_mask & kSkipHashIndex) &&
(option_config_ == kBlockBasedTableWithPrefixHashIndex ||
option_config_ == kBlockBasedTableWithWholeKeyHashIndex)) {
continue;
}
if ((skip_mask & kSkipHashCuckoo) && (option_config_ == kHashCuckoo)) {
continue;
}
if ((skip_mask & kSkipFIFOCompaction) &&
option_config_ == kFIFOCompaction) {
continue;
}
if ((skip_mask & kSkipMmapReads) && option_config_ == kWalDirAndMmapReads) {
continue;
}
break;
@@ -209,13 +204,6 @@ Options DBTestBase::CurrentOptions(
const anon::OptionsOverride& options_override) {
Options options;
options.write_buffer_size = 4090 * 4096;
options.target_file_size_base = 2 * 1024 * 1024;
options.max_bytes_for_level_base = 10 * 1024 * 1024;
options.max_open_files = 5000;
options.base_background_compactions = -1;
options.wal_recovery_mode = WALRecoveryMode::kTolerateCorruptedTailRecords;
options.compaction_pri = CompactionPri::kByCompensatedSize;
return CurrentOptions(options, options_override);
}
@@ -234,7 +222,6 @@ Options DBTestBase::CurrentOptions(
case kHashSkipList:
options.prefix_extractor.reset(NewFixedPrefixTransform(1));
options.memtable_factory.reset(NewHashSkipListRepFactory(16));
options.allow_concurrent_memtable_write = false;
break;
case kPlainTableFirstBytePrefix:
options.table_factory.reset(new PlainTableFactory());
@@ -266,18 +253,15 @@ Options DBTestBase::CurrentOptions(
break;
case kVectorRep:
options.memtable_factory.reset(new VectorRepFactory(100));
options.allow_concurrent_memtable_write = false;
break;
case kHashLinkList:
options.prefix_extractor.reset(NewFixedPrefixTransform(1));
options.memtable_factory.reset(
NewHashLinkListRepFactory(4, 0, 3, true, 4));
options.allow_concurrent_memtable_write = false;
break;
case kHashCuckoo:
options.memtable_factory.reset(
NewHashCuckooRepFactory(options.write_buffer_size));
options.allow_concurrent_memtable_write = false;
break;
#endif // ROCKSDB_LITE
case kMergePut:
@@ -311,9 +295,11 @@ Options DBTestBase::CurrentOptions(
case kPerfOptions:
options.soft_rate_limit = 2.0;
options.delayed_write_rate = 8 * 1024 * 1024;
options.report_bg_io_stats = true;
// TODO(3.13) -- test more options
break;
case kDeletesFilterFirst:
options.filter_deletes = true;
break;
case kUniversalCompaction:
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = 1;
@@ -347,10 +333,6 @@ Options DBTestBase::CurrentOptions(
options.prefix_extractor.reset(NewNoopTransform());
break;
}
case kBlockBasedTableWithIndexRestartInterval: {
table_options.index_block_restart_interval = 8;
break;
}
case kOptimizeFiltersForHits: {
options.optimize_filters_for_hits = true;
set_block_based_table_factory = true;
@@ -374,11 +356,6 @@ Options DBTestBase::CurrentOptions(
options.max_subcompactions = 4;
break;
}
case kConcurrentSkipList: {
options.allow_concurrent_memtable_write = true;
options.enable_write_thread_adaptive_yield = true;
break;
}
default:
break;
@@ -449,7 +426,7 @@ void DBTestBase::Reopen(const Options& options) {
void DBTestBase::Close() {
for (auto h : handles_) {
db_->DestroyColumnFamilyHandle(h);
delete h;
}
handles_.clear();
delete db_;
@@ -502,15 +479,6 @@ Status DBTestBase::Put(int cf, const Slice& k, const Slice& v,
}
}
Status DBTestBase::Merge(const Slice& k, const Slice& v, WriteOptions wo) {
return db_->Merge(wo, k, v);
}
Status DBTestBase::Merge(int cf, const Slice& k, const Slice& v,
WriteOptions wo) {
return db_->Merge(wo, handles_[cf], k, v);
}
Status DBTestBase::Delete(const std::string& k) {
return db_->Delete(WriteOptions(), k);
}
@@ -599,15 +567,11 @@ std::string DBTestBase::Contents(int cf) {
std::string DBTestBase::AllEntriesFor(const Slice& user_key, int cf) {
Arena arena;
auto options = CurrentOptions();
InternalKeyComparator icmp(options.comparator);
RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
ScopedArenaIterator iter;
if (cf == 0) {
iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg));
iter.set(dbfull()->NewInternalIterator(&arena));
} else {
iter.set(
dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[cf]));
iter.set(dbfull()->NewInternalIterator(&arena, handles_[cf]));
}
InternalKey target(user_key, kMaxSequenceNumber, kTypeValue);
iter->Seek(target.Encode());
@@ -697,14 +661,14 @@ uint64_t DBTestBase::SizeAtLevel(int level) {
return sum;
}
size_t DBTestBase::TotalLiveFiles(int cf) {
int DBTestBase::TotalLiveFiles(int cf) {
ColumnFamilyMetaData cf_meta;
if (cf == 0) {
db_->GetColumnFamilyMetaData(&cf_meta);
} else {
db_->GetColumnFamilyMetaData(handles_[cf], &cf_meta);
}
size_t num_files = 0;
int num_files = 0;
for (auto& level : cf_meta.levels) {
num_files += level.files.size();
}
@@ -732,25 +696,9 @@ int DBTestBase::NumTableFilesAtLevel(int level, int cf) {
return atoi(property.c_str());
}
double DBTestBase::CompressionRatioAtLevel(int level, int cf) {
std::string property;
if (cf == 0) {
// default cfd
EXPECT_TRUE(db_->GetProperty(
"rocksdb.compression-ratio-at-level" + NumberToString(level),
&property));
} else {
EXPECT_TRUE(db_->GetProperty(
handles_[cf],
"rocksdb.compression-ratio-at-level" + NumberToString(level),
&property));
}
return std::stod(property);
}
int DBTestBase::TotalTableFiles(int cf, int levels) {
if (levels == -1) {
levels = (cf == 0) ? db_->NumberLevels() : db_->NumberLevels(handles_[1]);
levels = CurrentOptions().num_levels;
}
int result = 0;
for (int level = 0; level < levels; level++) {
@@ -1008,14 +956,10 @@ UpdateStatus DBTestBase::updateInPlaceNoAction(char* prevValue,
void DBTestBase::validateNumberOfEntries(int numValues, int cf) {
ScopedArenaIterator iter;
Arena arena;
auto options = CurrentOptions();
InternalKeyComparator icmp(options.comparator);
RangeDelAggregator range_del_agg(icmp, {} /* snapshots */);
if (cf != 0) {
iter.set(
dbfull()->NewInternalIterator(&arena, &range_del_agg, handles_[cf]));
iter.set(dbfull()->NewInternalIterator(&arena, handles_[cf]));
} else {
iter.set(dbfull()->NewInternalIterator(&arena, &range_del_agg));
iter.set(dbfull()->NewInternalIterator(&arena));
}
iter->SeekToFirst();
ASSERT_EQ(iter->status().ok(), true);
@@ -1056,192 +1000,4 @@ void DBTestBase::CopyFile(const std::string& source,
ASSERT_OK(destfile->Close());
}
std::unordered_map<std::string, uint64_t> DBTestBase::GetAllSSTFiles(
uint64_t* total_size) {
std::unordered_map<std::string, uint64_t> res;
if (total_size) {
*total_size = 0;
}
std::vector<std::string> files;
env_->GetChildren(dbname_, &files);
for (auto& file_name : files) {
uint64_t number;
FileType type;
std::string file_path = dbname_ + "/" + file_name;
if (ParseFileName(file_name, &number, &type) && type == kTableFile) {
uint64_t file_size = 0;
env_->GetFileSize(file_path, &file_size);
res[file_path] = file_size;
if (total_size) {
*total_size += file_size;
}
}
}
return res;
}
std::vector<std::uint64_t> DBTestBase::ListTableFiles(Env* env,
const std::string& path) {
std::vector<std::string> files;
std::vector<uint64_t> file_numbers;
env->GetChildren(path, &files);
uint64_t number;
FileType type;
for (size_t i = 0; i < files.size(); ++i) {
if (ParseFileName(files[i], &number, &type)) {
if (type == kTableFile) {
file_numbers.push_back(number);
}
}
}
return file_numbers;
}
void DBTestBase::VerifyDBFromMap(std::map<std::string, std::string> true_data,
size_t* total_reads_res, bool tailing_iter,
std::map<std::string, Status> status) {
size_t total_reads = 0;
for (auto& kv : true_data) {
Status s = status[kv.first];
if (s.ok()) {
ASSERT_EQ(Get(kv.first), kv.second);
} else {
std::string value;
ASSERT_EQ(s, db_->Get(ReadOptions(), kv.first, &value));
}
total_reads++;
}
// Normal Iterator
{
int iter_cnt = 0;
ReadOptions ro;
ro.total_order_seek = true;
Iterator* iter = db_->NewIterator(ro);
// Verify Iterator::Next()
iter_cnt = 0;
auto data_iter = true_data.begin();
Status s;
for (iter->SeekToFirst(); iter->Valid(); iter->Next(), data_iter++) {
ASSERT_EQ(iter->key().ToString(), data_iter->first);
Status current_status = status[data_iter->first];
if (!current_status.ok()) {
s = current_status;
}
ASSERT_EQ(iter->status(), s);
if (current_status.ok()) {
ASSERT_EQ(iter->value().ToString(), data_iter->second);
}
iter_cnt++;
total_reads++;
}
ASSERT_EQ(data_iter, true_data.end()) << iter_cnt << " / "
<< true_data.size();
delete iter;
// Verify Iterator::Prev()
// Use a new iterator to make sure its status is clean.
iter = db_->NewIterator(ro);
iter_cnt = 0;
s = Status::OK();
auto data_rev = true_data.rbegin();
for (iter->SeekToLast(); iter->Valid(); iter->Prev(), data_rev++) {
ASSERT_EQ(iter->key().ToString(), data_rev->first);
Status current_status = status[data_rev->first];
if (!current_status.ok()) {
s = current_status;
}
ASSERT_EQ(iter->status(), s);
if (current_status.ok()) {
ASSERT_EQ(iter->value().ToString(), data_rev->second);
}
iter_cnt++;
total_reads++;
}
ASSERT_EQ(data_rev, true_data.rend()) << iter_cnt << " / "
<< true_data.size();
// Verify Iterator::Seek()
for (auto kv : true_data) {
iter->Seek(kv.first);
ASSERT_EQ(kv.first, iter->key().ToString());
ASSERT_EQ(kv.second, iter->value().ToString());
total_reads++;
}
delete iter;
}
if (tailing_iter) {
#ifndef ROCKSDB_LITE
// Tailing iterator
int iter_cnt = 0;
ReadOptions ro;
ro.tailing = true;
ro.total_order_seek = true;
Iterator* iter = db_->NewIterator(ro);
// Verify ForwardIterator::Next()
iter_cnt = 0;
auto data_iter = true_data.begin();
for (iter->SeekToFirst(); iter->Valid(); iter->Next(), data_iter++) {
ASSERT_EQ(iter->key().ToString(), data_iter->first);
ASSERT_EQ(iter->value().ToString(), data_iter->second);
iter_cnt++;
total_reads++;
}
ASSERT_EQ(data_iter, true_data.end()) << iter_cnt << " / "
<< true_data.size();
// Verify ForwardIterator::Seek()
for (auto kv : true_data) {
iter->Seek(kv.first);
ASSERT_EQ(kv.first, iter->key().ToString());
ASSERT_EQ(kv.second, iter->value().ToString());
total_reads++;
}
delete iter;
#endif // ROCKSDB_LITE
}
if (total_reads_res) {
*total_reads_res = total_reads;
}
}
void DBTestBase::VerifyDBInternal(
std::vector<std::pair<std::string, std::string>> true_data) {
Arena arena;
InternalKeyComparator icmp(last_options_.comparator);
RangeDelAggregator range_del_agg(icmp, {});
auto iter = dbfull()->NewInternalIterator(&arena, &range_del_agg);
iter->SeekToFirst();
for (auto p : true_data) {
ASSERT_TRUE(iter->Valid());
ParsedInternalKey ikey;
ASSERT_TRUE(ParseInternalKey(iter->key(), &ikey));
ASSERT_EQ(p.first, ikey.user_key);
ASSERT_EQ(p.second, iter->value());
iter->Next();
};
ASSERT_FALSE(iter->Valid());
iter->~InternalIterator();
}
#ifndef ROCKSDB_LITE
uint64_t DBTestBase::GetNumberOfSstFilesForColumnFamily(
DB* db, std::string column_family_name) {
std::vector<LiveFileMetaData> metadata;
db->GetLiveFilesMetaData(&metadata);
uint64_t result = 0;
for (auto& fileMetadata : metadata) {
result += (fileMetadata.column_family_name == column_family_name);
}
return result;
}
#endif // ROCKSDB_LITE
} // namespace rocksdb
+27 -125
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -19,7 +19,6 @@
#endif
#include <algorithm>
#include <map>
#include <set>
#include <string>
#include <thread>
@@ -28,6 +27,7 @@
#include <vector>
#include "db/db_impl.h"
#include "db/db_test_util.h"
#include "db/dbformat.h"
#include "db/filename.h"
#include "memtable/hash_linklist_rep.h"
@@ -39,8 +39,6 @@
#include "rocksdb/filter_policy.h"
#include "rocksdb/options.h"
#include "rocksdb/slice.h"
#include "rocksdb/sst_file_writer.h"
#include "rocksdb/statistics.h"
#include "rocksdb/table.h"
#include "rocksdb/utilities/checkpoint.h"
#include "table/block_based_table_factory.h"
@@ -193,10 +191,6 @@ class SpecialSkipListFactory : public MemTableRepFactory {
}
virtual const char* Name() const override { return "SkipListFactory"; }
bool IsInsertConcurrentlySupported() const override {
return factory_.IsInsertConcurrentlySupported();
}
private:
SkipListFactory factory_;
int num_entries_flush_;
@@ -225,7 +219,7 @@ class SpecialEnv : public EnvWrapper {
// Drop writes on the floor
return Status::OK();
} else if (env_->no_space_.load(std::memory_order_acquire)) {
return Status::NoSpace("No space left on device");
return Status::IOError("No space left on device");
} else {
env_->bytes_written_ += data.size();
return base_->Append(data);
@@ -289,10 +283,7 @@ class SpecialEnv : public EnvWrapper {
class WalFile : public WritableFile {
public:
WalFile(SpecialEnv* env, unique_ptr<WritableFile>&& b)
: env_(env), base_(std::move(b)) {
env_->num_open_wal_file_.fetch_add(1);
}
virtual ~WalFile() { env_->num_open_wal_file_.fetch_add(-1); }
: env_(env), base_(std::move(b)) {}
Status Append(const Slice& data) override {
#if !(defined NDEBUG) || !defined(OS_WIN)
TEST_SYNC_POINT("SpecialEnv::WalFile::Append:1");
@@ -314,18 +305,7 @@ class SpecialEnv : public EnvWrapper {
return s;
}
Status Truncate(uint64_t size) override { return base_->Truncate(size); }
Status Close() override {
// SyncPoint is not supported in Released Windows Mode.
#if !(defined NDEBUG) || !defined(OS_WIN)
// Check preallocation size
// preallocation size is never passed to base file.
size_t preallocation_size = preallocation_block_size();
TEST_SYNC_POINT_CALLBACK("DBTestWalFile.GetPreallocationStatus",
&preallocation_size);
#endif // !(defined NDEBUG) || !defined(OS_WIN)
return base_->Close();
}
Status Close() override { return base_->Close(); }
Status Flush() override { return base_->Flush(); }
Status Sync() override {
++env_->sync_counter_;
@@ -377,30 +357,23 @@ class SpecialEnv : public EnvWrapper {
class CountingFile : public RandomAccessFile {
public:
CountingFile(unique_ptr<RandomAccessFile>&& target,
anon::AtomicCounter* counter,
std::atomic<size_t>* bytes_read)
: target_(std::move(target)),
counter_(counter),
bytes_read_(bytes_read) {}
anon::AtomicCounter* counter)
: target_(std::move(target)), counter_(counter) {}
virtual Status Read(uint64_t offset, size_t n, Slice* result,
char* scratch) const override {
counter_->Increment();
Status s = target_->Read(offset, n, result, scratch);
*bytes_read_ += result->size();
return s;
return target_->Read(offset, n, result, scratch);
}
private:
unique_ptr<RandomAccessFile> target_;
anon::AtomicCounter* counter_;
std::atomic<size_t>* bytes_read_;
};
Status s = target()->NewRandomAccessFile(f, r, soptions);
random_file_open_counter_++;
if (s.ok() && count_random_reads_) {
r->reset(new CountingFile(std::move(*r), &random_read_counter_,
&random_read_bytes_counter_));
r->reset(new CountingFile(std::move(*r), &random_read_counter_));
}
return s;
}
@@ -432,10 +405,10 @@ class SpecialEnv : public EnvWrapper {
virtual void SleepForMicroseconds(int micros) override {
sleep_counter_.Increment();
if (no_slowdown_ || time_elapse_only_sleep_) {
if (no_sleep_ || time_elapse_only_sleep_) {
addon_time_.fetch_add(micros);
}
if (!no_slowdown_) {
if (!no_sleep_) {
target()->SleepForMicroseconds(micros);
}
}
@@ -461,11 +434,6 @@ class SpecialEnv : public EnvWrapper {
addon_time_.load();
}
virtual Status DeleteFile(const std::string& fname) override {
delete_count_.fetch_add(1);
return target()->DeleteFile(fname);
}
Random rnd_;
port::Mutex rnd_mutex_; // Lock to pretect rnd_
@@ -493,12 +461,8 @@ class SpecialEnv : public EnvWrapper {
// Slow down every log write, in micro-seconds.
std::atomic<int> log_write_slowdown_;
// Number of WAL files that are still open for write.
std::atomic<int> num_open_wal_file_;
bool count_random_reads_;
anon::AtomicCounter random_read_counter_;
std::atomic<size_t> random_read_bytes_counter_;
std::atomic<int> random_file_open_counter_;
bool count_sequential_reads_;
@@ -520,42 +484,13 @@ class SpecialEnv : public EnvWrapper {
std::atomic<int64_t> addon_time_;
std::atomic<int> delete_count_;
bool time_elapse_only_sleep_;
bool no_slowdown_;
bool no_sleep_;
std::atomic<bool> is_wal_sync_thread_safe_{true};
};
#ifndef ROCKSDB_LITE
class OnFileDeletionListener : public EventListener {
public:
OnFileDeletionListener() : matched_count_(0), expected_file_name_("") {}
void SetExpectedFileName(const std::string file_name) {
expected_file_name_ = file_name;
}
void VerifyMatchedCount(size_t expected_value) {
ASSERT_EQ(matched_count_, expected_value);
}
void OnTableFileDeleted(const TableFileDeletionInfo& info) override {
if (expected_file_name_ != "") {
ASSERT_EQ(expected_file_name_, info.file_path);
expected_file_name_ = "";
matched_count_++;
}
}
private:
size_t matched_count_;
std::string expected_file_name_;
};
#endif
class DBTestBase : public testing::Test {
protected:
// Sequence of option configurations to try
@@ -579,21 +514,20 @@ class DBTestBase : public testing::Test {
kWalDirAndMmapReads = 16,
kManifestFileSize = 17,
kPerfOptions = 18,
kHashSkipList = 19,
kUniversalCompaction = 20,
kUniversalCompactionMultiLevel = 21,
kCompressedBlockCache = 22,
kInfiniteMaxOpenFiles = 23,
kxxHashChecksum = 24,
kFIFOCompaction = 25,
kOptimizeFiltersForHits = 26,
kRowCache = 27,
kRecycleLogFiles = 28,
kConcurrentSkipList = 29,
kEnd = 30,
kLevelSubcompactions = 31,
kUniversalSubcompactions = 32,
kBlockBasedTableWithIndexRestartInterval = 33,
kDeletesFilterFirst = 19,
kHashSkipList = 20,
kUniversalCompaction = 21,
kUniversalCompactionMultiLevel = 22,
kCompressedBlockCache = 23,
kInfiniteMaxOpenFiles = 24,
kxxHashChecksum = 25,
kFIFOCompaction = 26,
kOptimizeFiltersForHits = 27,
kRowCache = 28,
kRecycleLogFiles = 29,
kLevelSubcompactions = 30,
kUniversalSubcompactions = 31,
kEnd = 30
};
int option_config_;
@@ -639,8 +573,6 @@ class DBTestBase : public testing::Test {
return std::string(buf);
}
static bool ShouldSkipOptions(int option_config, int skip_mask = kNoSkip);
// Switch to a fresh database with the next option configuration to
// test. Return false if there are no more configurations to test.
bool ChangeOptions(int skip_mask = kNoSkip);
@@ -699,12 +631,6 @@ class DBTestBase : public testing::Test {
Status Put(int cf, const Slice& k, const Slice& v,
WriteOptions wo = WriteOptions());
Status Merge(const Slice& k, const Slice& v,
WriteOptions wo = WriteOptions());
Status Merge(int cf, const Slice& k, const Slice& v,
WriteOptions wo = WriteOptions());
Status Delete(const std::string& k);
Status Delete(int cf, const std::string& k);
@@ -735,15 +661,13 @@ class DBTestBase : public testing::Test {
uint64_t SizeAtLevel(int level);
size_t TotalLiveFiles(int cf = 0);
int TotalLiveFiles(int cf = 0);
size_t CountLiveFiles();
#endif // ROCKSDB_LITE
int NumTableFilesAtLevel(int level, int cf = 0);
double CompressionRatioAtLevel(int level, int cf = 0);
int TotalTableFiles(int cf = 0, int levels = -1);
// Return spread of files per level
@@ -825,28 +749,6 @@ class DBTestBase : public testing::Test {
void CopyFile(const std::string& source, const std::string& destination,
uint64_t size = 0);
std::unordered_map<std::string, uint64_t> GetAllSSTFiles(
uint64_t* total_size = nullptr);
std::vector<std::uint64_t> ListTableFiles(Env* env, const std::string& path);
void VerifyDBFromMap(std::map<std::string, std::string> true_data,
size_t* total_reads_res = nullptr,
bool tailing_iter = false,
std::map<std::string, Status> status = {});
void VerifyDBInternal(
std::vector<std::pair<std::string, std::string>> true_data);
#ifndef ROCKSDB_LITE
uint64_t GetNumberOfSstFilesForColumnFamily(DB* db,
std::string column_family_name);
#endif // ROCKSDB_LITE
uint64_t TestGetTickerCount(const Options& options, Tickers ticker_type) {
return options.statistics->getTickerCount(ticker_type);
}
};
} // namespace rocksdb
+65 -254
View File
@@ -1,4 +1,4 @@
// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// Copyright (c) 2013, Facebook, Inc. All rights reserved.
// This source code is licensed under the BSD-style license found in the
// LICENSE file in the root directory of this source tree. An additional grant
// of patent rights can be found in the PATENTS file in the same directory.
@@ -22,16 +22,12 @@ static std::string CompressibleString(Random* rnd, int len) {
class DBTestUniversalCompactionBase
: public DBTestBase,
public ::testing::WithParamInterface<std::tuple<int, bool>> {
public ::testing::WithParamInterface<int> {
public:
explicit DBTestUniversalCompactionBase(
const std::string& path) : DBTestBase(path) {}
virtual void SetUp() override {
num_levels_ = std::get<0>(GetParam());
exclusive_manual_compaction_ = std::get<1>(GetParam());
}
virtual void SetUp() override { num_levels_ = GetParam(); }
int num_levels_;
bool exclusive_manual_compaction_;
};
class DBTestUniversalCompaction : public DBTestUniversalCompactionBase {
@@ -119,7 +115,8 @@ class DelayFilterFactory : public CompactionFilterFactory {
// Make sure we don't trigger a problem if the trigger conditon is given
// to be 0, which is invalid.
TEST_P(DBTestUniversalCompaction, UniversalCompactionSingleSortedRun) {
Options options = CurrentOptions();
Options options;
options = CurrentOptions(options);
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = num_levels_;
@@ -127,7 +124,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSingleSortedRun) {
options.level0_file_num_compaction_trigger = 0;
options.compaction_options_universal.size_ratio = 10;
options.compaction_options_universal.min_merge_width = 2;
options.compaction_options_universal.max_size_amplification_percent = 0;
options.compaction_options_universal.max_size_amplification_percent = 1;
options.write_buffer_size = 105 << 10; // 105KB
options.arena_block_size = 4 << 10;
@@ -151,74 +148,6 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSingleSortedRun) {
dbfull()->TEST_WaitForCompact();
ASSERT_EQ(NumSortedRuns(0), 1);
}
ASSERT_OK(Put(Key(key_idx), ""));
dbfull()->TEST_WaitForCompact();
ASSERT_EQ(NumSortedRuns(0), 1);
}
TEST_P(DBTestUniversalCompaction, OptimizeFiltersForHits) {
Options options = CurrentOptions();
options.compaction_style = kCompactionStyleUniversal;
options.compaction_options_universal.size_ratio = 5;
options.num_levels = num_levels_;
options.write_buffer_size = 105 << 10; // 105KB
options.arena_block_size = 4 << 10;
options.target_file_size_base = 32 << 10; // 32KB
// trigger compaction if there are >= 4 files
options.level0_file_num_compaction_trigger = 4;
BlockBasedTableOptions bbto;
bbto.cache_index_and_filter_blocks = true;
bbto.filter_policy.reset(NewBloomFilterPolicy(10, false));
bbto.whole_key_filtering = true;
options.table_factory.reset(NewBlockBasedTableFactory(bbto));
options.optimize_filters_for_hits = true;
options.statistics = rocksdb::CreateDBStatistics();
options.memtable_factory.reset(new SpecialSkipListFactory(3));
DestroyAndReopen(options);
// block compaction from happening
env_->SetBackgroundThreads(1, Env::LOW);
test::SleepingBackgroundTask sleeping_task_low;
env_->Schedule(&test::SleepingBackgroundTask::DoSleepTask, &sleeping_task_low,
Env::Priority::LOW);
for (int num = 0; num < options.level0_file_num_compaction_trigger; num++) {
Put(Key(num * 10), "val");
if (num) {
dbfull()->TEST_WaitForFlushMemTable();
}
Put(Key(30 + num * 10), "val");
Put(Key(60 + num * 10), "val");
}
Put("", "");
dbfull()->TEST_WaitForFlushMemTable();
// Query set of non existing keys
for (int i = 5; i < 90; i += 10) {
ASSERT_EQ(Get(Key(i)), "NOT_FOUND");
}
// Make sure bloom filter is used at least once.
ASSERT_GT(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
auto prev_counter = TestGetTickerCount(options, BLOOM_FILTER_USEFUL);
// Make sure bloom filter is used for all but the last L0 file when looking
// up a non-existent key that's in the range of all L0 files.
ASSERT_EQ(Get(Key(35)), "NOT_FOUND");
ASSERT_EQ(prev_counter + NumTableFilesAtLevel(0) - 1,
TestGetTickerCount(options, BLOOM_FILTER_USEFUL));
prev_counter = TestGetTickerCount(options, BLOOM_FILTER_USEFUL);
// Unblock compaction and wait it for happening.
sleeping_task_low.WakeUp();
dbfull()->TEST_WaitForCompact();
// The same queries will not trigger bloom filter
for (int i = 5; i < 90; i += 10) {
ASSERT_EQ(Get(Key(i)), "NOT_FOUND");
}
ASSERT_EQ(prev_counter, TestGetTickerCount(options, BLOOM_FILTER_USEFUL));
}
// TODO(kailiu) The tests on UniversalCompaction has some issues:
@@ -333,12 +262,13 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTrigger) {
}
TEST_P(DBTestUniversalCompaction, UniversalCompactionSizeAmplification) {
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = num_levels_;
options.write_buffer_size = 100 << 10; // 100KB
options.target_file_size_base = 32 << 10; // 32KB
options.level0_file_num_compaction_trigger = 3;
options = CurrentOptions(options);
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -383,12 +313,12 @@ TEST_P(DBTestUniversalCompaction, CompactFilesOnUniversalCompaction) {
ChangeCompactOptions();
Options options;
options.create_if_missing = true;
options.write_buffer_size = kEntrySize * kEntriesPerBuffer;
options.compaction_style = kCompactionStyleLevel;
options.num_levels = 1;
options.target_file_size_base = options.write_buffer_size;
options.compression = kNoCompression;
options = CurrentOptions(options);
options.write_buffer_size = kEntrySize * kEntriesPerBuffer;
CreateAndReopenWithCF({"pikachu"}, options);
ASSERT_EQ(options.compaction_style, kCompactionStyleUniversal);
Random rnd(301);
@@ -446,11 +376,12 @@ TEST_P(DBTestUniversalCompaction, CompactFilesOnUniversalCompaction) {
}
TEST_P(DBTestUniversalCompaction, UniversalCompactionTargetLevel) {
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.write_buffer_size = 100 << 10; // 100KB
options.num_levels = 7;
options.disable_auto_compactions = true;
options = CurrentOptions(options);
DestroyAndReopen(options);
// Generate 3 overlapping files
@@ -475,7 +406,6 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTargetLevel) {
CompactRangeOptions compact_options;
compact_options.change_level = true;
compact_options.target_level = 4;
compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
db_->CompactRange(compact_options, nullptr, nullptr);
ASSERT_EQ("0,0,0,0,1", FilesPerLevel(0));
}
@@ -490,13 +420,14 @@ class DBTestUniversalCompactionMultiLevels
};
TEST_P(DBTestUniversalCompactionMultiLevels, UniversalCompactionMultiLevels) {
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = num_levels_;
options.write_buffer_size = 100 << 10; // 100KB
options.level0_file_num_compaction_trigger = 8;
options.max_background_compactions = 3;
options.target_file_size_base = 32 * 1024;
options = CurrentOptions(options);
CreateAndReopenWithCF({"pikachu"}, options);
// Trigger compaction if size amplification exceeds 110%
@@ -532,7 +463,7 @@ TEST_P(DBTestUniversalCompactionMultiLevels, UniversalCompactionTrivialMove) {
});
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.compaction_options_universal.allow_trivial_move = true;
options.num_levels = 3;
@@ -540,6 +471,7 @@ TEST_P(DBTestUniversalCompactionMultiLevels, UniversalCompactionTrivialMove) {
options.level0_file_num_compaction_trigger = 3;
options.max_background_compactions = 2;
options.target_file_size_base = 32 * 1024;
options = CurrentOptions(options);
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -566,8 +498,7 @@ TEST_P(DBTestUniversalCompactionMultiLevels, UniversalCompactionTrivialMove) {
INSTANTIATE_TEST_CASE_P(DBTestUniversalCompactionMultiLevels,
DBTestUniversalCompactionMultiLevels,
::testing::Combine(::testing::Values(3, 20),
::testing::Bool()));
::testing::Values(3, 20));
class DBTestUniversalCompactionParallel :
public DBTestUniversalCompactionBase {
@@ -578,7 +509,7 @@ class DBTestUniversalCompactionParallel :
};
TEST_P(DBTestUniversalCompactionParallel, UniversalCompactionParallel) {
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = num_levels_;
options.write_buffer_size = 1 << 10; // 1KB
@@ -587,6 +518,7 @@ TEST_P(DBTestUniversalCompactionParallel, UniversalCompactionParallel) {
options.max_background_flushes = 3;
options.target_file_size_base = 1 * 1024;
options.compaction_options_universal.max_size_amplification_percent = 110;
options = CurrentOptions(options);
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -637,135 +569,12 @@ TEST_P(DBTestUniversalCompactionParallel, UniversalCompactionParallel) {
}
}
TEST_P(DBTestUniversalCompactionParallel, PickByFileNumberBug) {
Options options = CurrentOptions();
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = num_levels_;
options.write_buffer_size = 1 * 1024; // 1KB
options.level0_file_num_compaction_trigger = 7;
options.max_background_compactions = 2;
options.target_file_size_base = 1024 * 1024; // 1MB
// Disable size amplifiction compaction
options.compaction_options_universal.max_size_amplification_percent =
UINT_MAX;
DestroyAndReopen(options);
rocksdb::SyncPoint::GetInstance()->LoadDependency(
{{"DBTestUniversalCompactionParallel::PickByFileNumberBug:0",
"BackgroundCallCompaction:0"},
{"UniversalCompactionPicker::PickCompaction:Return",
"DBTestUniversalCompactionParallel::PickByFileNumberBug:1"},
{"DBTestUniversalCompactionParallel::PickByFileNumberBug:2",
"CompactionJob::Run():Start"}});
int total_picked_compactions = 0;
rocksdb::SyncPoint::GetInstance()->SetCallBack(
"UniversalCompactionPicker::PickCompaction:Return", [&](void* arg) {
if (arg) {
total_picked_compactions++;
}
});
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
// Write 7 files to trigger compaction
int key_idx = 1;
for (int i = 1; i <= 70; i++) {
std::string k = Key(key_idx++);
ASSERT_OK(Put(k, k));
if (i % 10 == 0) {
ASSERT_OK(Flush());
}
}
// Wait for the 1st background compaction process to start
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:0");
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:1");
rocksdb::SyncPoint::GetInstance()->ClearTrace();
// Write 3 files while 1st compaction is held
// These 3 files have different sizes to avoid compacting based on size_ratio
int num_keys = 1000;
for (int i = 0; i < 3; i++) {
for (int j = 1; j <= num_keys; j++) {
std::string k = Key(key_idx++);
ASSERT_OK(Put(k, k));
}
ASSERT_OK(Flush());
num_keys -= 100;
}
// Wait for the 2nd background compaction process to start
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:0");
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:1");
// Hold the 1st and 2nd compaction from finishing
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:2");
dbfull()->TEST_WaitForCompact();
// Although 2 compaction threads started, the second one did not compact
// anything because the number of files not being compacted is less than
// level0_file_num_compaction_trigger
EXPECT_EQ(total_picked_compactions, 1);
EXPECT_EQ(TotalTableFiles(), 4);
// Stop SyncPoint and destroy the DB and reopen it again
rocksdb::SyncPoint::GetInstance()->ClearTrace();
rocksdb::SyncPoint::GetInstance()->DisableProcessing();
key_idx = 1;
total_picked_compactions = 0;
DestroyAndReopen(options);
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
// Write 7 files to trigger compaction
for (int i = 1; i <= 70; i++) {
std::string k = Key(key_idx++);
ASSERT_OK(Put(k, k));
if (i % 10 == 0) {
ASSERT_OK(Flush());
}
}
// Wait for the 1st background compaction process to start
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:0");
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:1");
rocksdb::SyncPoint::GetInstance()->ClearTrace();
// Write 8 files while 1st compaction is held
// These 8 files have different sizes to avoid compacting based on size_ratio
num_keys = 1000;
for (int i = 0; i < 8; i++) {
for (int j = 1; j <= num_keys; j++) {
std::string k = Key(key_idx++);
ASSERT_OK(Put(k, k));
}
ASSERT_OK(Flush());
num_keys -= 100;
}
// Wait for the 2nd background compaction process to start
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:0");
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:1");
// Hold the 1st and 2nd compaction from finishing
TEST_SYNC_POINT("DBTestUniversalCompactionParallel::PickByFileNumberBug:2");
dbfull()->TEST_WaitForCompact();
// This time we will trigger a compaction because of size ratio and
// another compaction because of number of files that are not compacted
// greater than 7
EXPECT_GE(total_picked_compactions, 2);
}
INSTANTIATE_TEST_CASE_P(DBTestUniversalCompactionParallel,
DBTestUniversalCompactionParallel,
::testing::Combine(::testing::Values(1, 10),
::testing::Values(false)));
::testing::Values(1, 10));
TEST_P(DBTestUniversalCompaction, UniversalCompactionOptions) {
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.write_buffer_size = 105 << 10; // 105KB
options.arena_block_size = 4 << 10; // 4KB
@@ -773,6 +582,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionOptions) {
options.level0_file_num_compaction_trigger = 4;
options.num_levels = num_levels_;
options.compaction_options_universal.compression_size_percent = -1;
options = CurrentOptions(options);
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -885,13 +695,14 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionCompressRatio1) {
return;
}
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.write_buffer_size = 100 << 10; // 100KB
options.target_file_size_base = 32 << 10; // 32KB
options.level0_file_num_compaction_trigger = 2;
options.num_levels = num_levels_;
options.compaction_options_universal.compression_size_percent = 70;
options = CurrentOptions(options);
DestroyAndReopen(options);
Random rnd(301);
@@ -952,13 +763,14 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionCompressRatio2) {
if (!Snappy_Supported()) {
return;
}
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.write_buffer_size = 100 << 10; // 100KB
options.target_file_size_base = 32 << 10; // 32KB
options.level0_file_num_compaction_trigger = 2;
options.num_levels = num_levels_;
options.compaction_options_universal.compression_size_percent = 95;
options = CurrentOptions(options);
DestroyAndReopen(options);
Random rnd(301);
@@ -994,7 +806,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTrivialMoveTest1) {
});
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.compaction_options_universal.allow_trivial_move = true;
options.num_levels = 2;
@@ -1002,6 +814,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTrivialMoveTest1) {
options.level0_file_num_compaction_trigger = 3;
options.max_background_compactions = 1;
options.target_file_size_base = 32 * 1024;
options = CurrentOptions(options);
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -1038,14 +851,15 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTrivialMoveTest2) {
rocksdb::SyncPoint::GetInstance()->EnableProcessing();
Options options = CurrentOptions();
Options options;
options.compaction_style = kCompactionStyleUniversal;
options.compaction_options_universal.allow_trivial_move = true;
options.num_levels = 15;
options.write_buffer_size = 100 << 10; // 100KB
options.level0_file_num_compaction_trigger = 8;
options.max_background_compactions = 1;
options.max_background_compactions = 4;
options.target_file_size_base = 64 * 1024;
options = CurrentOptions(options);
DestroyAndReopen(options);
CreateAndReopenWithCF({"pikachu"}, options);
@@ -1071,7 +885,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionTrivialMoveTest2) {
}
TEST_P(DBTestUniversalCompaction, UniversalCompactionFourPaths) {
Options options = CurrentOptions();
Options options;
options.db_paths.emplace_back(dbname_, 300 * 1024);
options.db_paths.emplace_back(dbname_ + "_2", 300 * 1024);
options.db_paths.emplace_back(dbname_ + "_3", 500 * 1024);
@@ -1080,10 +894,11 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionFourPaths) {
new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
options.compaction_style = kCompactionStyleUniversal;
options.compaction_options_universal.size_ratio = 5;
options.write_buffer_size = 111 << 10; // 114KB
options.write_buffer_size = 110 << 10; // 105KB
options.arena_block_size = 4 << 10;
options.level0_file_num_compaction_trigger = 2;
options.num_levels = 1;
options = CurrentOptions(options);
std::vector<std::string> filenames;
env_->GetChildren(options.db_paths[1].path, &filenames);
@@ -1118,13 +933,13 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionFourPaths) {
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
// (1, 2, 4) -> (3, 4)
// (1, 2, 4)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[2].path));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
ASSERT_EQ(1, GetSstFileCount(dbname_));
// (1, 3, 4) -> (8)
// (1, 1, 2, 4) -> (8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[3].path));
@@ -1138,22 +953,22 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionFourPaths) {
ASSERT_EQ(1, GetSstFileCount(options.db_paths[3].path));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
// (1, 2, 8) -> (3, 8)
// (1, 2, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[3].path));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
ASSERT_EQ(1, GetSstFileCount(dbname_));
// (1, 3, 8) -> (4, 8)
// (1, 1, 2, 8) -> (4, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[2].path));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[3].path));
// (1, 4, 8) -> (5, 8)
// (1, 4, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[3].path));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[2].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
ASSERT_EQ(1, GetSstFileCount(dbname_));
for (int i = 0; i < key_idx; i++) {
auto v = Get(Key(i));
@@ -1210,12 +1025,16 @@ TEST_P(DBTestUniversalCompaction, IncreaseUniversalCompactionNumLevels) {
for (int i = 0; i <= max_key1; i++) {
// each value is 10K
ASSERT_OK(Put(1, Key(i), RandomString(&rnd, 10000)));
dbfull()->TEST_WaitForFlushMemTable(handles_[1]);
dbfull()->TEST_WaitForCompact();
}
ASSERT_OK(Flush(1));
dbfull()->TEST_WaitForCompact();
int non_level0_num_files = 0;
for (int i = 1; i < options.num_levels; i++) {
non_level0_num_files += NumTableFilesAtLevel(i, 1);
}
ASSERT_EQ(non_level0_num_files, 0);
// Stage 2: reopen with universal compaction, num_levels=4
options.compaction_style = kCompactionStyleUniversal;
options.num_levels = 4;
@@ -1228,8 +1047,6 @@ TEST_P(DBTestUniversalCompaction, IncreaseUniversalCompactionNumLevels) {
for (int i = max_key1 + 1; i <= max_key2; i++) {
// each value is 10K
ASSERT_OK(Put(1, Key(i), RandomString(&rnd, 10000)));
dbfull()->TEST_WaitForFlushMemTable(handles_[1]);
dbfull()->TEST_WaitForCompact();
}
ASSERT_OK(Flush(1));
dbfull()->TEST_WaitForCompact();
@@ -1246,7 +1063,6 @@ TEST_P(DBTestUniversalCompaction, IncreaseUniversalCompactionNumLevels) {
CompactRangeOptions compact_options;
compact_options.change_level = true;
compact_options.target_level = 0;
compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
dbfull()->CompactRange(compact_options, handles_[1], nullptr, nullptr);
// Need to restart it once to remove higher level records in manifest.
ReopenWithColumnFamilies({"default", "pikachu"}, options);
@@ -1260,8 +1076,6 @@ TEST_P(DBTestUniversalCompaction, IncreaseUniversalCompactionNumLevels) {
for (int i = max_key2 + 1; i <= max_key3; i++) {
// each value is 10K
ASSERT_OK(Put(1, Key(i), RandomString(&rnd, 10000)));
dbfull()->TEST_WaitForFlushMemTable(handles_[1]);
dbfull()->TEST_WaitForCompact();
}
ASSERT_OK(Flush(1));
dbfull()->TEST_WaitForCompact();
@@ -1273,17 +1087,19 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSecondPathRatio) {
if (!Snappy_Supported()) {
return;
}
Options options = CurrentOptions();
Options options;
options.db_paths.emplace_back(dbname_, 500 * 1024);
options.db_paths.emplace_back(dbname_ + "_2", 1024 * 1024 * 1024);
options.compaction_style = kCompactionStyleUniversal;
options.compaction_options_universal.size_ratio = 5;
options.write_buffer_size = 111 << 10; // 114KB
options.write_buffer_size = 110 << 10; // 105KB
options.arena_block_size = 4 * 1024;
options.arena_block_size = 4 << 10;
options.level0_file_num_compaction_trigger = 2;
options.num_levels = 1;
options.memtable_factory.reset(
new SpecialSkipListFactory(KNumKeysByGenerateNewFile - 1));
options = CurrentOptions(options);
std::vector<std::string> filenames;
env_->GetChildren(options.db_paths[1].path, &filenames);
@@ -1317,12 +1133,12 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSecondPathRatio) {
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(1, GetSstFileCount(dbname_));
// (1, 2, 4) -> (3, 4)
// (1, 2, 4)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(2, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(2, GetSstFileCount(dbname_));
// (1, 3, 4) -> (8)
// (1, 1, 2, 4) -> (8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
@@ -1337,20 +1153,20 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSecondPathRatio) {
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(1, GetSstFileCount(dbname_));
// (1, 2, 8) -> (3, 8)
// (1, 2, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(1, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(2, GetSstFileCount(dbname_));
// (1, 1, 2, 8) -> (4, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(2, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
// (1, 3, 8) -> (4, 8)
// (1, 4, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(2, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
// (1, 4, 8) -> (5, 8)
GenerateNewFile(&rnd, &key_idx);
ASSERT_EQ(2, GetSstFileCount(options.db_paths[1].path));
ASSERT_EQ(0, GetSstFileCount(dbname_));
ASSERT_EQ(1, GetSstFileCount(dbname_));
for (int i = 0; i < key_idx; i++) {
auto v = Get(Key(i));
@@ -1370,8 +1186,7 @@ TEST_P(DBTestUniversalCompaction, UniversalCompactionSecondPathRatio) {
}
INSTANTIATE_TEST_CASE_P(UniversalCompactionNumLevels, DBTestUniversalCompaction,
::testing::Combine(::testing::Values(1, 3, 5),
::testing::Bool()));
::testing::Values(1, 3, 5));
class DBTestUniversalManualCompactionOutputPathId
: public DBTestUniversalCompactionBase {
@@ -1403,7 +1218,6 @@ TEST_P(DBTestUniversalManualCompactionOutputPathId,
// Full compaction to DB path 0
CompactRangeOptions compact_options;
compact_options.target_path_id = 1;
compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);
ASSERT_EQ(1, TotalLiveFiles(1));
ASSERT_EQ(0, GetSstFileCount(options.db_paths[0].path));
@@ -1426,7 +1240,6 @@ TEST_P(DBTestUniversalManualCompactionOutputPathId,
// Full compaction to DB path 0
compact_options.target_path_id = 0;
compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
db_->CompactRange(compact_options, handles_[1], nullptr, nullptr);
ASSERT_EQ(1, TotalLiveFiles(1));
ASSERT_EQ(1, GetSstFileCount(options.db_paths[0].path));
@@ -1434,15 +1247,13 @@ TEST_P(DBTestUniversalManualCompactionOutputPathId,
// Fail when compacting to an invalid path ID
compact_options.target_path_id = 2;
compact_options.exclusive_manual_compaction = exclusive_manual_compaction_;
ASSERT_TRUE(db_->CompactRange(compact_options, handles_[1], nullptr, nullptr)
.IsInvalidArgument());
}
INSTANTIATE_TEST_CASE_P(DBTestUniversalManualCompactionOutputPathId,
DBTestUniversalManualCompactionOutputPathId,
::testing::Combine(::testing::Values(1, 8),
::testing::Bool()));
::testing::Values(1, 8));
} // namespace rocksdb
+1 -1087
View File
File diff suppressed because it is too large. Load diff
Loaded 100 of 1028 files, more files were not shown because too many files have changed in this diff. Show more