You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
rocksdb/db/compaction/compaction_picker_fifo.cc

454 lines
18 KiB

// Copyright (c) 2011-present, Facebook, Inc. All rights reserved.
// This source code is licensed under both the GPLv2 (found in the
// COPYING file in the root directory) and Apache 2.0 License
// (found in the LICENSE.Apache file in the root directory).
//
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
#include "db/compaction/compaction_picker_fifo.h"
#ifndef ROCKSDB_LITE
#include <cinttypes>
#include <string>
#include <vector>
#include "db/column_family.h"
#include "logging/log_buffer.h"
#include "logging/logging.h"
#include "util/string_util.h"
namespace ROCKSDB_NAMESPACE {
namespace {
uint64_t GetTotalFilesSize(const std::vector<FileMetaData*>& files) {
uint64_t total_size = 0;
for (const auto& f : files) {
total_size += f->fd.file_size;
}
return total_size;
}
} // anonymous namespace
bool FIFOCompactionPicker::NeedsCompaction(
const VersionStorageInfo* vstorage) const {
const int kLevel0 = 0;
return vstorage->CompactionScore(kLevel0) >= 1;
}
Compaction* FIFOCompactionPicker::PickTTLCompaction(
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
LogBuffer* log_buffer) {
assert(mutable_cf_options.ttl > 0);
const int kLevel0 = 0;
const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
uint64_t total_size = GetTotalFilesSize(level_files);
int64_t _current_time;
auto status = ioptions_.clock->GetCurrentTime(&_current_time);
if (!status.ok()) {
ROCKS_LOG_BUFFER(log_buffer,
"[%s] FIFO compaction: Couldn't get current time: %s. "
"Not doing compactions based on TTL. ",
cf_name.c_str(), status.ToString().c_str());
return nullptr;
}
const uint64_t current_time = static_cast<uint64_t>(_current_time);
Fix assertion failure in FIFO compaction with TTL (#5754) Summary: Before this PR, the following sequence of events can cause assertion failure as shown below. Stack trace (partial): ``` (gdb) bt 2 0x00007f59b350ad15 in __assert_fail_base (fmt=<optimized out>, assertion=assertion@entry=0x9f8390 "mark_as_compacted ? !inputs_[i][j]->being_compacted : inputs_[i][j]->being_compacted", file=file@entry=0x9e347c "db/compaction/compaction.cc", line=line@entry=395, function=function@entry=0xa21ec0 <rocksdb::Compaction::MarkFilesBeingCompacted(bool)::__PRETTY_FUNCTION__> "void rocksdb::Compaction::MarkFilesBeingCompacted(bool)") at assert.c:92 3 0x00007f59b350adc3 in __GI___assert_fail (assertion=assertion@entry=0x9f8390 "mark_as_compacted ? !inputs_[i][j]->being_compacted : inputs_[i][j]->being_compacted", file=file@entry=0x9e347c "db/compaction/compaction.cc", line=line@entry=395, function=function@entry=0xa21ec0 <rocksdb::Compaction::MarkFilesBeingCompacted(bool)::__PRETTY_FUNCTION__> "void rocksdb::Compaction::MarkFilesBeingCompacted(bool)") at assert.c:101 4 0x0000000000492ccd in rocksdb::Compaction::MarkFilesBeingCompacted (this=<optimized out>, mark_as_compacted=<optimized out>) at db/compaction/compaction.cc:394 5 0x000000000049467a in rocksdb::Compaction::Compaction (this=0x7f59af013000, vstorage=0x7f581af53030, _immutable_cf_options=..., _mutable_cf_options=..., _inputs=..., _output_level=<optimized out>, _target_file_size=0, _max_compaction_bytes=0, _output_path_id=0, _compression=<incomplete type>, _compression_opts=..., _max_subcompactions=0, _grandparents=..., _manual_compaction=false, _score=4, _deletion_compaction=true, _compaction_reason=rocksdb::CompactionReason::kFIFOTtl) at db/compaction/compaction.cc:241 6 0x00000000004af9bc in rocksdb::FIFOCompactionPicker::PickTTLCompaction (this=0x7f59b31a6900, cf_name=..., mutable_cf_options=..., vstorage=0x7f581af53030, log_buffer=log_buffer@entry=0x7f59b1bfa930) at db/compaction/compaction_picker_fifo.cc:101 7 0x00000000004b0771 in rocksdb::FIFOCompactionPicker::PickCompaction (this=0x7f59b31a6900, cf_name=..., mutable_cf_options=..., vstorage=0x7f581af53030, log_buffer=0x7f59b1bfa930) at db/compaction/compaction_picker_fifo.cc:201 8 0x00000000004838cc in rocksdb::ColumnFamilyData::PickCompaction (this=this@entry=0x7f59b31b3700, mutable_options=..., log_buffer=log_buffer@entry=0x7f59b1bfa930) at db/column_family.cc:933 9 0x00000000004f3645 in rocksdb::DBImpl::BackgroundCompaction (this=this@entry=0x7f59b3176000, made_progress=made_progress@entry=0x7f59b1bfa6bf, job_context=job_context@entry=0x7f59b1bfa760, log_buffer=log_buffer@entry=0x7f59b1bfa930, prepicked_compaction=prepicked_compaction@entry=0x0, thread_pri=rocksdb::Env::LOW) at db/db_impl/db_impl_compaction_flush.cc:2541 10 0x00000000004f5e2a in rocksdb::DBImpl::BackgroundCallCompaction (this=this@entry=0x7f59b3176000, prepicked_compaction=prepicked_compaction@entry=0x0, bg_thread_pri=bg_thread_pri@entry=rocksdb::Env::LOW) at db/db_impl/db_impl_compaction_flush.cc:2312 11 0x00000000004f648e in rocksdb::DBImpl::BGWorkCompaction (arg=<optimized out>) at db/db_impl/db_impl_compaction_flush.cc:2087 ``` This can be caused by the following sequence of events. ``` Time | thr bg_compact_thr1 bg_compact_thr2 | write | flush | mark all l0 as being compacted | write | flush | add cf to queue again | mark all l0 as being | compacted, fail the | assertion V ``` Test plan (on devserver) Since bg_compact_thr1 and bg_compact_thr2 are two threads executing the same code, it is difficult to use sync point dependency to coordinate their execution. Therefore, I choose to use db_stress. ``` $TEST_TMPDIR=/dev/shm/rocksdb ./db_stress --periodic_compaction_seconds=1 --max_background_compactions=20 --format_version=2 --memtablerep=skip_list --max_write_buffer_number=3 --cache_index_and_filter_blocks=1 --reopen=20 --recycle_log_file_num=0 --acquire_snapshot_one_in=10000 --delpercent=4 --log2_keys_per_lock=22 --compaction_ttl=1 --block_size=16384 --use_multiget=1 --compact_files_one_in=1000000 --target_file_size_multiplier=2 --clear_column_family_one_in=0 --max_bytes_for_level_base=10485760 --use_full_merge_v1=1 --target_file_size_base=2097152 --checkpoint_one_in=1000000 --mmap_read=0 --compression_type=zstd --writepercent=35 --readpercent=45 --subcompactions=4 --use_merge=0 --write_buffer_size=4194304 --test_batches_snapshots=0 --db=/dev/shm/rocksdb/rocksdb_crashtest_whitebox --use_direct_reads=0 --compact_range_one_in=1000000 --open_files=-1 --destroy_db_initially=0 --progress_reports=0 --compression_zstd_max_train_bytes=0 --snapshot_hold_ops=100000 --enable_pipelined_write=0 --nooverwritepercent=1 --compression_max_dict_bytes=0 --max_key=1000000 --prefixpercent=5 --flush_one_in=1000000 --ops_per_thread=40000 --index_block_restart_interval=7 --cache_size=1048576 --compaction_style=2 --verify_checksum=1 --delrangepercent=1 --use_direct_io_for_flush_and_compaction=0 ``` This should see no assertion failure. Last but not least, ``` $COMPILE_WITH_ASAN=1 make -j32 all $make check ``` Pull Request resolved: https://github.com/facebook/rocksdb/pull/5754 Differential Revision: D17109791 Pulled By: riversand963 fbshipit-source-id: 25fc46101235add158554e096540b72c324be078
5 years ago
if (!level0_compactions_in_progress_.empty()) {
ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: Already executing compaction. No need "
"to run parallel compactions since compactions are very fast",
cf_name.c_str());
return nullptr;
}
std::vector<CompactionInputFiles> inputs;
inputs.emplace_back();
inputs[0].level = 0;
// avoid underflow
if (current_time > mutable_cf_options.ttl) {
for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
FileMetaData* f = *ritr;
assert(f);
if (f->fd.table_reader && f->fd.table_reader->GetTableProperties()) {
uint64_t creation_time =
f->fd.table_reader->GetTableProperties()->creation_time;
if (creation_time == 0 ||
creation_time >= (current_time - mutable_cf_options.ttl)) {
break;
}
}
total_size -= f->fd.file_size;
inputs[0].files.push_back(f);
}
}
// Return a nullptr and proceed to size-based FIFO compaction if:
// 1. there are no files older than ttl OR
// 2. there are a few files older than ttl, but deleting them will not bring
// the total size to be less than max_table_files_size threshold.
if (inputs[0].files.empty() ||
total_size >
mutable_cf_options.compaction_options_fifo.max_table_files_size) {
return nullptr;
}
for (const auto& f : inputs[0].files) {
uint64_t creation_time = 0;
assert(f);
if (f->fd.table_reader && f->fd.table_reader->GetTableProperties()) {
creation_time = f->fd.table_reader->GetTableProperties()->creation_time;
}
ROCKS_LOG_BUFFER(log_buffer,
"[%s] FIFO compaction: picking file %" PRIu64
" with creation time %" PRIu64 " for deletion",
cf_name.c_str(), f->fd.GetNumber(), creation_time);
}
Compaction* c = new Compaction(
vstorage, ioptions_, mutable_cf_options, mutable_db_options,
std::move(inputs), 0, 0, 0, 0, kNoCompression,
mutable_cf_options.compression_opts, Temperature::kUnknown,
/* max_subcompactions */ 0, {}, /* is manual */ false,
/* trim_ts */ "", vstorage->CompactionScore(0),
/* is deletion compaction */ true, /* l0_files_might_overlap */ true,
CompactionReason::kFIFOTtl);
return c;
}
// The size-based compaction picker for FIFO.
//
// When the entire column family size exceeds max_table_files_size, FIFO will
// try to delete the oldest sst file(s) until the resulting column family size
// is smaller than max_table_files_size.
//
// This function also takes care the case where a DB is migrating from level /
// universal compaction to FIFO compaction. During the migration, the column
// family will also have non-L0 files while FIFO can only create L0 files.
// In this case, this function will first purge the sst files in the bottom-
// most non-empty level first, and the DB will eventually converge to the
// regular FIFO case where there're only L0 files. Note that during the
// migration case, the purge order will only be an approximation of "FIFO"
// as entries inside lower-level files might sometimes be newer than some
// entries inside upper-level files.
Compaction* FIFOCompactionPicker::PickSizeCompaction(
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
LogBuffer* log_buffer, const SequenceNumber earliest_mem_seqno) {
// compute the total size and identify the last non-empty level
int last_level = 0;
uint64_t total_size = 0;
for (int level = 0; level < vstorage->num_levels(); ++level) {
auto level_size = GetTotalFilesSize(vstorage->LevelFiles(level));
total_size += level_size;
if (level_size > 0) {
last_level = level;
}
}
const std::vector<FileMetaData*>& last_level_files =
vstorage->LevelFiles(last_level);
if (last_level == 0 &&
total_size <=
mutable_cf_options.compaction_options_fifo.max_table_files_size) {
// total size not exceeded, try to find intra level 0 compaction if enabled
const std::vector<FileMetaData*>& level0_files = vstorage->LevelFiles(0);
if (mutable_cf_options.compaction_options_fifo.allow_compaction &&
level0_files.size() > 0) {
CompactionInputFiles comp_inputs;
// try to prevent same files from being compacted multiple times, which
// could produce large files that may never TTL-expire. Achieve this by
// disallowing compactions with files larger than memtable (inflate its
// size by 10% to account for uncompressed L0 files that may have size
// slightly greater than memtable size limit).
size_t max_compact_bytes_per_del_file =
static_cast<size_t>(MultiplyCheckOverflow(
static_cast<uint64_t>(mutable_cf_options.write_buffer_size),
1.1));
if (FindIntraL0Compaction(
level0_files,
mutable_cf_options
.level0_file_num_compaction_trigger /* min_files_to_compact */
,
max_compact_bytes_per_del_file,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
mutable_cf_options.max_compaction_bytes, &comp_inputs,
earliest_mem_seqno)) {
Compaction* c = new Compaction(
vstorage, ioptions_, mutable_cf_options, mutable_db_options,
{comp_inputs}, 0, 16 * 1024 * 1024 /* output file size limit */,
0 /* max compaction bytes, not applicable */,
0 /* output path ID */, mutable_cf_options.compression,
mutable_cf_options.compression_opts, Temperature::kUnknown,
0 /* max_subcompactions */, {}, /* is manual */ false,
/* trim_ts */ "", vstorage->CompactionScore(0),
/* is deletion compaction */ false,
/* l0_files_might_overlap */ true,
CompactionReason::kFIFOReduceNumFiles);
return c;
}
}
ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: nothing to do. Total size %" PRIu64
", max size %" PRIu64 "\n",
cf_name.c_str(), total_size,
mutable_cf_options.compaction_options_fifo.max_table_files_size);
return nullptr;
}
if (!level0_compactions_in_progress_.empty()) {
ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: Already executing compaction. No need "
"to run parallel compactions since compactions are very fast",
cf_name.c_str());
return nullptr;
}
std::vector<CompactionInputFiles> inputs;
inputs.emplace_back();
inputs[0].level = last_level;
if (last_level == 0) {
// In L0, right-most files are the oldest files.
for (auto ritr = last_level_files.rbegin(); ritr != last_level_files.rend();
++ritr) {
auto f = *ritr;
total_size -= f->fd.file_size;
inputs[0].files.push_back(f);
char tmp_fsize[16];
AppendHumanBytes(f->fd.GetFileSize(), tmp_fsize, sizeof(tmp_fsize));
ROCKS_LOG_BUFFER(log_buffer,
"[%s] FIFO compaction: picking file %" PRIu64
" with size %s for deletion",
cf_name.c_str(), f->fd.GetNumber(), tmp_fsize);
if (total_size <=
mutable_cf_options.compaction_options_fifo.max_table_files_size) {
break;
}
}
Trigger FIFO file deletion in non L0 only if exceeding max_table_files_size (#10955) Summary: **Context** https://github.com/facebook/rocksdb/pull/10348 allows multi-level FIFO but accidentally made change to the logic of deleting files in `FIFOCompactionPicker::PickSizeCompaction`. With [this](https://github.com/facebook/rocksdb/pull/10348/files#diff-d8fb3d50749aa69b378de447e3d9cf2f48abe0281437f010b5d61365a7b813fdR156) and [this](https://github.com/facebook/rocksdb/pull/10348/files#diff-d8fb3d50749aa69b378de447e3d9cf2f48abe0281437f010b5d61365a7b813fdR235) together, it deletes one file in non-L0 even when `total_size <= mutable_cf_options.compaction_options_fifo.max_table_files_size`, which is incorrect. As a consequence, FIFO exercises more file deletion in our crash testing, which is not able to verify correctly on deleted keys in the file deleted by compaction. This results in errors `error : inconsistent values for key 000000000000239F000000000000012B000000000000028B: expected state has the key, Get() returns NotFound. Verification failed :(` or `Expected state has key 00000000000023A90000000000000003787878, iterator is at key 00000000000023A9000000000000004178 Column family: default, op_logs: S 00000000000023A90000000000000003787878` **Summary**: - Delete file for non-L0 only if `total_size <= mutable_cf_options.compaction_options_fifo.max_table_files_size` - Add some helpful log to LOG file Pull Request resolved: https://github.com/facebook/rocksdb/pull/10955 Test Plan: - Errors repro-ed by ``` ./db_stress --preserve_unverified_changes=1 --acquire_snapshot_one_in=10000 --adaptive_readahead=0 --allow_concurrent_memtable_write=0 --allow_data_in_errors=True --async_io=0 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=1 --backup_max_size=104857600 --backup_one_in=100000 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=10 --bottommost_compression_type=none --bytes_per_sync=0 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=1 --charge_file_metadata=1 --charge_filter_construction=0 --charge_table_reader=1 --checkpoint_one_in=1000000 --checksum_type=kxxHash --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=1000000 --compact_range_one_in=1000000 --compaction_pri=3 --compaction_style=2 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8589934591 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=xpress --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test/rocksdb_crashtest_whitebox --db_write_buffer_size=1048576 --delpercent=0 --delrangepercent=0 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --expected_values_dir=/dev/shm/rocksdb_test/rocksdb_crashtest_expected --fail_if_options_file_error=1 --fifo_allow_compaction=1 --file_checksum_impl=xxh64 --flush_one_in=1000000 --format_version=4 --get_current_wal_file_one_in=0 --get_live_files_one_in=1000000 --get_property_one_in=1000000 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=10 --index_type=2 --ingest_external_file_one_in=1000000 --initial_auto_readahead_size=16384 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=False --log2_keys_per_lock=10 --long_running_snapshots=0 --manual_wal_flush_one_in=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=524288 --max_background_compactions=1 --max_bytes_for_level_base=67108864 --max_key=25000000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=0 --memtable_prefix_bloom_size_ratio=0.01 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --min_write_buffer_number_to_merge=2 --mmap_read=0 --mock_direct_io=True --nooverwritepercent=0 --num_file_reads_for_auto_readahead=2 --open_files=-1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=0 --open_write_fault_one_in=0 --ops_per_thread=40000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=3 --pause_background_one_in=1000000 --periodic_compaction_seconds=0 --prefix_size=7 --prefixpercent=5 --prepopulate_block_cache=0 --preserve_internal_time_seconds=3600 --progress_reports=0 --read_fault_one_in=1000 --readahead_size=0 --readpercent=65 --recycle_log_file_num=1 --reopen=0 --ribbon_starting_level=999 --secondary_cache_fault_one_in=0 --snapshot_hold_ops=100000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --stats_dump_period_sec=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=16777216 --target_file_size_multiplier=1 --test_batches_snapshots=0 --top_level_index_pinning=1 --unpartitioned_pinning=1 --use_direct_io_for_flush_and_compaction=1 --use_direct_reads=1 --use_full_merge_v1=1 --use_merge=0 --use_multiget=0 --use_put_entity_one_in=0 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=1000000 --verify_db_one_in=100000 --verify_iterator_with_expected_state_one_in=0 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=none --write_buffer_size=33554432 --write_dbid_to_manifest=1 --writepercent=20 ``` is gone after this fix - CI Reviewed By: ajkr Differential Revision: D41319441 Pulled By: hx235 fbshipit-source-id: 6939753767007f7449ea7055b1420aabd03d7709
2 years ago
} else if (total_size >
mutable_cf_options.compaction_options_fifo.max_table_files_size) {
// If the last level is non-L0, we actually don't know which file is
// logically the oldest since the file creation time only represents
// when this file was compacted to this level, which is independent
// to when the entries in this file were first inserted.
//
// As a result, we delete files from the left instead. This means the sst
// file with the smallest key will be deleted first. This design decision
// better serves a major type of FIFO use cases where smaller keys are
// associated with older data.
for (const auto& f : last_level_files) {
total_size -= f->fd.file_size;
inputs[0].files.push_back(f);
char tmp_fsize[16];
AppendHumanBytes(f->fd.GetFileSize(), tmp_fsize, sizeof(tmp_fsize));
Trigger FIFO file deletion in non L0 only if exceeding max_table_files_size (#10955) Summary: **Context** https://github.com/facebook/rocksdb/pull/10348 allows multi-level FIFO but accidentally made change to the logic of deleting files in `FIFOCompactionPicker::PickSizeCompaction`. With [this](https://github.com/facebook/rocksdb/pull/10348/files#diff-d8fb3d50749aa69b378de447e3d9cf2f48abe0281437f010b5d61365a7b813fdR156) and [this](https://github.com/facebook/rocksdb/pull/10348/files#diff-d8fb3d50749aa69b378de447e3d9cf2f48abe0281437f010b5d61365a7b813fdR235) together, it deletes one file in non-L0 even when `total_size <= mutable_cf_options.compaction_options_fifo.max_table_files_size`, which is incorrect. As a consequence, FIFO exercises more file deletion in our crash testing, which is not able to verify correctly on deleted keys in the file deleted by compaction. This results in errors `error : inconsistent values for key 000000000000239F000000000000012B000000000000028B: expected state has the key, Get() returns NotFound. Verification failed :(` or `Expected state has key 00000000000023A90000000000000003787878, iterator is at key 00000000000023A9000000000000004178 Column family: default, op_logs: S 00000000000023A90000000000000003787878` **Summary**: - Delete file for non-L0 only if `total_size <= mutable_cf_options.compaction_options_fifo.max_table_files_size` - Add some helpful log to LOG file Pull Request resolved: https://github.com/facebook/rocksdb/pull/10955 Test Plan: - Errors repro-ed by ``` ./db_stress --preserve_unverified_changes=1 --acquire_snapshot_one_in=10000 --adaptive_readahead=0 --allow_concurrent_memtable_write=0 --allow_data_in_errors=True --async_io=0 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=1 --backup_max_size=104857600 --backup_one_in=100000 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=10 --bottommost_compression_type=none --bytes_per_sync=0 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=1 --charge_file_metadata=1 --charge_filter_construction=0 --charge_table_reader=1 --checkpoint_one_in=1000000 --checksum_type=kxxHash --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=1000000 --compact_range_one_in=1000000 --compaction_pri=3 --compaction_style=2 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8589934591 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=xpress --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test/rocksdb_crashtest_whitebox --db_write_buffer_size=1048576 --delpercent=0 --delrangepercent=0 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --expected_values_dir=/dev/shm/rocksdb_test/rocksdb_crashtest_expected --fail_if_options_file_error=1 --fifo_allow_compaction=1 --file_checksum_impl=xxh64 --flush_one_in=1000000 --format_version=4 --get_current_wal_file_one_in=0 --get_live_files_one_in=1000000 --get_property_one_in=1000000 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=10 --index_type=2 --ingest_external_file_one_in=1000000 --initial_auto_readahead_size=16384 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=False --log2_keys_per_lock=10 --long_running_snapshots=0 --manual_wal_flush_one_in=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=524288 --max_background_compactions=1 --max_bytes_for_level_base=67108864 --max_key=25000000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=0 --memtable_prefix_bloom_size_ratio=0.01 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --min_write_buffer_number_to_merge=2 --mmap_read=0 --mock_direct_io=True --nooverwritepercent=0 --num_file_reads_for_auto_readahead=2 --open_files=-1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=0 --open_write_fault_one_in=0 --ops_per_thread=40000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=3 --pause_background_one_in=1000000 --periodic_compaction_seconds=0 --prefix_size=7 --prefixpercent=5 --prepopulate_block_cache=0 --preserve_internal_time_seconds=3600 --progress_reports=0 --read_fault_one_in=1000 --readahead_size=0 --readpercent=65 --recycle_log_file_num=1 --reopen=0 --ribbon_starting_level=999 --secondary_cache_fault_one_in=0 --snapshot_hold_ops=100000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --stats_dump_period_sec=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=16777216 --target_file_size_multiplier=1 --test_batches_snapshots=0 --top_level_index_pinning=1 --unpartitioned_pinning=1 --use_direct_io_for_flush_and_compaction=1 --use_direct_reads=1 --use_full_merge_v1=1 --use_merge=0 --use_multiget=0 --use_put_entity_one_in=0 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=1000000 --verify_db_one_in=100000 --verify_iterator_with_expected_state_one_in=0 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=none --write_buffer_size=33554432 --write_dbid_to_manifest=1 --writepercent=20 ``` is gone after this fix - CI Reviewed By: ajkr Differential Revision: D41319441 Pulled By: hx235 fbshipit-source-id: 6939753767007f7449ea7055b1420aabd03d7709
2 years ago
ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: picking file %" PRIu64
" with size %s for deletion under total size %" PRIu64
" vs max table files size %" PRIu64,
cf_name.c_str(), f->fd.GetNumber(), tmp_fsize, total_size,
mutable_cf_options.compaction_options_fifo.max_table_files_size);
if (total_size <=
mutable_cf_options.compaction_options_fifo.max_table_files_size) {
break;
}
}
Trigger FIFO file deletion in non L0 only if exceeding max_table_files_size (#10955) Summary: **Context** https://github.com/facebook/rocksdb/pull/10348 allows multi-level FIFO but accidentally made change to the logic of deleting files in `FIFOCompactionPicker::PickSizeCompaction`. With [this](https://github.com/facebook/rocksdb/pull/10348/files#diff-d8fb3d50749aa69b378de447e3d9cf2f48abe0281437f010b5d61365a7b813fdR156) and [this](https://github.com/facebook/rocksdb/pull/10348/files#diff-d8fb3d50749aa69b378de447e3d9cf2f48abe0281437f010b5d61365a7b813fdR235) together, it deletes one file in non-L0 even when `total_size <= mutable_cf_options.compaction_options_fifo.max_table_files_size`, which is incorrect. As a consequence, FIFO exercises more file deletion in our crash testing, which is not able to verify correctly on deleted keys in the file deleted by compaction. This results in errors `error : inconsistent values for key 000000000000239F000000000000012B000000000000028B: expected state has the key, Get() returns NotFound. Verification failed :(` or `Expected state has key 00000000000023A90000000000000003787878, iterator is at key 00000000000023A9000000000000004178 Column family: default, op_logs: S 00000000000023A90000000000000003787878` **Summary**: - Delete file for non-L0 only if `total_size <= mutable_cf_options.compaction_options_fifo.max_table_files_size` - Add some helpful log to LOG file Pull Request resolved: https://github.com/facebook/rocksdb/pull/10955 Test Plan: - Errors repro-ed by ``` ./db_stress --preserve_unverified_changes=1 --acquire_snapshot_one_in=10000 --adaptive_readahead=0 --allow_concurrent_memtable_write=0 --allow_data_in_errors=True --async_io=0 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=1 --backup_max_size=104857600 --backup_one_in=100000 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=10 --bottommost_compression_type=none --bytes_per_sync=0 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=1 --charge_file_metadata=1 --charge_filter_construction=0 --charge_table_reader=1 --checkpoint_one_in=1000000 --checksum_type=kxxHash --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=1000000 --compact_range_one_in=1000000 --compaction_pri=3 --compaction_style=2 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8589934591 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=xpress --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test/rocksdb_crashtest_whitebox --db_write_buffer_size=1048576 --delpercent=0 --delrangepercent=0 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --expected_values_dir=/dev/shm/rocksdb_test/rocksdb_crashtest_expected --fail_if_options_file_error=1 --fifo_allow_compaction=1 --file_checksum_impl=xxh64 --flush_one_in=1000000 --format_version=4 --get_current_wal_file_one_in=0 --get_live_files_one_in=1000000 --get_property_one_in=1000000 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=10 --index_type=2 --ingest_external_file_one_in=1000000 --initial_auto_readahead_size=16384 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=False --log2_keys_per_lock=10 --long_running_snapshots=0 --manual_wal_flush_one_in=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=524288 --max_background_compactions=1 --max_bytes_for_level_base=67108864 --max_key=25000000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=0 --memtable_prefix_bloom_size_ratio=0.01 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --min_write_buffer_number_to_merge=2 --mmap_read=0 --mock_direct_io=True --nooverwritepercent=0 --num_file_reads_for_auto_readahead=2 --open_files=-1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=0 --open_write_fault_one_in=0 --ops_per_thread=40000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=3 --pause_background_one_in=1000000 --periodic_compaction_seconds=0 --prefix_size=7 --prefixpercent=5 --prepopulate_block_cache=0 --preserve_internal_time_seconds=3600 --progress_reports=0 --read_fault_one_in=1000 --readahead_size=0 --readpercent=65 --recycle_log_file_num=1 --reopen=0 --ribbon_starting_level=999 --secondary_cache_fault_one_in=0 --snapshot_hold_ops=100000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --stats_dump_period_sec=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=16777216 --target_file_size_multiplier=1 --test_batches_snapshots=0 --top_level_index_pinning=1 --unpartitioned_pinning=1 --use_direct_io_for_flush_and_compaction=1 --use_direct_reads=1 --use_full_merge_v1=1 --use_merge=0 --use_multiget=0 --use_put_entity_one_in=0 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=1000000 --verify_db_one_in=100000 --verify_iterator_with_expected_state_one_in=0 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=none --write_buffer_size=33554432 --write_dbid_to_manifest=1 --writepercent=20 ``` is gone after this fix - CI Reviewed By: ajkr Differential Revision: D41319441 Pulled By: hx235 fbshipit-source-id: 6939753767007f7449ea7055b1420aabd03d7709
2 years ago
} else {
ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: nothing to do. Total size %" PRIu64
", max size %" PRIu64 "\n",
cf_name.c_str(), total_size,
mutable_cf_options.compaction_options_fifo.max_table_files_size);
return nullptr;
}
Compaction* c = new Compaction(
vstorage, ioptions_, mutable_cf_options, mutable_db_options,
std::move(inputs), last_level,
/* target_file_size */ 0,
/* max_compaction_bytes */ 0,
/* output_path_id */ 0, kNoCompression,
mutable_cf_options.compression_opts, Temperature::kUnknown,
/* max_subcompactions */ 0, {}, /* is manual */ false,
/* trim_ts */ "", vstorage->CompactionScore(0),
/* is deletion compaction */ true,
/* l0_files_might_overlap */ true, CompactionReason::kFIFOMaxSize);
return c;
}
Compaction* FIFOCompactionPicker::PickCompactionToWarm(
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
LogBuffer* log_buffer, const SequenceNumber earliest_mem_seqno) {
TEST_SYNC_POINT("PickCompactionToWarm");
if (mutable_cf_options.compaction_options_fifo.age_for_warm == 0) {
return nullptr;
}
// PickCompactionToWarm is only triggered if there is no non-L0 files.
for (int level = 1; level < vstorage->num_levels(); ++level) {
if (GetTotalFilesSize(vstorage->LevelFiles(level)) > 0) {
return nullptr;
}
}
const int kLevel0 = 0;
const std::vector<FileMetaData*>& level_files = vstorage->LevelFiles(kLevel0);
int64_t _current_time;
auto status = ioptions_.clock->GetCurrentTime(&_current_time);
if (!status.ok()) {
ROCKS_LOG_BUFFER(log_buffer,
"[%s] FIFO compaction: Couldn't get current time: %s. "
"Not doing compactions based on warm threshold. ",
cf_name.c_str(), status.ToString().c_str());
return nullptr;
}
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
TEST_SYNC_POINT_CALLBACK("PickCompactionToWarm::BeforeGetCurrentTime",
&_current_time);
const uint64_t current_time = static_cast<uint64_t>(_current_time);
if (!level0_compactions_in_progress_.empty()) {
ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: Already executing compaction. Parallel "
"compactions are not supported",
cf_name.c_str());
return nullptr;
}
std::vector<CompactionInputFiles> inputs;
inputs.emplace_back();
inputs[0].level = 0;
// avoid underflow
if (current_time > mutable_cf_options.compaction_options_fifo.age_for_warm) {
uint64_t create_time_threshold =
current_time - mutable_cf_options.compaction_options_fifo.age_for_warm;
uint64_t compaction_size = 0;
// We will ideally identify a file qualifying for warm tier by knowing
// the timestamp for the youngest entry in the file. However, right now
// we don't have the information. We infer it by looking at timestamp
// of the next file's (which is just younger) oldest entry's timestamp.
FileMetaData* prev_file = nullptr;
for (auto ritr = level_files.rbegin(); ritr != level_files.rend(); ++ritr) {
FileMetaData* f = *ritr;
assert(f);
if (f->being_compacted) {
// Right now this probably won't happen as we never try to schedule
// two compactions in parallel, so here we just simply don't schedule
// anything.
return nullptr;
}
uint64_t oldest_ancester_time = f->TryGetOldestAncesterTime();
if (oldest_ancester_time == kUnknownOldestAncesterTime) {
// Older files might not have enough information. It is possible to
// handle these files by looking at newer files, but maintaining the
// logic isn't worth it.
break;
}
if (oldest_ancester_time > create_time_threshold) {
// The previous file (which has slightly older data) doesn't qualify
// for warm tier.
break;
}
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
if (prev_file != nullptr &&
prev_file->fd.largest_seqno <= earliest_mem_seqno) {
compaction_size += prev_file->fd.GetFileSize();
if (compaction_size > mutable_cf_options.max_compaction_bytes) {
break;
}
inputs[0].files.push_back(prev_file);
ROCKS_LOG_BUFFER(log_buffer,
"[%s] FIFO compaction: picking file %" PRIu64
" with next file's oldest time %" PRIu64 " for warm",
cf_name.c_str(), prev_file->fd.GetNumber(),
oldest_ancester_time);
}
if (f->temperature == Temperature::kUnknown ||
f->temperature == Temperature::kHot) {
prev_file = f;
} else if (!inputs[0].files.empty()) {
// A warm file newer than files picked.
break;
} else {
assert(prev_file == nullptr);
}
}
}
if (inputs[0].files.empty()) {
return nullptr;
}
Compaction* c = new Compaction(
vstorage, ioptions_, mutable_cf_options, mutable_db_options,
std::move(inputs), 0, 0 /* output file size limit */,
0 /* max compaction bytes, not applicable */, 0 /* output path ID */,
mutable_cf_options.compression, mutable_cf_options.compression_opts,
Temperature::kWarm,
/* max_subcompactions */ 0, {}, /* is manual */ false, /* trim_ts */ "",
vstorage->CompactionScore(0),
/* is deletion compaction */ false, /* l0_files_might_overlap */ true,
CompactionReason::kChangeTemperature);
return c;
}
Compaction* FIFOCompactionPicker::PickCompaction(
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
LogBuffer* log_buffer, const SequenceNumber earliest_mem_seqno) {
Compaction* c = nullptr;
if (mutable_cf_options.ttl > 0) {
c = PickTTLCompaction(cf_name, mutable_cf_options, mutable_db_options,
vstorage, log_buffer);
}
if (c == nullptr) {
c = PickSizeCompaction(cf_name, mutable_cf_options, mutable_db_options,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
vstorage, log_buffer, earliest_mem_seqno);
}
if (c == nullptr) {
c = PickCompactionToWarm(cf_name, mutable_cf_options, mutable_db_options,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
vstorage, log_buffer, earliest_mem_seqno);
}
RegisterCompaction(c);
return c;
}
Compaction* FIFOCompactionPicker::CompactRange(
const std::string& cf_name, const MutableCFOptions& mutable_cf_options,
const MutableDBOptions& mutable_db_options, VersionStorageInfo* vstorage,
int input_level, int output_level,
const CompactRangeOptions& /*compact_range_options*/,
const InternalKey* /*begin*/, const InternalKey* /*end*/,
InternalKey** compaction_end, bool* /*manual_conflict*/,
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
uint64_t /*max_file_num_to_ignore*/, const std::string& /*trim_ts*/,
const SequenceNumber earliest_mem_seqno) {
#ifdef NDEBUG
(void)input_level;
(void)output_level;
#endif
assert(input_level == 0);
assert(output_level == 0);
*compaction_end = nullptr;
LogBuffer log_buffer(InfoLogLevel::INFO_LEVEL, ioptions_.logger);
Fix FIFO causing overlapping seqnos in L0 files due to overlapped seqnos between ingested files and memtable's (#10777) Summary: **Context:** Same as https://github.com/facebook/rocksdb/pull/5958#issue-511150930 but apply the fix to FIFO Compaction case Repro: ``` COERCE_CONTEXT_SWICH=1 make -j56 db_stress ./db_stress --acquire_snapshot_one_in=0 --adaptive_readahead=0 --allow_data_in_errors=True --async_io=1 --avoid_flush_during_recovery=0 --avoid_unnecessary_blocking_io=0 --backup_max_size=104857600 --backup_one_in=0 --batch_protection_bytes_per_key=0 --block_size=16384 --bloom_bits=18 --bottommost_compression_type=disable --bytes_per_sync=262144 --cache_index_and_filter_blocks=0 --cache_size=8388608 --cache_type=lru_cache --charge_compression_dictionary_building_buffer=0 --charge_file_metadata=1 --charge_filter_construction=1 --charge_table_reader=1 --checkpoint_one_in=0 --checksum_type=kCRC32c --clear_column_family_one_in=0 --column_families=1 --compact_files_one_in=0 --compact_range_one_in=1000 --compaction_pri=3 --open_files=-1 --compaction_style=2 --fifo_allow_compaction=1 --compaction_ttl=0 --compression_max_dict_buffer_bytes=8388607 --compression_max_dict_bytes=16384 --compression_parallel_threads=1 --compression_type=zlib --compression_use_zstd_dict_trainer=1 --compression_zstd_max_train_bytes=0 --continuous_verification_interval=0 --data_block_index_type=0 --db=/dev/shm/rocksdb_test0/rocksdb_crashtest_whitebox --db_write_buffer_size=8388608 --delpercent=4 --delrangepercent=1 --destroy_db_initially=1 --detect_filter_construct_corruption=0 --disable_wal=0 --enable_compaction_filter=0 --enable_pipelined_write=1 --fail_if_options_file_error=1 --file_checksum_impl=none --flush_one_in=1000 --format_version=5 --get_current_wal_file_one_in=0 --get_live_files_one_in=0 --get_property_one_in=0 --get_sorted_wal_files_one_in=0 --index_block_restart_interval=15 --index_type=3 --ingest_external_file_one_in=100 --initial_auto_readahead_size=0 --iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True --log2_keys_per_lock=10 --long_running_snapshots=0 --mark_for_compaction_one_file_in=10 --max_auto_readahead_size=16384 --max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000 --max_key_len=3 --max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=1048576 --max_write_buffer_number=3 --max_write_buffer_size_to_maintain=4194304 --memtable_prefix_bloom_size_ratio=0.5 --memtable_protection_bytes_per_key=1 --memtable_whole_key_filtering=1 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False --nooverwritepercent=1 --num_file_reads_for_auto_readahead=0 --num_levels=1 --open_metadata_write_fault_one_in=0 --open_read_fault_one_in=32 --open_write_fault_one_in=0 --ops_per_thread=200000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=0 --partition_pinning=1 --pause_background_one_in=0 --periodic_compaction_seconds=0 --prefix_size=8 --prefixpercent=5 --prepopulate_block_cache=0 --progress_reports=0 --read_fault_one_in=0 --readahead_size=16384 --readpercent=45 --recycle_log_file_num=1 --reopen=20 --ribbon_starting_level=999 --snapshot_hold_ops=1000 --sst_file_manager_bytes_per_sec=0 --sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=0 --target_file_size_base=524288 --target_file_size_multiplier=2 --test_batches_snapshots=0 --top_level_index_pinning=3 --unpartitioned_pinning=0 --use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=1 --use_merge=0 --use_multiget=1 --user_timestamp_size=0 --value_size_mult=32 --verify_checksum=1 --verify_checksum_one_in=0 --verify_db_one_in=1000 --verify_sst_unique_id_in_manifest=1 --wal_bytes_per_sync=0 --wal_compression=zstd --write_buffer_size=524288 --write_dbid_to_manifest=0 --writepercent=35 put or merge error: Corruption: force_consistency_checks(DEBUG): VersionBuilder: L0 file https://github.com/facebook/rocksdb/issues/479 with seqno 23711 29070 vs. file https://github.com/facebook/rocksdb/issues/482 with seqno 27138 29049 ``` **Summary:** FIFO only does intra-L0 compaction in the following four cases. For other cases, FIFO drops data instead of compacting on data, which is irrelevant to the overlapping seqno issue we are solving. - [FIFOCompactionPicker::PickSizeCompaction](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L155) when `total size < compaction_options_fifo.max_table_files_size` and `compaction_options_fifo.allow_compaction == true` - For this path, we simply reuse the fix in `FindIntraL0Compaction` https://github.com/facebook/rocksdb/pull/5958/files#diff-c261f77d6dd2134333c4a955c311cf4a196a08d3c2bb6ce24fd6801407877c89R56 - This path was not stress-tested at all. Therefore we covered `fifo.allow_compaction` in stress test to surface the overlapping seqno issue we are fixing here. - [FIFOCompactionPicker::PickCompactionToWarm](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L313) when `compaction_options_fifo.age_for_warm > 0` - For this path, we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 and skip files of largest seqno greater than `earliest_mem_seqno` - This path was not stress-tested at all. However covering `age_for_warm` option worths a separate PR to deal with db stress compatibility. Therefore we manually tested this path for this PR - [FIFOCompactionPicker::CompactRange](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker_fifo.cc#L365) that ends up picking one of the above two compactions - [CompactionPicker::CompactFiles](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.cc#L378) - Since `SanitizeCompactionInputFiles()` will be called [before](https://github.com/facebook/rocksdb/blob/7.6.fb/db/compaction/compaction_picker.h#L111-L113) `CompactionPicker::CompactFiles` , we simply replicate the idea in https://github.com/facebook/rocksdb/pull/5958#issue-511150930 in `SanitizeCompactionInputFiles()`. To simplify implementation, we return `Stats::Abort()` on encountering seqno-overlapped file when doing compaction to L0 instead of skipping the file and proceed with the compaction. Some additional clean-up included in this PR: - Renamed `earliest_memtable_seqno` to `earliest_mem_seqno` for consistent naming - Added comment about `earliest_memtable_seqno` in related APIs - Made parameter `earliest_memtable_seqno` constant and required Pull Request resolved: https://github.com/facebook/rocksdb/pull/10777 Test Plan: - make check - New unit test `TEST_P(DBCompactionTestFIFOCheckConsistencyWithParam, FlushAfterIntraL0CompactionWithIngestedFile)`corresponding to the above 4 cases, which will fail accordingly without the fix - Regular CI stress run on this PR + stress test with aggressive value https://github.com/facebook/rocksdb/pull/10761 and on FIFO compaction only Reviewed By: ajkr Differential Revision: D40090485 Pulled By: hx235 fbshipit-source-id: 52624186952ee7109117788741aeeac86b624a4f
2 years ago
Compaction* c =
PickCompaction(cf_name, mutable_cf_options, mutable_db_options, vstorage,
&log_buffer, earliest_mem_seqno);
log_buffer.FlushBufferToLog();
return c;
}
} // namespace ROCKSDB_NAMESPACE
#endif // !ROCKSDB_LITE