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
main
Hui Xiao 2 years ago committed by Facebook GitHub Bot
parent ed23fd7591
commit d8c043f7ad
  1. 1
      HISTORY.md
  2. 21
      db/compaction/compaction_picker_fifo.cc

@ -18,6 +18,7 @@
* Fixed a memory safety bug when using a SecondaryCache with `block_cache_compressed`. `block_cache_compressed` no longer attempts to use SecondaryCache features. * Fixed a memory safety bug when using a SecondaryCache with `block_cache_compressed`. `block_cache_compressed` no longer attempts to use SecondaryCache features.
* Fixed a regression in scan for async_io. During seek, valid buffers were getting cleared causing a regression. * Fixed a regression in scan for async_io. During seek, valid buffers were getting cleared causing a regression.
* Tiered Storage: fixed excessive keys written to penultimate level in non-debug builds. * Tiered Storage: fixed excessive keys written to penultimate level in non-debug builds.
* Fixed a bug that multi-level FIFO compaction deletes one file in non-L0 even when `CompactionOptionsFIFO::max_table_files_size` is no exceeded since #10348 or 7.8.0.
### New Features ### New Features
* Add basic support for user-defined timestamp to Merge (#10819). * Add basic support for user-defined timestamp to Merge (#10819).

@ -233,7 +233,8 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(
break; break;
} }
} }
} else { } 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 // 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 // logically the oldest since the file creation time only represents
// when this file was compacted to this level, which is independent // when this file was compacted to this level, which is independent
@ -248,15 +249,27 @@ Compaction* FIFOCompactionPicker::PickSizeCompaction(
inputs[0].files.push_back(f); inputs[0].files.push_back(f);
char tmp_fsize[16]; char tmp_fsize[16];
AppendHumanBytes(f->fd.GetFileSize(), tmp_fsize, sizeof(tmp_fsize)); AppendHumanBytes(f->fd.GetFileSize(), tmp_fsize, sizeof(tmp_fsize));
ROCKS_LOG_BUFFER(log_buffer, ROCKS_LOG_BUFFER(
log_buffer,
"[%s] FIFO compaction: picking file %" PRIu64 "[%s] FIFO compaction: picking file %" PRIu64
" with size %s for deletion", " with size %s for deletion under total size %" PRIu64
cf_name.c_str(), f->fd.GetNumber(), tmp_fsize); " 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 <= if (total_size <=
mutable_cf_options.compaction_options_fifo.max_table_files_size) { mutable_cf_options.compaction_options_fifo.max_table_files_size) {
break; break;
} }
} }
} 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( Compaction* c = new Compaction(

Loading…
Cancel
Save