rocksdb/table/block_based
Peter Dillinger 249eff0f30 Stats for redundant insertions into block cache (#6681)
Summary:
Since read threads do not coordinate on loading data into block
cache, two threads between Lookup and Insert can end up loading and
inserting the same data. This is particularly concerning with
cache_index_and_filter_blocks since those are hot and more likely to
be race targets if ejected from (or not pre-populated in) the cache.

Particularly with moves toward disaggregated / network storage, the cost
of redundant retrieval might be high, and we should at least have some
hard statistics from which we can estimate impact.

Example with full filter thrashing "cliff":

    $ ./db_bench --benchmarks=fillrandom --num=15000000 --cache_index_and_filter_blocks -bloom_bits=10
    ...
    $ ./db_bench --db=/tmp/rocksdbtest-172704/dbbench --use_existing_db --benchmarks=readrandom,stats --num=200000 --cache_index_and_filter_blocks --cache_size=$((130 * 1024 * 1024)) --bloom_bits=10 --threads=16 -statistics 2>&1 | egrep '^rocksdb.block.cache.(.*add|.*redundant)' | grep -v compress | sort
    rocksdb.block.cache.add COUNT : 14181
    rocksdb.block.cache.add.failures COUNT : 0
    rocksdb.block.cache.add.redundant COUNT : 476
    rocksdb.block.cache.data.add COUNT : 12749
    rocksdb.block.cache.data.add.redundant COUNT : 18
    rocksdb.block.cache.filter.add COUNT : 1003
    rocksdb.block.cache.filter.add.redundant COUNT : 217
    rocksdb.block.cache.index.add COUNT : 429
    rocksdb.block.cache.index.add.redundant COUNT : 241
    $ ./db_bench --db=/tmp/rocksdbtest-172704/dbbench --use_existing_db --benchmarks=readrandom,stats --num=200000 --cache_index_and_filter_blocks --cache_size=$((120 * 1024 * 1024)) --bloom_bits=10 --threads=16 -statistics 2>&1 | egrep '^rocksdb.block.cache.(.*add|.*redundant)' | grep -v compress | sort
    rocksdb.block.cache.add COUNT : 1182223
    rocksdb.block.cache.add.failures COUNT : 0
    rocksdb.block.cache.add.redundant COUNT : 302728
    rocksdb.block.cache.data.add COUNT : 31425
    rocksdb.block.cache.data.add.redundant COUNT : 12
    rocksdb.block.cache.filter.add COUNT : 795455
    rocksdb.block.cache.filter.add.redundant COUNT : 130238
    rocksdb.block.cache.index.add COUNT : 355343
    rocksdb.block.cache.index.add.redundant COUNT : 172478
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6681

Test Plan: Some manual testing (above) and unit test covering key metrics is included

Reviewed By: ltamasi

Differential Revision: D21134113

Pulled By: pdillinger

fbshipit-source-id: c11497b5f00f4ffdfe919823904e52d0a1a91d87
2020-04-27 13:20:27 -07:00
..
binary_search_index_reader.cc Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
binary_search_index_reader.h Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
block.cc Fix some typos in code comments (#6733) 2020-04-23 12:28:49 -07:00
block.h fix comparison count for format_version=3 indexes (#6650) 2020-04-13 11:18:37 -07:00
block_based_filter_block.cc Silence false alarms in db_stress fault injection (#6741) 2020-04-24 13:06:12 -07:00
block_based_filter_block.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
block_based_filter_block_test.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
block_based_table_builder.cc C++20 compatibility (#6697) 2020-04-20 13:24:25 -07:00
block_based_table_builder.h In block based table builder, make variables for estimating file size atomic (#6636) 2020-04-02 16:16:24 -07:00
block_based_table_factory.cc Add a ConfigOptions for use in comparing objects and converting to/from strings (#6389) 2020-04-21 17:38:17 -07:00
block_based_table_factory.h Add a ConfigOptions for use in comparing objects and converting to/from strings (#6389) 2020-04-21 17:38:17 -07:00
block_based_table_iterator.cc Properly report IO errors when IndexType::kBinarySearchWithFirstKey is used (#6621) 2020-04-15 17:40:44 -07:00
block_based_table_iterator.h Properly report IO errors when IndexType::kBinarySearchWithFirstKey is used (#6621) 2020-04-15 17:40:44 -07:00
block_based_table_reader.cc Stats for redundant insertions into block cache (#6681) 2020-04-27 13:20:27 -07:00
block_based_table_reader.h Stats for redundant insertions into block cache (#6681) 2020-04-27 13:20:27 -07:00
block_based_table_reader_impl.h Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
block_builder.cc Add pipelined & parallel compression optimization (#6262) 2020-04-01 16:40:18 -07:00
block_builder.h Add pipelined & parallel compression optimization (#6262) 2020-04-01 16:40:18 -07:00
block_prefetcher.cc De-template block based table iterator (#6531) 2020-03-16 12:20:50 -07:00
block_prefetcher.h De-template block based table iterator (#6531) 2020-03-16 12:20:50 -07:00
block_prefix_index.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
block_prefix_index.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
block_test.cc Fix range deletion tombstone ingestion with global seqno (#6429) 2020-02-25 15:31:48 -08:00
block_type.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
cachable_entry.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
data_block_footer.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
data_block_footer.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
data_block_hash_index.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
data_block_hash_index.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
data_block_hash_index_test.cc Fix range deletion tombstone ingestion with global seqno (#6429) 2020-02-25 15:31:48 -08:00
filter_block.h Fix iterator reading filter block despite read_tier == kBlockCacheTier (#6562) 2020-03-26 15:21:26 -07:00
filter_block_reader_common.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
filter_block_reader_common.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
filter_policy.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
filter_policy_internal.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
flush_block_policy.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
flush_block_policy.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
full_filter_block.cc Silence false alarms in db_stress fault injection (#6741) 2020-04-24 13:06:12 -07:00
full_filter_block.h Fix iterator reading filter block despite read_tier == kBlockCacheTier (#6562) 2020-03-26 15:21:26 -07:00
full_filter_block_test.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
hash_index_reader.cc Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
hash_index_reader.h Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
index_builder.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
index_builder.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
index_reader_common.cc Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
index_reader_common.h Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
mock_block_based_table.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
parsed_full_filter_block.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
parsed_full_filter_block.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
partitioned_filter_block.cc Silence false alarms in db_stress fault injection (#6741) 2020-04-24 13:06:12 -07:00
partitioned_filter_block.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
partitioned_filter_block_test.cc Fix range deletion tombstone ingestion with global seqno (#6429) 2020-02-25 15:31:48 -08:00
partitioned_index_iterator.cc Fix regression bug in partitioned index reseek caused by #6531 (#6551) 2020-03-17 12:33:10 -07:00
partitioned_index_iterator.h Fix compiler warning treated as error (#6547) 2020-03-17 09:59:28 -07:00
partitioned_index_reader.cc Silence false alarms in db_stress fault injection (#6741) 2020-04-24 13:06:12 -07:00
partitioned_index_reader.h Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
reader_common.cc Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
reader_common.h Divide block_based_table_reader.cc (#6527) 2020-03-12 21:41:50 -07:00
uncompression_dict_reader.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
uncompression_dict_reader.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00