rocksdb/tools
Peter Dillinger 5b2bbacb6f Minimize memory internal fragmentation for Bloom filters (#6427)
Summary:
New experimental option BBTO::optimize_filters_for_memory builds
filters that maximize their use of "usable size" from malloc_usable_size,
which is also used to compute block cache charges.

Rather than always "rounding up," we track state in the
BloomFilterPolicy object to mix essentially "rounding down" and
"rounding up" so that the average FP rate of all generated filters is
the same as without the option. (YMMV as heavily accessed filters might
be unluckily lower accuracy.)

Thus, the option near-minimizes what the block cache considers as
"memory used" for a given target Bloom filter false positive rate and
Bloom filter implementation. There are no forward or backward
compatibility issues with this change, though it only works on the
format_version=5 Bloom filter.

With Jemalloc, we see about 10% reduction in memory footprint (and block
cache charge) for Bloom filters, but 1-2% increase in storage footprint,
due to encoding efficiency losses (FP rate is non-linear with bits/key).

Why not weighted random round up/down rather than state tracking? By
only requiring malloc_usable_size, we don't actually know what the next
larger and next smaller usable sizes for the allocator are. We pick a
requested size, accept and use whatever usable size it has, and use the
difference to inform our next choice. This allows us to narrow in on the
right balance without tracking/predicting usable sizes.

Why not weight history of generated filter false positive rates by
number of keys? This could lead to excess skew in small filters after
generating a large filter.

Results from filter_bench with jemalloc (irrelevant details omitted):

    (normal keys/filter, but high variance)
    $ ./filter_bench -quick -impl=2 -average_keys_per_filter=30000 -vary_key_count_ratio=0.9
    Build avg ns/key: 29.6278
    Number of filters: 5516
    Total size (MB): 200.046
    Reported total allocated memory (MB): 220.597
    Reported internal fragmentation: 10.2732%
    Bits/key stored: 10.0097
    Average FP rate %: 0.965228
    $ ./filter_bench -quick -impl=2 -average_keys_per_filter=30000 -vary_key_count_ratio=0.9 -optimize_filters_for_memory
    Build avg ns/key: 30.5104
    Number of filters: 5464
    Total size (MB): 200.015
    Reported total allocated memory (MB): 200.322
    Reported internal fragmentation: 0.153709%
    Bits/key stored: 10.1011
    Average FP rate %: 0.966313

    (very few keys / filter, optimization not as effective due to ~59 byte
     internal fragmentation in blocked Bloom filter representation)
    $ ./filter_bench -quick -impl=2 -average_keys_per_filter=1000 -vary_key_count_ratio=0.9
    Build avg ns/key: 29.5649
    Number of filters: 162950
    Total size (MB): 200.001
    Reported total allocated memory (MB): 224.624
    Reported internal fragmentation: 12.3117%
    Bits/key stored: 10.2951
    Average FP rate %: 0.821534
    $ ./filter_bench -quick -impl=2 -average_keys_per_filter=1000 -vary_key_count_ratio=0.9 -optimize_filters_for_memory
    Build avg ns/key: 31.8057
    Number of filters: 159849
    Total size (MB): 200
    Reported total allocated memory (MB): 208.846
    Reported internal fragmentation: 4.42297%
    Bits/key stored: 10.4948
    Average FP rate %: 0.811006

    (high keys/filter)
    $ ./filter_bench -quick -impl=2 -average_keys_per_filter=1000000 -vary_key_count_ratio=0.9
    Build avg ns/key: 29.7017
    Number of filters: 164
    Total size (MB): 200.352
    Reported total allocated memory (MB): 221.5
    Reported internal fragmentation: 10.5552%
    Bits/key stored: 10.0003
    Average FP rate %: 0.969358
    $ ./filter_bench -quick -impl=2 -average_keys_per_filter=1000000 -vary_key_count_ratio=0.9 -optimize_filters_for_memory
    Build avg ns/key: 30.7131
    Number of filters: 160
    Total size (MB): 200.928
    Reported total allocated memory (MB): 200.938
    Reported internal fragmentation: 0.00448054%
    Bits/key stored: 10.1852
    Average FP rate %: 0.963387

And from db_bench (block cache) with jemalloc:

    $ ./db_bench -db=/dev/shm/dbbench.no_optimize -benchmarks=fillrandom -format_version=5 -value_size=90 -bloom_bits=10 -num=2000000 -threads=8 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=false
    $ ./db_bench -db=/dev/shm/dbbench -benchmarks=fillrandom -format_version=5 -value_size=90 -bloom_bits=10 -num=2000000 -threads=8 -optimize_filters_for_memory -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=false
    $ (for FILE in /dev/shm/dbbench.no_optimize/*.sst; do ./sst_dump --file=$FILE --show_properties | grep 'filter block' ; done) | awk '{ t += $4; } END { print t; }'
    17063835
    $ (for FILE in /dev/shm/dbbench/*.sst; do ./sst_dump --file=$FILE --show_properties | grep 'filter block' ; done) | awk '{ t += $4; } END { print t; }'
    17430747
    $ #^ 2.1% additional filter storage
    $ ./db_bench -db=/dev/shm/dbbench.no_optimize -use_existing_db -benchmarks=readrandom,stats -statistics -bloom_bits=10 -num=2000000 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=false -duration=10 -cache_index_and_filter_blocks -cache_size=1000000000
    rocksdb.block.cache.index.add COUNT : 33
    rocksdb.block.cache.index.bytes.insert COUNT : 8440400
    rocksdb.block.cache.filter.add COUNT : 33
    rocksdb.block.cache.filter.bytes.insert COUNT : 21087528
    rocksdb.bloom.filter.useful COUNT : 4963889
    rocksdb.bloom.filter.full.positive COUNT : 1214081
    rocksdb.bloom.filter.full.true.positive COUNT : 1161999
    $ #^ 1.04 % observed FP rate
    $ ./db_bench -db=/dev/shm/dbbench -use_existing_db -benchmarks=readrandom,stats -statistics -bloom_bits=10 -num=2000000 -compaction_style=2 -fifo_compaction_max_table_files_size_mb=10000 -fifo_compaction_allow_compaction=false -optimize_filters_for_memory -duration=10 -cache_index_and_filter_blocks -cache_size=1000000000
    rocksdb.block.cache.index.add COUNT : 33
    rocksdb.block.cache.index.bytes.insert COUNT : 8448592
    rocksdb.block.cache.filter.add COUNT : 33
    rocksdb.block.cache.filter.bytes.insert COUNT : 18220328
    rocksdb.bloom.filter.useful COUNT : 5360933
    rocksdb.bloom.filter.full.positive COUNT : 1321315
    rocksdb.bloom.filter.full.true.positive COUNT : 1262999
    $ #^ 1.08 % observed FP rate, 13.6% less memory usage for filters

(Due to specific key density, this example tends to generate filters that are "worse than average" for internal fragmentation. "Better than average" cases can show little or no improvement.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6427

Test Plan: unit test added, 'make check' with gcc, clang and valgrind

Reviewed By: siying

Differential Revision: D22124374

Pulled By: pdillinger

fbshipit-source-id: f3e3aa152f9043ddf4fae25799e76341d0d8714e
2020-06-22 13:32:07 -07:00
..
advisor Rules Advisor: some fixes to support fetching stats from ODS (#4223) 2018-08-02 15:42:42 -07:00
block_cache_analyzer C++20 compatibility (#6697) 2020-04-20 13:24:25 -07:00
dump Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
rdb Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
analyze_txn_stress_test.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
auto_sanity_test.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
benchmark_leveldb.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
benchmark.sh Fixed typo in benchmark.sh (#6434) 2020-02-19 17:08:02 -08:00
blob_dump.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
check_all_python.py Allow missing "unversioned" python, as in CentOS 8 (#6883) 2020-05-29 11:29:23 -07:00
check_format_compatible.sh Two Improvements to tools/check_format_compatible.sh (#6702) 2020-04-15 11:28:11 -07:00
CMakeLists.txt Mark dependencies as PRIVATE and fix missing dependencies in tools. (#6790) 2020-05-12 21:07:55 -07:00
db_bench_tool_test.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
db_bench_tool.cc Minimize memory internal fragmentation for Bloom filters (#6427) 2020-06-22 13:32:07 -07:00
db_bench.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
db_crashtest.py Minimize memory internal fragmentation for Bloom filters (#6427) 2020-06-22 13:32:07 -07:00
db_repl_stress.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
db_sanity_test.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
dbench_monitor Fix /bin/bash shebangs 2017-08-03 15:56:46 -07:00
Dockerfile adding docker build script and dockerfile 2015-05-22 16:03:39 -07:00
generate_random_db.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
ingest_external_sst.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
ldb_cmd_impl.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
ldb_cmd_test.cc Add --version and --help to ldb and sst_dump (#6951) 2020-06-09 10:04:01 -07:00
ldb_cmd.cc Add --version and --help to ldb and sst_dump (#6951) 2020-06-09 10:04:01 -07:00
ldb_test.py Allow missing "unversioned" python, as in CentOS 8 (#6883) 2020-05-29 11:29:23 -07:00
ldb_tool.cc Add --version and --help to ldb and sst_dump (#6951) 2020-06-09 10:04:01 -07:00
ldb.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
pflag Fix /bin/bash shebangs 2017-08-03 15:56:46 -07:00
reduce_levels_test.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
regression_test.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
report_lite_binary_size.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
rocksdb_dump_test.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
run_flash_bench.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
run_leveldb.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
sample-dump.dmp First version of rocksdb_dump and rocksdb_undump. 2015-06-19 16:24:36 -07:00
sst_dump_test.cc Add --version and --help to ldb and sst_dump (#6951) 2020-06-09 10:04:01 -07:00
sst_dump_tool_imp.h sst_dump to reduce number of file reads (#6836) 2020-05-12 18:23:33 -07:00
sst_dump_tool.cc Add --version and --help to ldb and sst_dump (#6951) 2020-06-09 10:04:01 -07:00
sst_dump.cc Implement a new subcommand "identify" for sst_dump (#6943) 2020-06-08 13:58:28 -07:00
trace_analyzer_test.cc decouple the dependency of trace_analyzer_test unit test (#6941) 2020-06-08 10:36:43 -07:00
trace_analyzer_tool.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
trace_analyzer_tool.h Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
trace_analyzer.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00
verify_random_db.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
write_external_sst.sh Add copyright headers per FB open-source checkup tool. (#5199) 2019-04-18 10:55:01 -07:00
write_stress_runner.py Allow missing "unversioned" python, as in CentOS 8 (#6883) 2020-05-29 11:29:23 -07:00
write_stress.cc Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433) 2020-02-20 12:09:57 -08:00