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
Summary:
Based on https://github.com/facebook/rocksdb/issues/6648 (CLA Signed), but heavily modified / extended:
* Implicit capture of this via [=] deprecated in C++20, and [=,this] not standard before C++20 -> now using explicit capture lists
* Implicit copy operator deprecated in gcc 9 -> add explicit '= default' definition
* std::random_shuffle deprecated in C++17 and removed in C++20 -> migrated to a replacement in RocksDB random.h API
* Add the ability to build with different std version though -DCMAKE_CXX_STANDARD=11/14/17/20 on the cmake command line
* Minimal rebuild flag of MSVC is deprecated and is forbidden with /std:c++latest (C++20)
* Added MSVC 2019 C++11 & MSVC 2019 C++20 in AppVeyor
* Added GCC 9 C++11 & GCC9 C++20 in Travis
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6697
Test Plan: make check and CI
Reviewed By: cheng-chang
Differential Revision: D21020318
Pulled By: pdillinger
fbshipit-source-id: 12311be5dbd8675a0e2c817f7ec50fa11c18ab91
Summary:
Add NewFileChecksumGenCrc32cFactory to file checksum public interface such that applications can use the build in crc32 checksum factory.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6688
Test Plan: pass make asan_check
Reviewed By: riversand963
Differential Revision: D21006859
Pulled By: zhichao-cao
fbshipit-source-id: ea8a45196a8b77c310728ab05f6cc0f49f3baef0
Summary:
Adding a simple timer support to schedule work at a fixed time.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6543
Test Plan: TODO: clean up the unit tests, and make them better.
Reviewed By: siying
Differential Revision: D20465390
Pulled By: sagar0
fbshipit-source-id: cba143f70b6339863e1d0f8b8bf92e51c2b3d678
Summary:
I suspect LRUCache could use some optimization, and to support
such an effort, a good benchmarking tool is needed. The existing
cache_bench was heavily skewed toward insertion and lookup misses, and
did not saturate memory with other work. This change should improve
those things to better resemble a real workload.
(All below using clang compiler, for some consistency, but not
necessarily same version and settings.)
The real workload is from production MySQL on RocksDB, filtering stacks
containing "LRU", "ShardedCache" or "CacheShard."
Lookup inclusive: 66%
Insert inclusive: 17%
Release inclusive: 15%
An alternate simulated workload is MySQL running a LinkBench read test:
Lookup inclusive: 54%
Insert inclusive: 24%
Release inclusive: 21%
cache_bench default settings, prior to this change:
Lookup inclusive: 35.8%
Insert inclusive: 63.6%
Release inclusive: 0%
cache_bench after this change (intended as somewhat "tighter" workload
than average production, more like LinkBench):
Lookup inclusive: 52%
Insert inclusive: 20%
Release inclusive: 26%
And top exclusive stacks (portion of stack samples as filtered above):
Production MySQL:
LRUHandleTable::FindPointer: 25.3%
rocksdb::operator==: 15.1% <-- Slice ==
LRUCacheShard::LRU_Remove: 13.8%
ShardedCache::Lookup: 8.9%
__pthread_mutex_lock: 7.1%
LRUCacheShard::LRU_Insert: 6.3%
MurmurHash64A: 4.8% <-- Since upgraded to XXH3p
...
Old cache_bench:
LRUHandleTable::FindPointer: 23.6%
__pthread_mutex_lock: 15.0%
__pthread_mutex_unlock_usercnt: 11.7%
__lll_lock_wait: 8.6%
__lll_unlock_wake: 6.8%
LRUCacheShard::LRU_Insert: 6.0%
ShardedCache::Lookup: 4.4%
LRUCacheShard::LRU_Remove: 2.8%
...
rocksdb::operator==: 0.2% <-- Slice ==
...
New cache_bench:
LRUHandleTable::FindPointer: 22.8%
__pthread_mutex_unlock_usercnt: 14.3%
rocksdb::operator==: 10.5% <-- Slice ==
LRUCacheShard::LRU_Insert: 9.0%
__pthread_mutex_lock: 5.9%
LRUCacheShard::LRU_Remove: 5.0%
...
ShardedCache::Lookup: 2.9%
...
So there's a bit more lock contention in the benchmark than in
production, but otherwise looks similar enough to me. At least it's a
big improvement over the existing code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6629
Test Plan: No production code changes, ran cache_bench with ASAN
Reviewed By: ltamasi
Differential Revision: D20824318
Pulled By: pdillinger
fbshipit-source-id: 6f8dc5891ead0f87edbed3a615ecd5289d9abe12
Summary:
This PR adds support for pipelined & parallel compression optimization for `BlockBasedTableBuilder`. This optimization makes block building, block compression and block appending a pipeline, and uses multiple threads to accelerate block compression. Users can set `CompressionOptions::parallel_threads` greater than 1 to enable compression parallelism.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6262
Reviewed By: ajkr
Differential Revision: D20651306
fbshipit-source-id: 62125590a9c15b6d9071def9dc72589c1696a4cb
Summary:
In the current implementation, sst file checksum is calculated by a shared checksum function object, which may make some checksum function hard to be applied here such as SHA1. In this implementation, each sst file will have its own checksum generator obejct, created by FileChecksumGenFactory. User needs to implement its own FilechecksumGenerator and Factory to plugin the in checksum calculation method.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6600
Test Plan: tested with make asan_check
Reviewed By: riversand963
Differential Revision: D20717670
Pulled By: zhichao-cao
fbshipit-source-id: 2a74c1c280ac11a07a1980185b43b671acaa71c6
Summary:
When creating a database backup, the background threads will not only consume IO resources by copying files, but also consuming CPU such as by computing checksums. During peak times, the CPU consumption by the background threads might affect online queries.
This PR makes it possible to decrease CPU priority of these threads when creating a new backup.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6602
Test Plan: make check
Reviewed By: siying, zhichao-cao
Differential Revision: D20683216
Pulled By: cheng-chang
fbshipit-source-id: 9978b9ed9488e8ce135e90ca083e5b4b7221fd84
Summary:
In automatic compaction, if a compaction is bottommost, it goes to bottom thread pool. We should do the same for manual compaction too.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6593
Test Plan: Add a unit test. See all existing tests pass.
Reviewed By: ajkr
Differential Revision: D20637408
fbshipit-source-id: cb03031e8f895085f7acf6d2d65e69e84c9ddef3
Summary:
When building RocksDB on VS2015, an error shows up with
hash_map.h(39): error C2719: 'value': formal parameter with requested alignment of 8 won't be aligned
Making the reference a reference can solve the problem, and there isn't a reason we can't do that, at least for the current use of the hash map.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6567
Test Plan: See CI tests pass.
Reviewed By: pdillinger
Differential Revision: D20548543
fbshipit-source-id: 255b55d74cf68a0b324e6f504c56608a97ea6276
Summary:
There was an alignment bug in our copy of the streaming APIs
for XXH3 (which we dubbed "XXH3p" for "preview" release). Since those
APIs are unused and some values for XXH3 have changed since XXH3p, I'm
simply removing those APIs, expecting it's better to use finalized XXH3
function if/when we decide to use those APIs (e.g. for checksums).
Fixes https://github.com/facebook/rocksdb/issues/6508
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6540
Test Plan: make check
Differential Revision: D20479271
Pulled By: pdillinger
fbshipit-source-id: 246cf1690d614d3b31042b563d249de32dec1e0d
Summary:
fix a few build warnings that are treated as failures with more strict MSVC warning settings
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6517
Differential Revision: D20401325
Pulled By: pdillinger
fbshipit-source-id: b44979dfaafdc7b3b8cb44a565400a99b331dd30
Summary:
When users fail to open a DB with file lock failure, it is sometimes hard for users to debug. We now include the time the lock is acquired and the thread ID that acquired the lock, to help users debug problems like this. Default Env's thread ID is used.
Since type of lockedFiles is changed, rename it to follow naming convention too.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6507
Test Plan: Add a unit test and improve an existing test to validate the case.
Differential Revision: D20378333
fbshipit-source-id: 312fe0e9733fd1d1e9969c321b90ce523cf4708a
Summary:
Preliminary support for iterator with user timestamp. Current implementation does not consider merge operator and reverse iterator. Auto compaction is also disabled in unit tests.
Create an iterator with timestamp.
```
...
read_opts.timestamp = &ts;
auto* iter = db->NewIterator(read_opts);
// target is key without timestamp.
for (iter->Seek(target); iter->Valid(); iter->Next()) {}
for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {}
delete iter;
read_opts.timestamp = &ts1;
// lower_bound and upper_bound are without timestamp.
read_opts.iterate_lower_bound = &lower_bound;
read_opts.iterate_upper_bound = &upper_bound;
auto* iter1 = db->NewIterator(read_opts);
// Do Seek or SeekToFirst()
delete iter1;
```
Test plan (dev server)
```
$make check
```
Simple benchmarking (dev server)
1. The overhead introduced by this PR even when timestamp is disabled.
key size: 16 bytes
value size: 100 bytes
Entries: 1000000
Data reside in main memory, and try to stress iterator.
Repeated three times on master and this PR.
- Seek without next
```
./db_bench -db=/dev/shm/rocksdbtest-1000 -benchmarks=fillseq,seekrandom -enable_pipelined_write=false -disable_wal=true -format_version=3
```
master: 159047.0 ops/sec
this PR: 158922.3 ops/sec (2% drop in throughput)
- Seek and next 10 times
```
./db_bench -db=/dev/shm/rocksdbtest-1000 -benchmarks=fillseq,seekrandom -enable_pipelined_write=false -disable_wal=true -format_version=3 -seek_nexts=10
```
master: 109539.3 ops/sec
this PR: 107519.7 ops/sec (2% drop in throughput)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6255
Differential Revision: D19438227
Pulled By: riversand963
fbshipit-source-id: b66b4979486f8474619f4aa6bdd88598870b0746
Summary:
In direct IO mode, RandomAccessFileReader::Read allocates an internal aligned buffer, and then copies the result into the scratch buffer. If the result is only temporarily used inside a function, there is no need to do the memcpy and just let the result Slice refer to the internally allocated buffer.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6455
Test Plan: make check
Differential Revision: D20106753
Pulled By: cheng-chang
fbshipit-source-id: 44f505843837bba47a56e3fa2c4dd3bd76486b58
Summary:
Check for sys/auxv.h and getauxval before using them as they are not
always available (for example on uclibc)
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6359
Differential Revision: D20239797
fbshipit-source-id: 175a098094d81545628c2372e7c388e70a32fd48
Summary:
When dynamically linking two binaries together, different builds of RocksDB from two sources might cause errors. To provide a tool for user to solve the problem, the RocksDB namespace is changed to a flag which can be overridden in build time.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6433
Test Plan: Build release, all and jtest. Try to build with ROCKSDB_NAMESPACE with another flag.
Differential Revision: D19977691
fbshipit-source-id: aa7f2d0972e1c31d75339ac48478f34f6cfcfb3e
Summary:
1. remove AssertEmpty because calling methods on moved objects is discouraged.
2. add a test to assert that the internal buffer is moved instead of being copied.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6399
Test Plan:
make slice_test && ./slice_test
USE_CLANG=1 make analyze
Differential Revision: D19825372
Pulled By: cheng-chang
fbshipit-source-id: 2e26f8ce5ec3edbfce067db045e80bd433e704f4
Summary:
Add a utility class `Defer` to defer the execution of a function until the Defer object goes out of scope.
Used in VersionSet:: ProcessManifestWrites as an example.
The inline comments for class `Defer` have more details.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6382
Test Plan: `make defer_test version_set_test && ./defer_test && ./version_set_test`
Differential Revision: D19797538
Pulled By: cheng-chang
fbshipit-source-id: b1a9b7306e4fd4f48ec2ab55783caa561a315f0f
Summary:
In the current code base, RocksDB generate the checksum for each block and verify the checksum at usage. Current PR enable SST file checksum. After a SST file is generated by Flush or Compaction, RocksDB generate the SST file checksum and store the checksum value and checksum method name in the vs_info and MANIFEST as part for the FileMetadata.
Added the enable_sst_file_checksum to Options to enable or disable file checksum. Added sst_file_checksum to Options such that user can plugin their own SST file checksum calculate method via overriding the SstFileChecksum class. The checksum information inlcuding uint32_t checksum value and a checksum name (string). A new tool is added to LDB such that user can dump out a list of file checksum information from MANIFEST. If user enables the file checksum but does not provide the sst_file_checksum instance, RocksDB will use the default crc32checksum implemented in table/sst_file_checksum_crc32c.h
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6216
Test Plan: Added the testing case in table_test and ldb_cmd_test to verify checksum is correct in different level. Pass make asan_check.
Differential Revision: D19171461
Pulled By: zhichao-cao
fbshipit-source-id: b2e53479eefc5bb0437189eaa1941670e5ba8b87
Summary:
Right, when reading from option files, no readahead is used and 8KB buffer is used. It might introduce high latency if the file system provide high latency and doesn't do readahead. Instead, introduce a readahead to the file. When calling inside DB, infer the value from options.log_readahead. Otherwise, a default 512KB readahead size is used.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6372
Test Plan: Add --log_readahead_size in db_bench. Run it with several options and observe read size from option files using strace.
Differential Revision: D19727739
fbshipit-source-id: e6d8053b0a64259abc087f1f388b9cd66fa8a583
Summary:
It's logically correct for PinnableSlice to support move semantics to transfer ownership of the pinned memory region. This PR adds both move constructor and move assignment to PinnableSlice.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6374
Test Plan:
A set of unit tests for the move semantics are added in slice_test.
So `make slice_test && ./slice_test`.
Differential Revision: D19739254
Pulled By: cheng-chang
fbshipit-source-id: f898bd811bb05b2d87384ec58b645e9915e8e0b1
Summary:
Unit test names, together with other components, are used to create log files
during some internal testing. Overly long names cause infra failure due to file
names being too long.
Look for internal tests.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6352
Differential Revision: D19649307
Pulled By: riversand963
fbshipit-source-id: 6f29de096e33c0eaa87d9c8702f810eda50059e7
Summary:
With many millions of keys, the old Bloom filter implementation
for the block-based table (format_version <= 4) would have excessive FP
rate due to the limitations of feeding the Bloom filter with a 32-bit hash.
This change computes an estimated inflated FP rate due to this effect
and warns in the log whenever an SST filter is constructed (almost
certainly a "full" not "partitioned" filter) that exceeds 1.5x FP rate
due to this effect. The detailed condition is only checked if 3 million
keys or more have been added to a filter, as this should be a lower
bound for common bits/key settings (< 20).
Recommended remedies include smaller SST file size, using
format_version >= 5 (for new Bloom filter), or using partitioned
filters.
This does not change behavior other than generating warnings for some
constructed filters using the old implementation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6317
Test Plan:
Example with warning, 15M keys @ 15 bits / key: (working_mem_size_mb is just to stop after building one filter if it's large)
$ ./filter_bench -quick -impl=0 -working_mem_size_mb=1 -bits_per_key=15 -average_keys_per_filter=15000000 2>&1 | grep 'FP rate'
[WARN] [/block_based/filter_policy.cc:292] Using legacy SST/BBT Bloom filter with excessive key count (15.0M @ 15bpk), causing estimated 1.8x higher filter FP rate. Consider using new Bloom with format_version>=5, smaller SST file size, or partitioned filters.
Predicted FP rate %: 0.766702
Average FP rate %: 0.66846
Example without warning (150K keys):
$ ./filter_bench -quick -impl=0 -working_mem_size_mb=1 -bits_per_key=15 -average_keys_per_filter=150000 2>&1 | grep 'FP rate'
Predicted FP rate %: 0.422857
Average FP rate %: 0.379301
$
With more samples at 15 bits/key:
150K keys -> no warning; actual: 0.379% FP rate (baseline)
1M keys -> no warning; actual: 0.396% FP rate, 1.045x
9M keys -> no warning; actual: 0.563% FP rate, 1.485x
10M keys -> warning (1.5x); actual: 0.564% FP rate, 1.488x
15M keys -> warning (1.8x); actual: 0.668% FP rate, 1.76x
25M keys -> warning (2.4x); actual: 0.880% FP rate, 2.32x
At 10 bits/key:
150K keys -> no warning; actual: 1.17% FP rate (baseline)
1M keys -> no warning; actual: 1.16% FP rate
10M keys -> no warning; actual: 1.32% FP rate, 1.13x
25M keys -> no warning; actual: 1.63% FP rate, 1.39x
35M keys -> warning (1.6x); actual: 1.81% FP rate, 1.55x
At 5 bits/key:
150K keys -> no warning; actual: 9.32% FP rate (baseline)
25M keys -> no warning; actual: 9.62% FP rate, 1.03x
200M keys -> no warning; actual: 12.2% FP rate, 1.31x
250M keys -> warning (1.5x); actual: 12.8% FP rate, 1.37x
300M keys -> warning (1.6x); actual: 13.4% FP rate, 1.43x
The reason for the modest inaccuracy at low bits/key is that the assumption of independence between a collision between 32-hash values feeding the filter and an FP in the filter is not quite true for implementations using "simple" logic to compute indices from the stock hash result. There's math on this in my dissertation, but I don't think it's worth the effort just for these extreme cases (> 100 million keys and low-ish bits/key).
Differential Revision: D19471715
Pulled By: pdillinger
fbshipit-source-id: f80c96893a09bf1152630ff0b964e5cdd7e35c68
Summary:
Help users that would benefit most from new Bloom filter
implementation by logging a warning that recommends the using
format_version >= 5.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6312
Test Plan:
$ (for BPK in 10 13 14 19 20 50; do ./filter_bench -quick -impl=0 -bits_per_key=$BPK -m_queries=1 2>&1; done) | grep 'its/key'
Bits/key actual: 10.0647
Bits/key actual: 13.0593
[WARN] [/block_based/filter_policy.cc:546] Using legacy Bloom filter with high (14) bits/key. Significant filter space and/or accuracy improvement is available with format_verion>=5.
Bits/key actual: 14.0581
[WARN] [/block_based/filter_policy.cc:546] Using legacy Bloom filter with high (19) bits/key. Significant filter space and/or accuracy improvement is available with format_verion>=5.
Bits/key actual: 19.0542
[WARN] [/block_based/filter_policy.cc:546] Using legacy Bloom filter with high (20) bits/key. Dramatic filter space and/or accuracy improvement is available with format_verion>=5.
Bits/key actual: 20.0584
[WARN] [/block_based/filter_policy.cc:546] Using legacy Bloom filter with high (50) bits/key. Dramatic filter space and/or accuracy improvement is available with format_verion>=5.
Bits/key actual: 50.0577
Differential Revision: D19457191
Pulled By: pdillinger
fbshipit-source-id: 073d94cde5c70e03a160f953e1100c15ea83eda4
Summary:
Several improvements to crash_test/stress_test:
(1) Stress_test to support an parameter of bottommost compression
(2) Rename those FLAGS_* variables that are not gflags to avoid confusion
(3) Crash_test to randomly generate compression type for bottommost compression with half the chance.
(4) Stress_test to sanitize unsupported compression type to snappy, so that crash_test to cover all possible compression types and people don't need to worry about they don't support all comrpession types in their environment.
(5) In crash_test, when generating db_stress command, sort arguments in alphabeta order, so that it is easier to find value for a specific argument.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6215
Test Plan: Run "make crash_test" for a while and see the botommost option shown in LOG files.
Differential Revision: D19171255
fbshipit-source-id: d7001e246c4ff9ee5760776eea0be97738650735
Summary:
The filter bits builder collects all the hashes to add in memory before adding them (because the number of keys is not known until we've walked over all the keys). Existing code uses a std::vector for this, which can mean up to 2x than necessary space allocated (and not freed) and up to ~2x write amplification in memory. Using std::deque uses close to minimal space (for large filters, the only time it matters), no write amplification, frees memory while building, and no need for large contiguous memory area. The only cost is more calls to allocator, which does not appear to matter, at least in benchmark test.
For now, this change only applies to the new (format_version=5) Bloom filter implementation, to ease before-and-after comparison downstream.
Temporary memory use during build is about the only way the new Bloom filter could regress vs. the old (because of upgrade to 64-bit hash) and that should only matter for full filters. This change should largely mitigate that potential regression.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6175
Test Plan:
Using filter_bench with -new_builder option and 6M keys per filter is like large full filter (improvement). 10k keys and no -new_builder is like partitioned filters (about the same). (Corresponding configurations run simultaneously on devserver.)
std::vector impl (before)
$ /usr/bin/time -v ./filter_bench -impl=2 -quick -new_builder -working_mem_size_mb=1000 -
average_keys_per_filter=6000000
Build avg ns/key: 52.2027
Maximum resident set size (kbytes): 1105016
$ /usr/bin/time -v ./filter_bench -impl=2 -quick -working_mem_size_mb=1000 -
average_keys_per_filter=10000
Build avg ns/key: 30.5694
Maximum resident set size (kbytes): 1208152
std::deque impl (after)
$ /usr/bin/time -v ./filter_bench -impl=2 -quick -new_builder -working_mem_size_mb=1000 -
average_keys_per_filter=6000000
Build avg ns/key: 39.0697
Maximum resident set size (kbytes): 1087196
$ /usr/bin/time -v ./filter_bench -impl=2 -quick -working_mem_size_mb=1000 -
average_keys_per_filter=10000
Build avg ns/key: 30.9348
Maximum resident set size (kbytes): 1207980
Differential Revision: D19053431
Pulled By: pdillinger
fbshipit-source-id: 2888e748723a19d9ea40403934f13cbb8483430c
Summary:
The current Env API encompasses both storage/file operations, as well as OS related operations. Most of the APIs return a Status, which does not have enough metadata about an error, such as whether its retry-able or not, scope (i.e fault domain) of the error etc., that may be required in order to properly handle a storage error. The file APIs also do not provide enough control over the IO SLA, such as timeout, prioritization, hinting about placement and redundancy etc.
This PR separates out the file/storage APIs from Env into a new FileSystem class. The APIs are updated to return an IOStatus with metadata about the error, as well as to take an IOOptions structure as input in order to allow more control over the IO.
The user can set both ```options.env``` and ```options.file_system``` to specify that RocksDB should use the former for OS related operations and the latter for storage operations. Internally, a ```CompositeEnvWrapper``` has been introduced that inherits from ```Env``` and redirects individual methods to either an ```Env``` implementation or the ```FileSystem``` as appropriate. When options are sanitized during ```DB::Open```, ```options.env``` is replaced with a newly allocated ```CompositeEnvWrapper``` instance if both env and file_system have been specified. This way, the rest of the RocksDB code can continue to function as before.
This PR also ports PosixEnv to the new API by splitting it into two - PosixEnv and PosixFileSystem. PosixEnv is defined as a sub-class of CompositeEnvWrapper, and threading/time functions are overridden with Posix specific implementations in order to avoid an extra level of indirection.
The ```CompositeEnvWrapper``` translates ```IOStatus``` return code to ```Status```, and sets the severity to ```kSoftError``` if the io_status is retryable. The error handling code in RocksDB can then recover the DB automatically.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5761
Differential Revision: D18868376
Pulled By: anand1976
fbshipit-source-id: 39efe18a162ea746fabac6360ff529baba48486f
Summary:
And clean up related code, especially in stress test.
(More clean up of db_stress_test_base.cc coming after this.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6154
Test Plan: make check, make blackbox_crash_test for a bit
Differential Revision: D18938180
Pulled By: pdillinger
fbshipit-source-id: 524d27621b8dbb25f6dff40f1081e7c00630357e
Summary:
This change fixes a source issue that caused compile time error which breaks build for many fbcode services in that setup. The size() member function of channel is a const member, so member variables accessed within it are implicitly const as well. This caused error when clang fails to resolve to a constructor that takes std::mutex because the suitable constructor got rejected due to loss of constness for its argument. The fix is to add mutable modifier to the lock_ member of channel.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6161
Differential Revision: D18967685
Pulled By: maysamyabandeh
fbshipit-source-id: 698b6a5153c3c92eeacb842c467aa28cc350d432
Summary:
thread_local_test now fails because it asserts no thread local instance is created when the test started. However, right now a thread local instance might be created when creating PosixEnv as a static variable. Fix the test by relaxing the assumption of starting from 0.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6136
Test Plan: Find an environment where the test fails, and see it passes with the fix applied.
Differential Revision: D18889224
fbshipit-source-id: 7946f3bfea81d236f7bb1554076696705b211b92
Summary:
From the reset of the code, it looks this this maybe can be unconditionally given the attribute? But I couldn't test with MSVC so I defensively put under CPP.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6075
Differential Revision: D18723749
fbshipit-source-id: 45fc8732c28dd29aab1644225d68f3c6f39bd69b
Summary:
This change enables custom implementations of FilterPolicy to
wrap a variety of NewBloomFilterPolicy and select among them based on
contextual information such as table level and compaction style.
* Moves FilterBuildingContext to public API and elaborates it with more
useful data. (It would be nice to put more general options-like data,
but at the time this object is constructed, we are using internal APIs
ImmutableCFOptions and MutableCFOptions and don't have easy access to
ColumnFamilyOptions that I can tell.)
* Renames BloomFilterPolicy::GetFilterBitsBuilderInternal to
GetBuilderWithContext, because it's now public.
* Plumbs through the table's "level_at_creation" for filter building
context.
* Simplified some tests by adding GetBuilder() to
MockBlockBasedTableTester.
* Adds test as DBBloomFilterTest.ContextCustomFilterPolicy, including
sample wrapper class LevelAndStyleCustomFilterPolicy.
* Fixes a cross-test bug in DBBloomFilterTest.OptimizeFiltersForHits
where it does not reset perf context.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6088
Test Plan: make check, valgrind on db_bloom_filter_test
Differential Revision: D18697817
Pulled By: pdillinger
fbshipit-source-id: 5f987a2d7b07cc7a33670bc08ca6b4ca698c1cf4
Summary:
There's no technological impediment to allowing the Bloom
filter bits/key to be non-integer (fractional/decimal) values, and it
provides finer control over the memory vs. accuracy trade-off. This is
especially handy in using the format_version=5 Bloom filter in place
of the old one, because bits_per_key=9.55 provides the same accuracy as
the old bits_per_key=10.
This change not only requires refining the logic for choosing the best
num_probes for a given bits/key setting, it revealed a flaw in that logic.
As bits/key gets higher, the best num_probes for a cache-local Bloom
filter is closer to bpk / 2 than to bpk * 0.69, the best choice for a
standard Bloom filter. For example, at 16 bits per key, the best
num_probes is 9 (FP rate = 0.0843%) not 11 (FP rate = 0.0884%).
This change fixes and refines that logic (for the format_version=5
Bloom filter only, just in case) based on empirical tests to find
accuracy inflection points between each num_probes.
Although bits_per_key is now specified as a double, the new Bloom
filter converts/rounds this to "millibits / key" for predictable/precise
internal computations. Just in case of unforeseen compatibility
issues, we round to the nearest whole number bits / key for the
legacy Bloom filter, so as not to unlock new behaviors for it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6092
Test Plan: unit tests included
Differential Revision: D18711313
Pulled By: pdillinger
fbshipit-source-id: 1aa73295f152a995328cb846ef9157ae8a05522a
Summary:
As described in detail in issue https://github.com/facebook/rocksdb/issues/6048, iterators' dereference operators
(`*`, `->`, and `[]`) should return `pointer`s/`reference`s (as opposed to
`const_pointer`s/`const_reference`s) even if the iterator itself is `const`
to be in sync with the standard's iterator concept.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6057
Test Plan: make check
Differential Revision: D18623235
Pulled By: ltamasi
fbshipit-source-id: 04e82d73bc0c67fb0ded018383af8dfc332050cc
Summary:
This is a required operator for random-access iterators, and an upcoming update for Visual Studio 2019 will change the C++ Standard Library's heap algorithms to use this operator.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6047
Differential Revision: D18618531
Pulled By: ltamasi
fbshipit-source-id: 08d10bc85bf2dbc3f7ef0fa3c777e99f1e927ef5
Summary:
Adds an improved, replacement Bloom filter implementation (FastLocalBloom) for full and partitioned filters in the block-based table. This replacement is faster and more accurate, especially for high bits per key or millions of keys in a single filter.
Speed
The improved speed, at least on recent x86_64, comes from
* Using fastrange instead of modulo (%)
* Using our new hash function (XXH3 preview, added in a previous commit), which is much faster for large keys and only *slightly* slower on keys around 12 bytes if hashing the same size many thousands of times in a row.
* Optimizing the Bloom filter queries with AVX2 SIMD operations. (Added AVX2 to the USE_SSE=1 build.) Careful design was required to support (a) SIMD-optimized queries, (b) compatible non-SIMD code that's simple and efficient, (c) flexible choice of number of probes, and (d) essentially maximized accuracy for a cache-local Bloom filter. Probes are made eight at a time, so any number of probes up to 8 is the same speed, then up to 16, etc.
* Prefetching cache lines when building the filter. Although this optimization could be applied to the old structure as well, it seems to balance out the small added cost of accumulating 64 bit hashes for adding to the filter rather than 32 bit hashes.
Here's nominal speed data from filter_bench (200MB in filters, about 10k keys each, 10 bits filter data / key, 6 probes, avg key size 24 bytes, includes hashing time) on Skylake DE (relatively low clock speed):
$ ./filter_bench -quick -impl=2 -net_includes_hashing # New Bloom filter
Build avg ns/key: 47.7135
Mixed inside/outside queries...
Single filter net ns/op: 26.2825
Random filter net ns/op: 150.459
Average FP rate %: 0.954651
$ ./filter_bench -quick -impl=0 -net_includes_hashing # Old Bloom filter
Build avg ns/key: 47.2245
Mixed inside/outside queries...
Single filter net ns/op: 63.2978
Random filter net ns/op: 188.038
Average FP rate %: 1.13823
Similar build time but dramatically faster query times on hot data (63 ns to 26 ns), and somewhat faster on stale data (188 ns to 150 ns). Performance differences on batched and skewed query loads are between these extremes as expected.
The only other interesting thing about speed is "inside" (query key was added to filter) vs. "outside" (query key was not added to filter) query times. The non-SIMD implementations are substantially slower when most queries are "outside" vs. "inside". This goes against what one might expect or would have observed years ago, as "outside" queries only need about two probes on average, due to short-circuiting, while "inside" always have num_probes (say 6). The problem is probably the nastily unpredictable branch. The SIMD implementation has few branches (very predictable) and has pretty consistent running time regardless of query outcome.
Accuracy
The generally improved accuracy (re: Issue https://github.com/facebook/rocksdb/issues/5857) comes from a better design for probing indices
within a cache line (re: Issue https://github.com/facebook/rocksdb/issues/4120) and improved accuracy for millions of keys in a single filter from using a 64-bit hash function (XXH3p). Design details in code comments.
Accuracy data (generalizes, except old impl gets worse with millions of keys):
Memory bits per key: FP rate percent old impl -> FP rate percent new impl
6: 5.70953 -> 5.69888
8: 2.45766 -> 2.29709
10: 1.13977 -> 0.959254
12: 0.662498 -> 0.411593
16: 0.353023 -> 0.0873754
24: 0.261552 -> 0.0060971
50: 0.225453 -> ~0.00003 (less than 1 in a million queries are FP)
Fixes https://github.com/facebook/rocksdb/issues/5857
Fixes https://github.com/facebook/rocksdb/issues/4120
Unlike the old implementation, this implementation has a fixed cache line size (64 bytes). At 10 bits per key, the accuracy of this new implementation is very close to the old implementation with 128-byte cache line size. If there's sufficient demand, this implementation could be generalized.
Compatibility
Although old releases would see the new structure as corrupt filter data and read the table as if there's no filter, we've decided only to enable the new Bloom filter with new format_version=5. This provides a smooth path for automatic adoption over time, with an option for early opt-in.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6007
Test Plan: filter_bench has been used thoroughly to validate speed, accuracy, and correctness. Unit tests have been carefully updated to exercise new and old implementations, as well as the logic to select an implementation based on context (format_version).
Differential Revision: D18294749
Pulled By: pdillinger
fbshipit-source-id: d44c9db3696e4d0a17caaec47075b7755c262c5f
Summary:
For upcoming new SST filter implementations, we will use a new
64-bit hash function (XXH3 preview, slightly modified). This change
updates hash.{h,cc} for that change, adds unit tests, and out-of-lines
the implementations to keep hash.h as clean/small as possible.
In developing the unit tests, I discovered that the XXH3 preview always
returns zero for the empty string. Zero is problematic for some
algorithms (including an upcoming SST filter implementation) if it
occurs more often than at the "natural" rate, so it should not be
returned from trivial values using trivial seeds. I modified our fork
of XXH3 to return a modest hash of the seed for the empty string.
With hash function details out-of-lines in hash.h, it makes sense to
enable XXH_INLINE_ALL, so that direct calls to XXH64/XXH32/XXH3p
are inlined. To fix array-bounds warnings on some inline calls, I
injected some casts to uintptr_t in xxhash.cc. (Issue reported to Yann.)
Revised: Reverted using XXH_INLINE_ALL for now. Some Facebook
checks are unhappy about #include on xxhash.cc file. I would
fix that by rename to xxhash_cc.h, but to best preserve history I want
to do that in a separate commit (PR) from the uintptr casts.
Also updated filter_bench for this change, improving the performance
predictability of dry run hashing and adding support for 64-bit hash
(for upcoming new SST filter implementations, minor dead code in the
tool for now).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5984
Differential Revision: D18246567
Pulled By: pdillinger
fbshipit-source-id: 6162fbf6381d63c8cc611dd7ec70e1ddc883fbb8
Summary:
This change sets up for alternate implementations underlying
BloomFilterPolicy:
* Refactor BloomFilterPolicy and expose in internal .h file so that it's easy to iterate over / select implementations for testing, regardless of what the best public interface will look like. Most notably updated db_bloom_filter_test to use this.
* Hide FullFilterBitsBuilder from unit tests (alternate derived classes planned); expose the part important for testing (CalculateSpace), as abstract class BuiltinFilterBitsBuilder. (Also cleaned up internally exposed interface to CalculateSpace.)
* Rename BloomTest -> BlockBasedBloomTest for clarity (despite ongoing confusion between block-based table and block-based filter)
* Assert that block-based filter construction interface is only used on BloomFilterPolicy appropriately constructed. (A couple of tests updated to add ", true".)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5967
Test Plan: make check
Differential Revision: D18138704
Pulled By: pdillinger
fbshipit-source-id: 55ef9273423b0696309e251f50b8c1b5e9ec7597
Summary:
filter_bench is a specialized micro-benchmarking tool that
should not be needed with ROCKSDB_LITE. This should fix the LITE build.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5978
Test Plan: make LITE=1 check
Differential Revision: D18177941
Pulled By: pdillinger
fbshipit-source-id: b73a171404661e09e018bc99afcf8d4bf1e2949c
Summary:
* Adds support for plain table filter. This is not critical right now, but does add a -impl flag that will be useful for new filter implementations initially targeted at block-based table (and maybe later ported to plain table)
* Better mixing of inside vs. outside queries, for more realism
* A -best_case option handy for implementation tuning inner loop
* Option for whether to include hashing time in dry run / net timings
No modifications to production code, just filter_bench.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5968
Differential Revision: D18139872
Pulled By: pdillinger
fbshipit-source-id: 5b09eba963111b48f9e0525a706e9921070990e8
Summary:
Some filtering tests were unfriendly to new implementations of
FilterBitsBuilder because of dynamic_cast to FullFilterBitsBuilder. Most
of those have now been cleaned up, worked around, or at least changed
from crash on dynamic_cast failure to individual test failure.
Also put some clarifying comments on filter-related APIs.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5960
Test Plan: make check
Differential Revision: D18121223
Pulled By: pdillinger
fbshipit-source-id: e83827d9d5d96315d96f8e25a99cd70f497d802c
Peter Dillinger