Summary:
According to https://github.com/facebook/rocksdb/issues/5907, each filter partition "should include the bloom of the prefix of the last
key in the previous partition" so that SeekForPrev() in prefix mode can return correct result.
The prefix of the last key in the previous partition does not necessarily have the same prefix
as the first key in the current partition. Regardless of the first key in current partition, the
prefix of the last key in the previous partition should be added. The existing code, however,
does not follow this. Furthermore, there is another issue: when finishing current filter partition,
`FullFilterBlockBuilder::AddPrefix()` is called for the first key in next filter partition, which effectively
overwrites `last_prefix_str_` prematurely. Consequently, when the filter block builder proceeds
to the next partition, `last_prefix_str_` will be the prefix of its first key, leaving no way of adding
the bloom of the prefix of the last key of the previous partition.
Prefix extractor is FixedLength.2.
```
[ filter part 1 ] [ filter part 2 ]
abc d
```
When SeekForPrev("abcd"), checking the filter partition will land on filter part 2 because "abcd" > "abc"
but smaller than "d".
If the filter in filter part 2 happens to return false for the test for "ab", then SeekForPrev("abcd") will build
incorrect iterator tree in non-total-order mode.
Also fix a unit test which starts to fail following this PR. `InDomain` should not fail due to assertion
error when checking on an arbitrary key.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8137
Test Plan:
```
make check
```
Without this fix, the following command will fail pretty soon.
```
./db_stress --acquire_snapshot_one_in=10000 --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=17 \
--bottommost_compression_type=disable --cache_index_and_filter_blocks=1 --cache_size=1048576 \
--checkpoint_one_in=0 --checksum_type=kxxHash64 --clear_column_family_one_in=0 \
--compact_files_one_in=1000000 --compact_range_one_in=1000000 --compaction_ttl=0 \
--compression_max_dict_buffer_bytes=0 --compression_max_dict_bytes=0 \
--compression_parallel_threads=1 --compression_type=zstd --compression_zstd_max_train_bytes=0 \
--continuous_verification_interval=0 --db=/dev/shm/rocksdb/rocksdb_crashtest_whitebox \
--db_write_buffer_size=8388608 --delpercent=5 --delrangepercent=0 --destroy_db_initially=0 --enable_blob_files=0 \
--enable_compaction_filter=0 --enable_pipelined_write=1 --file_checksum_impl=big --flush_one_in=1000000 \
--format_version=5 --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=4 --index_type=2 --ingest_external_file_one_in=0 \
--iterpercent=10 --key_len_percent_dist=1,30,69 --level_compaction_dynamic_level_bytes=True \
--log2_keys_per_lock=10 --long_running_snapshots=1 --mark_for_compaction_one_file_in=0 \
--max_background_compactions=20 --max_bytes_for_level_base=10485760 --max_key=100000000 --max_key_len=3 \
--max_manifest_file_size=1073741824 --max_write_batch_group_size_bytes=16777216 --max_write_buffer_number=3 \
--max_write_buffer_size_to_maintain=8388608 --memtablerep=skip_list --mmap_read=1 --mock_direct_io=False \
--nooverwritepercent=0 --open_files=500000 --ops_per_thread=20000000 --optimize_filters_for_memory=0 --paranoid_file_checks=1 --partition_filters=1 --partition_pinning=0 --pause_background_one_in=1000000 \
--periodic_compaction_seconds=0 --prefixpercent=5 --progress_reports=0 --read_fault_one_in=0 --read_only=0 \
--readpercent=45 --recycle_log_file_num=0 --reopen=20 --secondary_catch_up_one_in=0 \
--snapshot_hold_ops=100000 --sst_file_manager_bytes_per_sec=104857600 \
--sst_file_manager_bytes_per_truncate=0 --subcompactions=2 --sync=0 --sync_fault_injection=False \
--target_file_size_base=2097152 --target_file_size_multiplier=2 --test_batches_snapshots=0 --test_cf_consistency=0 \
--top_level_index_pinning=0 --unpartitioned_pinning=1 --use_blob_db=0 --use_block_based_filter=0 \
--use_direct_io_for_flush_and_compaction=0 --use_direct_reads=0 --use_full_merge_v1=0 --use_merge=0 \
--use_multiget=0 --use_ribbon_filter=0 --use_txn=0 --user_timestamp_size=8 --verify_checksum=1 \
--verify_checksum_one_in=1000000 --verify_db_one_in=100000 --write_buffer_size=4194304 \
--write_dbid_to_manifest=1 --writepercent=35
```
Reviewed By: pdillinger
Differential Revision: D27553054
Pulled By: riversand963
fbshipit-source-id: 60e391e4a2d8d98a9a3172ec5d6176b90ec3de98
Summary:
Return early in case there are zero data blocks when
`BlockBasedTableBuilder::EnterUnbuffered()` is called. This crash can
only be triggered by applying dictionary compression to SST files that
contain only range tombstones. It cannot be triggered by a low buffer
limit alone since we only consider entering unbuffered mode after
buffering a data block causing the limit to be breached, or `Finish()`ing the file. It also cannot
be triggered by a totally empty file because those go through
`Abandon()` rather than `Finish()` so unbuffered mode is never entered.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8141
Test Plan: added a unit test that repro'd the "Floating point exception"
Reviewed By: riversand963
Differential Revision: D27495640
Pulled By: ajkr
fbshipit-source-id: a463cfba476919dc5c5c380800a75a86c31ffa23
Summary:
When timestamp is enabled, key comparison should take this into account.
In `BlockBasedTableReader::Get()`, `BlockBasedTableReader::MultiGet()`,
assume the target key is `key`, and the timestamp upper bound is `ts`.
The highest key in current block is (key, ts1), while the lowest key in next
block is (key, ts2).
If
```
ts1 > ts > ts2
```
then
```
(key, ts1) < (key, ts) < (key, ts2)
```
It can be shown that if `Compare()` is used, then we will mistakenly skip the next
block. Instead, we should use `CompareWithoutTimestamp()`.
The majority of this PR makes some existing tests in `db_with_timestamp_basic_test.cc`
parameterized so that different index types can be tested. A new unit test is
also added for more coverage.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8062
Test Plan: make check
Reviewed By: ltamasi
Differential Revision: D27057557
Pulled By: riversand963
fbshipit-source-id: c1062fa7c159ed600a1ad7e461531d52265021f1
Summary:
For performance purposes, the lower level routines were changed to use a SystemClock* instead of a std::shared_ptr<SystemClock>. The shared ptr has some performance degradation on certain hardware classes.
For most of the system, there is no risk of the pointer being deleted/invalid because the shared_ptr will be stored elsewhere. For example, the ImmutableDBOptions stores the Env which has a std::shared_ptr<SystemClock> in it. The SystemClock* within the ImmutableDBOptions is essentially a "short cut" to gain access to this constant resource.
There were a few classes (PeriodicWorkScheduler?) where the "short cut" property did not hold. In those cases, the shared pointer was preserved.
Using db_bench readrandom perf_level=3 on my EC2 box, this change performed as well or better than 6.17:
6.17: readrandom : 28.046 micros/op 854902 ops/sec; 61.3 MB/s (355999 of 355999 found)
6.18: readrandom : 32.615 micros/op 735306 ops/sec; 52.7 MB/s (290999 of 290999 found)
PR: readrandom : 27.500 micros/op 871909 ops/sec; 62.5 MB/s (367999 of 367999 found)
(Note that the times for 6.18 are prior to revert of the SystemClock).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8033
Reviewed By: pdillinger
Differential Revision: D27014563
Pulled By: mrambacher
fbshipit-source-id: ad0459eba03182e454391b5926bf5cdd45657b67
Summary:
This change only affects non-schema-critical aspects of the production candidate Ribbon filter. Specifically, it refines choice of internal configuration parameters based on inputs. The changes are minor enough that the schema tests in bloom_test, some of which depend on this, are unaffected. There are also some minor optimizations and refactorings.
This would be a schema change for "smash" Ribbon, to fix some known issues with small filters, but "smash" Ribbon is not accessible in public APIs. Unit test CompactnessAndBacktrackAndFpRate updated to test small and medium-large filters. Run with --thoroughness=100 or so for much better detection power (not appropriate for continuous regression testing).
Homogenous Ribbon:
This change adds internally a Ribbon filter variant we call Homogeneous Ribbon, in collaboration with Stefan Walzer. The expected "result" value for every key is zero, instead of computed from a hash. Entropy for queries not to be false positives comes from free variables ("overhead") in the solution structure, which are populated pseudorandomly. Construction is slightly faster for not tracking result values, and never fails. Instead, FP rate can jump up whenever and whereever entries are packed too tightly. For small structures, we can choose overhead to make this FP rate jump unlikely, as seen in updated unit test CompactnessAndBacktrackAndFpRate.
Unlike standard Ribbon, Homogeneous Ribbon seems to scale to arbitrary number of keys when accepting an FP rate penalty for small pockets of high FP rate in the structure. For example, 64-bit ribbon with 8 solution columns and 10% allocated space overhead for slots seems to achieve about 10.5% space overhead vs. information-theoretic minimum based on its observed FP rate with expected pockets of degradation. (FP rate is close to 1/256.) If targeting a higher FP rate with fewer solution columns, Homogeneous Ribbon can be even more space efficient, because the penalty from degradation is relatively smaller. If targeting a lower FP rate, Homogeneous Ribbon is less space efficient, as more allocated overhead is needed to keep the FP rate impact of degradation relatively under control. The new OptimizeHomogAtScale tool in ribbon_test helps to find these optimal allocation overheads for different numbers of solution columns. And Ribbon widths, with 128-bit Ribbon apparently cutting space overheads in half vs. 64-bit.
Other misc item specifics:
* Ribbon APIs in util/ribbon_config.h now provide configuration data for not just 5% construction failure rate (95% success), but also 50% and 0.1%.
* Note that the Ribbon structure does not exhibit "threshold" behavior as standard Xor filter does, so there is a roughly fixed space penalty to cut construction failure rate in half. Thus, there isn't really an "almost sure" setting.
* Although we can extrapolate settings for large filters, we don't have a good formula for configuring smaller filters (< 2^17 slots or so), and efforts to summarize with a formula have failed. Thus, small data is hard-coded from updated FindOccupancy tool.
* Enhances ApproximateNumEntries for public API Ribbon using more precise data (new API GetNumToAdd), thus a more accurate but not perfect reversal of CalculateSpace. (bloom_test updated to expect the greater precision)
* Move EndianSwapValue from coding.h to coding_lean.h to keep Ribbon code easily transferable from RocksDB
* Add some missing 'const' to member functions
* Small optimization to 128-bit BitParity
* Small refactoring of BandingStorage in ribbon_alg.h to support Homogeneous Ribbon
* CompactnessAndBacktrackAndFpRate now has an "expand" test: on construction failure, a possible alternative to re-seeding hash functions is simply to increase the number of slots (allocated space overhead) and try again with essentially the same hash values. (Start locations will be different roundings of the same scaled hash values--because fastrange not mod.) This seems to be as effective or more effective than re-seeding, as long as we increase the number of slots (m) by roughly m += m/w where w is the Ribbon width. This way, there is effectively an expansion by one slot for each ribbon-width window in the banding. (This approach assumes that getting "bad data" from your hash function is as unlikely as it naturally should be, e.g. no adversary.)
* 32-bit and 16-bit Ribbon configurations are added to ribbon_test for understanding their behavior, e.g. with FindOccupancy. They are not considered useful at this time and not tested with CompactnessAndBacktrackAndFpRate.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7879
Test Plan: unit test updates included
Reviewed By: jay-zhuang
Differential Revision: D26371245
Pulled By: pdillinger
fbshipit-source-id: da6600d90a3785b99ad17a88b2a3027710b4ea3a
Summary:
In the adapter class `WritableFileStringStreamAdapter`, which wraps WritableFile to be used for std::ostream, previouly only `std::endl` is considered a special case because `endl` is written by `os.put()` directly without going through `xsputn()`. `os.put()` will call `sputc()` and if we further check the internal implementation of `sputc()`, we will see it is
```
int_type __CLR_OR_THIS_CALL sputc(_Elem _Ch) { // put a character
return 0 < _Pnavail() ? _Traits::to_int_type(*_Pninc() = _Ch) : overflow(_Traits::to_int_type(_Ch));
```
As we explicitly disabled buffering, _Pnavail() is always 0. Thus every write, not captured by xsputn, becomes an overflow.
When I run tests on Windows, I found not only `std::endl` will drop into this case, writing an unsigned long long will also call `os.put()` then followed by `sputc()` and eventually call `overflow()`. Therefore, instead of only checking `std::endl`, we should try to append other characters as well unless the appending operation fails.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7991
Reviewed By: jay-zhuang
Differential Revision: D26615692
Pulled By: ajkr
fbshipit-source-id: 4c0003de1645b9531545b23df69b000e07014468
Summary:
RocksDB does auto-readahead for iterators on noticing more
than two reads for a table file. The readahead starts at 8KB and doubles on every
additional read upto BlockBasedTable::kMaxAutoReadAheadSize which is
256*1024.
This PR adds a new option BlockBasedTableOptions::max_auto_readahead_size which
replaces BlockBasedTable::kMaxAutoReadAheadSize and the new option can be
configured.
If max_auto_readahead_size is set 0 then no implicit auto prefetching will
be done. If max_auto_readahead_size provided is less than
8KB (which is initial readahead size used by rocksdb in case of
auto-readahead), readahead size will remain same as max_auto_readahead_size.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7951
Test Plan: Add new unit test case.
Reviewed By: anand1976
Differential Revision: D26568085
Pulled By: akankshamahajan15
fbshipit-source-id: b6543520fc74e97d859f2002328d4c5254d417af
Summary:
The sample selection technique taken in https://github.com/facebook/rocksdb/issues/7970 was problematic
because it had two code paths for sample selection depending on the
number of data blocks, and one of those code paths involved an
allocation. Using prime numbers, we can consolidate into one code path
without allocation. The downside is there will be values of N (number of
data blocks buffered) that suffer from poor spread in the selected
samples.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7987
Test Plan: `make check -j48`
Reviewed By: pdillinger
Differential Revision: D26586147
Pulled By: ajkr
fbshipit-source-id: 62028e54336fadb6e2c7a7fe6747daa05a263d32
Summary:
For dictionary compression, we need to collect some representative samples of the data to be compressed, which we use to either generate or train (when `CompressionOptions::zstd_max_train_bytes > 0`) a dictionary. Previously, the strategy was to buffer all the data blocks during flush, and up to the target file size during compaction. That strategy allowed us to randomly pick samples from as wide a range as possible that'd be guaranteed to land in a single output file.
However, some users try to make huge files in memory-constrained environments, where this strategy can cause OOM. This PR introduces an option, `CompressionOptions::max_dict_buffer_bytes`, that limits how much data blocks are buffered before we switch to unbuffered mode (which means creating the per-SST dictionary, writing out the buffered data, and compressing/writing new blocks as soon as they are built). It is not strict as we currently buffer more than just data blocks -- also keys are buffered. But it does make a step towards giving users predictable memory usage.
Related changes include:
- Changed sampling for dictionary compression to select unique data blocks when there is limited availability of data blocks
- Made use of `BlockBuilder::SwapAndReset()` to save an allocation+memcpy when buffering data blocks for building a dictionary
- Changed `ParseBoolean()` to accept an input containing characters after the boolean. This is necessary since, with this PR, a value for `CompressionOptions::enabled` is no longer necessarily the final component in the `CompressionOptions` string.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7970
Test Plan:
- updated `CompressionOptions` unit tests to verify limit is respected (to the extent expected in the current implementation) in various scenarios of flush/compaction to bottommost/non-bottommost level
- looked at jemalloc heap profiles right before and after switching to unbuffered mode during flush/compaction. Verified memory usage in buffering is proportional to the limit set.
Reviewed By: pdillinger
Differential Revision: D26467994
Pulled By: ajkr
fbshipit-source-id: 3da4ef9fba59974e4ef40e40c01611002c861465
Summary:
Adds support for prefetching data in Ribbon queries,
which especially optimizes batched Ribbon queries for MultiGet
(~222ns/key to ~97ns/key) but also single key queries on cold memory
(~333ns to ~226ns) because many queries span more than one cache line.
This required some refactoring of the query algorithm, and there
does not appear to be a noticeable regression in "hot memory" query
times (perhaps from 48ns to 50ns).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7889
Test Plan:
existing unit tests, plus performance validation with
filter_bench:
Each data point is the best of two runs. I saturated the machine
CPUs with other filter_bench runs in the background.
Before:
$ ./filter_bench -impl=3 -m_keys_total_max=200 -average_keys_per_filter=100000 -m_queries=50
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 125.86
Number of filters: 1993
Total size (MB): 168.166
Reported total allocated memory (MB): 183.211
Reported internal fragmentation: 8.94626%
Bits/key stored: 7.05341
Prelim FP rate %: 0.951827
----------------------------
Mixed inside/outside queries...
Single filter net ns/op: 48.0111
Batched, prepared net ns/op: 222.384
Batched, unprepared net ns/op: 343.908
Skewed 50% in 1% net ns/op: 252.916
Skewed 80% in 20% net ns/op: 320.579
Random filter net ns/op: 332.957
After:
$ ./filter_bench -impl=3 -m_keys_total_max=200 -average_keys_per_filter=100000 -m_queries=50
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 128.117
Number of filters: 1993
Total size (MB): 168.166
Reported total allocated memory (MB): 183.211
Reported internal fragmentation: 8.94626%
Bits/key stored: 7.05341
Prelim FP rate %: 0.951827
----------------------------
Mixed inside/outside queries...
Single filter net ns/op: 49.8812
Batched, prepared net ns/op: 97.1514
Batched, unprepared net ns/op: 222.025
Skewed 50% in 1% net ns/op: 197.48
Skewed 80% in 20% net ns/op: 212.457
Random filter net ns/op: 226.464
Bloom comparison, for reference:
$ ./filter_bench -impl=2 -m_keys_total_max=200 -average_keys_per_filter=100000 -m_queries=50
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 35.3042
Number of filters: 1993
Total size (MB): 238.488
Reported total allocated memory (MB): 262.875
Reported internal fragmentation: 10.2255%
Bits/key stored: 10.0029
Prelim FP rate %: 0.965327
----------------------------
Mixed inside/outside queries...
Single filter net ns/op: 9.09931
Batched, prepared net ns/op: 34.21
Batched, unprepared net ns/op: 88.8564
Skewed 50% in 1% net ns/op: 139.75
Skewed 80% in 20% net ns/op: 181.264
Random filter net ns/op: 173.88
Reviewed By: jay-zhuang
Differential Revision: D26378710
Pulled By: pdillinger
fbshipit-source-id: 058428967c55ed763698284cd3b4bbe3351b6e69
Summary:
Introduces and uses a SystemClock class to RocksDB. This class contains the time-related functions of an Env and these functions can be redirected from the Env to the SystemClock.
Many of the places that used an Env (Timer, PerfStepTimer, RepeatableThread, RateLimiter, WriteController) for time-related functions have been changed to use SystemClock instead. There are likely more places that can be changed, but this is a start to show what can/should be done. Over time it would be nice to migrate most (if not all) of the uses of the time functions from the Env to the SystemClock.
There are several Env classes that implement these functions. Most of these have not been converted yet to SystemClock implementations; that will come in a subsequent PR. It would be good to unify many of the Mock Timer implementations, so that they behave similarly and be tested similarly (some override Sleep, some use a MockSleep, etc).
Additionally, this change will allow new methods to be introduced to the SystemClock (like https://github.com/facebook/rocksdb/issues/7101 WaitFor) in a consistent manner across a smaller number of classes.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7858
Reviewed By: pdillinger
Differential Revision: D26006406
Pulled By: mrambacher
fbshipit-source-id: ed10a8abbdab7ff2e23d69d85bd25b3e7e899e90
Summary:
Change the StringEnv and related classes to be based on FileSystem APIs rather than the corresponding Env ones. The StringSink and StringSource classes were changed to be based on the corresponding FS file classes.
Part of a cleanup to use the newer interfaces. This change also eliminates some of the casts/wrappers to LegacyFile classes.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7786
Reviewed By: jay-zhuang
Differential Revision: D25761460
Pulled By: anand1976
fbshipit-source-id: 428ae8e32b3db97dbeeca08c9d3bb0d9d4d3a38f
Summary:
RetrieveMultipleBlocks which is used by MultiGet to read data blocks is not updating num_data_read stat in
GetContextStats.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7770
Test Plan: make check -j64
Reviewed By: anand1976
Differential Revision: D25538982
Pulled By: akankshamahajan15
fbshipit-source-id: e3daedb035b1be8ab6af6f115cb3793ccc7b1ec6
Summary:
This disables Linux/amd64 builds in Travis for PRs, and adds a
gcc-10+c++20 build in CircleCI, which should fill out sufficient coverage
vs. what we had in Travis
Fixed a use of std::is_pod, which is deprecated in c++20
Fixed ++ on a volatile in db_repl_stress.cc, with bigger refactoring.
Although ++ on this volatile was probably ok with one thread writer and
one thread reader, the code was still overly complex. There was a
deadcode check for error
`if (replThread.no_read < dataPump.no_records)` which can be proven
never to happen based on the structure of the code. It infinite loops
instead for the case intended to be checked. I just simplified the code
for what should be the same checking power.
Also most configurations seem to be using make parallelism = 2 * vcores,
so fixing / using that.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7791
Test Plan:
CI
and `while ./db_repl_stress; do echo again; done` for a while
Reviewed By: siying
Differential Revision: D25669834
Pulled By: pdillinger
fbshipit-source-id: b2c688053d0b1d52c989903449d3cd27a04130d6
Summary:
Primarily this change refactors the optimize_filters_for_memory
code for Bloom filters, based on malloc_usable_size, to also work for
Ribbon filters.
This change also replaces the somewhat slow but general
BuiltinFilterBitsBuilder::ApproximateNumEntries with
implementation-specific versions for Ribbon (new) and Legacy Bloom
(based on a recently deleted version). The reason is to emphasize
speed in ApproximateNumEntries rather than 100% accuracy.
Justification: ApproximateNumEntries (formerly CalculateNumEntry) is
only used by RocksDB for range-partitioned filters, called each time we
start to construct one. (In theory, it should be possible to reuse the
estimate, but the abstractions provided by FilterPolicy don't really
make that workable.) But this is only used as a heuristic estimate for
hitting a desired partitioned filter size because of alignment to data
blocks, which have various numbers of unique keys or prefixes. The two
factors lead us to prioritize reasonable speed over 100% accuracy.
optimize_filters_for_memory adds extra complication, because precisely
calculating num_entries for some allowed number of bytes depends on state
with optimize_filters_for_memory enabled. And the allocator-agnostic
implementation of optimize_filters_for_memory, using malloc_usable_size,
means we would have to actually allocate memory, many times, just to
precisely determine how many entries (keys) could be added and stay below
some size budget, for the current state. (In a draft, I got this
working, and then realized the balance of speed vs. accuracy was all
wrong.)
So related to that, I have made CalculateSpace, an internal-only API
only used for testing, non-authoritative also if
optimize_filters_for_memory is enabled. This simplifies some code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7774
Test Plan:
unit test updated, and for FilterSize test, range of tested
values is greatly expanded (still super fast)
Also tested `db_bench -benchmarks=fillrandom,stats -bloom_bits=10 -num=1000000 -partition_index_and_filters -format_version=5 [-optimize_filters_for_memory] [-use_ribbon_filter]` with temporary debug output of generated filter sizes.
Bloom+optimize_filters_for_memory:
1 Filter size: 197 (224 in memory)
134 Filter size: 3525 (3584 in memory)
107 Filter size: 4037 (4096 in memory)
Total on disk: 904,506
Total in memory: 918,752
Ribbon+optimize_filters_for_memory:
1 Filter size: 3061 (3072 in memory)
110 Filter size: 3573 (3584 in memory)
58 Filter size: 4085 (4096 in memory)
Total on disk: 633,021 (-30.0%)
Total in memory: 634,880 (-30.9%)
Bloom (no offm):
1 Filter size: 261 (320 in memory)
1 Filter size: 3333 (3584 in memory)
240 Filter size: 3717 (4096 in memory)
Total on disk: 895,674 (-1% on disk vs. +offm; known tolerable overhead of offm)
Total in memory: 986,944 (+7.4% vs. +offm)
Ribbon (no offm):
1 Filter size: 2949 (3072 in memory)
1 Filter size: 3381 (3584 in memory)
167 Filter size: 3701 (4096 in memory)
Total on disk: 624,397 (-30.3% vs. Bloom)
Total in memory: 690,688 (-30.0% vs. Bloom)
Note that optimize_filters_for_memory is even more effective for Ribbon filter than for cache-local Bloom, because it can close the unused memory gap even tighter than Bloom filter, because of 16 byte increments for Ribbon vs. 64 byte increments for Bloom.
Reviewed By: jay-zhuang
Differential Revision: D25592970
Pulled By: pdillinger
fbshipit-source-id: 606fdaa025bb790d7e9c21601e8ea86e10541912
Summary:
Deprecate CalculateNumEntry and replace with
ApproximateNumEntries (better name) using size_t instead of int and
uint32_t, to minimize confusing casts and bad overflow behavior
(possible though probably not realistic). Bloom sizes are now explicitly
capped at max size supported by implementations: just under 4GiB for
fv=5 Bloom, and just under 512MiB for fv<5 Legacy Bloom. This
hardening could help to set up for fuzzing.
Also, since RocksDB only uses this information as an approximation
for trying to hit certain sizes for partitioned filters, it's more important
that the function be reasonably fast than for it to be completely
accurate. It's hard enough to be 100% accurate for Ribbon (currently
reversing CalculateSpace) that adding optimize_filters_for_memory
into the mix is just not worth trying to be 100% accurate for num
entries for bytes.
Also:
- Cleaned up filter_policy.h to remove MSVC warning handling and
potentially unsafe use of exception for "not implemented"
- Correct the number of entries limit beyond which current Ribbon
implementation falls back on Bloom instead.
- Consistently use "num_entries" rather than "num_entry"
- Remove LegacyBloomBitsBuilder::CalculateNumEntry as it's essentially
obsolete from general implementation
BuiltinFilterBitsBuilder::CalculateNumEntries.
- Fix filter_bench to skip some tests that don't make sense when only
one or a small number of filters has been generated.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7726
Test Plan:
expanded existing unit tests for CalculateSpace /
ApproximateNumEntries. Also manually used filter_bench to verify Legacy and
fv=5 Bloom size caps work (much too expensive for unit test). Note that
the actual bits per key is below requested due to space cap.
$ ./filter_bench -impl=0 -bits_per_key=20 -average_keys_per_filter=256000000 -vary_key_count_ratio=0 -m_keys_total_max=256 -allow_bad_fp_rate
...
Total size (MB): 511.992
Bits/key stored: 16.777
...
$ ./filter_bench -impl=2 -bits_per_key=20 -average_keys_per_filter=2000000000 -vary_key_count_ratio=0 -m_keys_total_max=2000
...
Total size (MB): 4096
Bits/key stored: 17.1799
...
$
Reviewed By: jay-zhuang
Differential Revision: D25239800
Pulled By: pdillinger
fbshipit-source-id: f94e6d065efd31e05ec630ae1a82e6400d8390c4
Summary:
Ensure that when direct IO is enabled and a compressed block cache is
configured, MultiGet inserts compressed data blocks into the compressed
block cache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7756
Test Plan: Add unit test to db_basic_test
Reviewed By: cheng-chang
Differential Revision: D25416240
Pulled By: anand1976
fbshipit-source-id: 75d57526370c9c0a45ff72651f3278dbd8a9086f
Summary:
Instead of using `EncodeFixed32` which always serialize a integer to
little endian, we should use the local machine's endianness when
populating a native data structure during options parsing.
Without this fix, `read_amp_bytes_per_bit` may be populated incorrectly
on big-endian machines.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7680
Test Plan: make check
Reviewed By: pdillinger
Differential Revision: D24999166
Pulled By: riversand963
fbshipit-source-id: dc603cff6e17f8fa32479ce6df93b93082e6b0c4
Summary:
A temporary hack to work around a bug in 6.10, 6.11, 6.12, 6.13 and
6.14. The bug will write out 8 bytes to OPTIONS file from the starting
address of BlockBasedTableOptions.read_amp_bytes_per_bit which is
actually a uint32. Consequently, the value of read_amp_bytes_per_bit
written in the OPTIONS file is wrong. From 6.15, RocksDB will
try to parse the read_amp_bytes_per_bit from OPTIONS file as a uint32.
To be able to load OPTIONS file generated by affected releases before
the fix, we need to manually parse read_amp_bytes_per_bit with this hack.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7659
Test Plan:
Generate a db with current 6.14.fb (head at b6db05dbb5). Maybe use db_stress.
Checkout this PR, run
```
~/rocksdb/ldb --db=. --try_load_options --ignore_unknown_options idump --count_only
```
Expect success, and should not see
```
Failed: Invalid argument: Error parsing read_amp_bytes_per_bit:17179869184
```
Also
make check
Reviewed By: anand1976
Differential Revision: D24954752
Pulled By: riversand963
fbshipit-source-id: c7b802fc3e52acd050a4fc1cd475016122234394
Summary:
Added experimental public API for Ribbon filter:
NewExperimentalRibbonFilterPolicy(). This experimental API will
take a "Bloom equivalent" bits per key, and configure the Ribbon
filter for the same FP rate as Bloom would have but ~30% space
savings. (Note: optimize_filters_for_memory is not yet implemented
for Ribbon filter. That can be added with no effect on schema.)
Internally, the Ribbon filter is configured using a "one_in_fp_rate"
value, which is 1 over desired FP rate. For example, use 100 for 1%
FP rate. I'm expecting this will be used in the future for configuring
Bloom-like filters, as I expect people to more commonly hold constant
the filter accuracy and change the space vs. time trade-off, rather than
hold constant the space (per key) and change the accuracy vs. time
trade-off, though we might make that available.
### Benchmarking
```
$ ./filter_bench -impl=2 -quick -m_keys_total_max=200 -average_keys_per_filter=100000 -net_includes_hashing
Building...
Build avg ns/key: 34.1341
Number of filters: 1993
Total size (MB): 238.488
Reported total allocated memory (MB): 262.875
Reported internal fragmentation: 10.2255%
Bits/key stored: 10.0029
----------------------------
Mixed inside/outside queries...
Single filter net ns/op: 18.7508
Random filter net ns/op: 258.246
Average FP rate %: 0.968672
----------------------------
Done. (For more info, run with -legend or -help.)
$ ./filter_bench -impl=3 -quick -m_keys_total_max=200 -average_keys_per_filter=100000 -net_includes_hashing
Building...
Build avg ns/key: 130.851
Number of filters: 1993
Total size (MB): 168.166
Reported total allocated memory (MB): 183.211
Reported internal fragmentation: 8.94626%
Bits/key stored: 7.05341
----------------------------
Mixed inside/outside queries...
Single filter net ns/op: 58.4523
Random filter net ns/op: 363.717
Average FP rate %: 0.952978
----------------------------
Done. (For more info, run with -legend or -help.)
```
168.166 / 238.488 = 0.705 -> 29.5% space reduction
130.851 / 34.1341 = 3.83x construction time for this Ribbon filter vs. lastest Bloom filter (could make that as little as about 2.5x for less space reduction)
### Working around a hashing "flaw"
bloom_test discovered a flaw in the simple hashing applied in
StandardHasher when num_starts == 1 (num_slots == 128), showing an
excessively high FP rate. The problem is that when many entries, on the
order of number of hash bits or kCoeffBits, are associated with the same
start location, the correlation between the CoeffRow and ResultRow (for
efficiency) can lead to a solution that is "universal," or nearly so, for
entries mapping to that start location. (Normally, variance in start
location breaks the effective association between CoeffRow and
ResultRow; the same value for CoeffRow is effectively different if start
locations are different.) Without kUseSmash and with num_starts > 1 (thus
num_starts ~= num_slots), this flaw should be completely irrelevant. Even
with 10M slots, the chances of a single slot having just 16 (or more)
entries map to it--not enough to cause an FP problem, which would be local
to that slot if it happened--is 1 in millions. This spreadsheet formula
shows that: =1/(10000000*(1 - POISSON(15, 1, TRUE)))
As kUseSmash==false (the setting for Standard128RibbonBitsBuilder) is
intended for CPU efficiency of filters with many more entries/slots than
kCoeffBits, a very reasonable work-around is to disallow num_starts==1
when !kUseSmash, by making the minimum non-zero number of slots
2*kCoeffBits. This is the work-around I've applied. This also means that
the new Ribbon filter schema (Standard128RibbonBitsBuilder) is not
space-efficient for less than a few hundred entries. Because of this, I
have made it fall back on constructing a Bloom filter, under existing
schema, when that is more space efficient for small filters. (We can
change this in the future if we want.)
TODO: better unit tests for this case in ribbon_test, and probably
update StandardHasher for kUseSmash case so that it can scale nicely to
small filters.
### Other related changes
* Add Ribbon filter to stress/crash test
* Add Ribbon filter to filter_bench as -impl=3
* Add option string support, as in "filter_policy=experimental_ribbon:5.678;"
where 5.678 is the Bloom equivalent bits per key.
* Rename internal mode BloomFilterPolicy::kAuto to kAutoBloom
* Add a general BuiltinFilterBitsBuilder::CalculateNumEntry based on
binary searching CalculateSpace (inefficient), so that subclasses
(especially experimental ones) don't have to provide an efficient
implementation inverting CalculateSpace.
* Minor refactor FastLocalBloomBitsBuilder for new base class
XXH3pFilterBitsBuilder shared with new Standard128RibbonBitsBuilder,
which allows the latter to fall back on Bloom construction in some
extreme cases.
* Mostly updated bloom_test for Ribbon filter, though a test like
FullBloomTest::Schema is a next TODO to ensure schema stability
(in case this becomes production-ready schema as it is).
* Add some APIs to ribbon_impl.h for configuring Ribbon filters.
Although these are reasonably covered by bloom_test, TODO more unit
tests in ribbon_test
* Added a "tool" FindOccupancyForSuccessRate to ribbon_test to get data
for constructing the linear approximations in GetNumSlotsFor95PctSuccess.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7658
Test Plan:
Some unit tests updated but other testing is left TODO. This
is considered experimental but laying down schema compatibility as early
as possible in case it proves production-quality. Also tested in
stress/crash test.
Reviewed By: jay-zhuang
Differential Revision: D24899349
Pulled By: pdillinger
fbshipit-source-id: 9715f3e6371c959d923aea8077c9423c7a9f82b8
Summary:
The Customizable class is an extension of the Configurable class and allows instances to be created by a name/ID. Classes that extend customizable can define their Type (e.g. "TableFactory", "Cache") and a method to instantiate them (TableFactory::CreateFromString). Customizable objects can be registered with the ObjectRegistry and created dynamically.
Future PRs will make more types of objects extend Customizable.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6590
Reviewed By: cheng-chang
Differential Revision: D24841553
Pulled By: zhichao-cao
fbshipit-source-id: d0c2132bd932e971cbfe2c908ca2e5db30c5e155
Summary:
The filter query key should not contain timestamp. The timestamp is
stripped for Get(), but not MultiGet().
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7589
Reviewed By: riversand963
Differential Revision: D24494661
Pulled By: jay-zhuang
fbshipit-source-id: fc5ff40f9d683a89a760c6ff0ab3aed05a70c317
Summary:
This is a PR generated **semi-automatically** by an internal tool to remove unused includes and `using` statements.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7604
Test Plan: make check
Reviewed By: ajkr
Differential Revision: D24579392
Pulled By: riversand963
fbshipit-source-id: c4bfa6c6b08da1de186690d37eb73d8fff45aecd
Summary:
Fixes Issue https://github.com/facebook/rocksdb/issues/7497
When allow_data_in_errors db_options is set, log error key details in `ParseInternalKey()`
Have fixed most of the calls. Have few TODOs still pending - because have to make more deeper changes to pass in the allow_data_in_errors flag. Will do those in a separate PR later.
Tests:
- make check
- some of the existing tests that exercise the "internal key too small" condition are: dbformat_test, cuckoo_table_builder_test
- some of the existing tests that exercise the corrupted key path are: corruption_test, merge_helper_test, compaction_iterator_test
Example of new status returns:
- Key too small - `Corrupted Key: Internal Key too small. Size=5`
- Corrupt key with allow_data_in_errors option set to false: `Corrupted Key: '<redacted>' seq:3, type:3`
- Corrupt key with allow_data_in_errors option set to true: `Corrupted Key: '61' seq:3, type:3`
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7515
Reviewed By: ajkr
Differential Revision: D24240264
Pulled By: ramvadiv
fbshipit-source-id: bc48f5d4475ac19d7713e16df37505b31aac42e7
Summary:
This PR does a few things:
1. The MockFileSystem class was split out from the MockEnv. This change would theoretically allow a MockFileSystem to be used by other Environments as well (if we created a means of constructing one). The MockFileSystem implements a FileSystem in its entirety and does not rely on any Wrapper implementation.
2. Make the RocksDB test suite work when MOCK_ENV=1 and ENCRYPTED_ENV=1 are set. To accomplish this, a few things were needed:
- The tests that tried to use the "wrong" environment (Env::Default() instead of env_) were updated
- The MockFileSystem was changed to support the features it was missing or mishandled (such as recursively deleting files in a directory or supporting renaming of a directory).
3. Updated the test framework to have a ROCKSDB_GTEST_SKIP macro. This can be used to flag tests that are skipped. Currently, this defaults to doing nothing (marks the test as SUCCESS) but will mark the tests as SKIPPED when RocksDB is upgraded to a version of gtest that supports this (gtest-1.10).
I have run a full "make check" with MEM_ENV, ENCRYPTED_ENV, both, and neither under both MacOS and RedHat. A few tests were disabled/skipped for the MEM/ENCRYPTED cases. The error_handler_fs_test fails/hangs for MEM_ENV (presumably a timing problem) and I will introduce another PR/issue to track that problem. (I will also push a change to disable those tests soon). There is one more test in DBTest2 that also fails which I need to investigate or skip before this PR is merged.
Theoretically, this PR should also allow the test suite to run against an Env loaded from the registry, though I do not have one to try it with currently.
Finally, once this is accepted, it would be nice if there was a CircleCI job to run these tests on a checkin so this effort does not become stale. I do not know how to do that, so if someone could write that job, it would be appreciated :)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7566
Reviewed By: zhichao-cao
Differential Revision: D24408980
Pulled By: jay-zhuang
fbshipit-source-id: 911b1554a4d0da06fd51feca0c090a4abdcb4a5f
Summary:
This commit makes https://github.com/facebook/rocksdb/issues/6262's code change tidier and easier to understand by:
1. Wrapping parallel compression initialization and termination into
common methods;
2. Wrapping BlockRep initialization, push/pop into common methods;
3. Wrapping file size estimation into common methods;
4. Fixing function declarations that use non-const reference;
5. Fixing some uninitialized variables;
6. Fixing first_block data race;
7. Making BlockRep::status check in BlockBasedTableBuilder::Finish only present
if ok();
8. Making assert(ok()) in BlockBasedTableBuilder::CompressAndVerifyBlock only
present in non-parallel compression mode. In parallel compression mode,
compression will abort if status is not OK;
9. Eliminating potential data race caused by BlockBasedTableBuilder::GetStatus()
and BlockBasedTableBuilder::GetIOStatus() by returning status copy instead of
unprotected reference.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6888
Reviewed By: ajkr
Differential Revision: D21957110
Pulled By: jay-zhuang
fbshipit-source-id: 3a29892f249209513f030349756cecd7736eae80
Summary:
Further refinement of the earlier PR. Now the Status is NotFound with a subcode of PathNotFound. Also the existing functions for options parsing/loading are reverted to return InvalidArgument no matter in which way the user-provided arguments are deemed invalid.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7563
Reviewed By: zhichao-cao
Differential Revision: D24422491
Pulled By: ajkr
fbshipit-source-id: ba6b237cd0584d3f925c5ba0d349aeb8c250af67
Summary:
Ignore read error in 'FilePrefetchBuffer::TryReadFromCache' as status is ignored
and bool value is returned. Return error if prefetch fails in
'PrefetchTail' as we have planned to return Prefetch failures to users.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7574
Test Plan:
make check -j64,
python -u tools/db_crashtest.py blackbox
Reviewed By: zhichao-cao
Differential Revision: D24408825
Pulled By: akankshamahajan15
fbshipit-source-id: feebda771415998253fbe54632f13e6e75b7a243
Summary:
This PR adds support for writing a location identifier of the DB host to SST files as a table property. By default, the hostname is used, but can be overridden by the user. There have been some recent corruptions in files written by ```SstFileWriter``` before checksumming, so this property can be used to trace it back to the writing host and checking the host for hardware isues.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7479
Test Plan: Add new unit tests
Reviewed By: pdillinger
Differential Revision: D24340671
Pulled By: anand1976
fbshipit-source-id: 2038949fd8d160c0633ccb4f9da77740f19fa2a2
Summary:
Right now all I/O failures under
PartitionFilterBlock::CacheDependencies() is swallowed. Return error in
case prefetch fails.
On returning error in PartitionedFilterBlockReader::CacheDependencies was causing stress test failure because PrefetchBuffer is initialized with enable_ = true, as result when PosixMmapReadableFile::Read is called from Prefetch, scratch is ignored causing buffer to fill with garbage values. Initializing prefetch buffer by CreatePrefetchBuffer that sets enable_ with !ioptions.allow_mmap_reads fixed the problem as it returns without prefetching data if allow_mmap_reads is set.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7543
Test Plan:
make check -j64;
python -u tools/db_crashtest.py --simple blackbox
Reviewed By: anand1976
Differential Revision: D24284596
Pulled By: akankshamahajan15
fbshipit-source-id: f3f0fd44b59dcf60645730436f28564a07884868
Summary:
The old flag-based APIs (`BlockBasedTableOptions::pin_l0_filter_and_index_blocks_in_cache` and `BlockBasedTableOptions::pin_top_level_index_and_filter`) were insufficient for our needs. For example, it was impossible to pin only unpartitioned meta-blocks, which could prevent block cache contention when turning on dictionary compression or during a migration to partitioned indexes/filters. It was also impossible to pin all meta-blocks in memory while having predictable memory usage via block cache. If we had continued adding flags to address these scenarios, they would have had significant overlap causing confusion. Instead, this PR deprecates the flags and starts a new API with non-overlapping options.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7520
Test Plan:
- new unit test
- added new options to stress/crash test and ran for a while: `$ python tools/db_crashtest.py blackbox --simple --max_key=1000000 -write_buffer_size=1048576 -target_file_size_base=1048576 -max_bytes_for_level_base=4194304 --interval=10 -value_size_mult=33 -column_families=1 -reopen=0`
Reviewed By: pdillinger
Differential Revision: D24200034
Pulled By: ajkr
fbshipit-source-id: 3fa7cfc71e7960f7a867511dd6ae5834dd73b13e
Summary:
Add following stats for MultiGet in Histogram to get more insight on MultiGet.
1. Number of index and filter blocks read from file as part of MultiGet
request per level.
2. Number of data blocks read from file per level.
3. Number of SST files loaded from file system per level.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7366
Reviewed By: anand1976
Differential Revision: D24127040
Pulled By: akankshamahajan15
fbshipit-source-id: e63a003056b833729b277edc0639c08fb432756b
Summary:
We just used a hacky way to fix db_basic_test: suppress status code in ~BlockBasedTableBuilder. Rather, we should pass them back in Finish() and suppress them in Abandon().
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7507
Test Plan: Watch existing tests to succeed.
Reviewed By: riversand963
Differential Revision: D24119527
fbshipit-source-id: 71c4d4a81c0fd1c5595224692275f20f7759973a
Summary:
Right now all I/O failures under
PartitionFilterBlock::CacheDependencies() is swallowed. Return error in
case prefetch fails.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7463
Test Plan: make check -j64
Reviewed By: anand1976
Differential Revision: D24008226
Pulled By: akankshamahajan15
fbshipit-source-id: b65d63b2d01465db92500b78de7ad58650ec9b3b
Summary:
Add db_basic_test status check list. Some of the warnings are suppressed. It is possible that some of them are due to real bugs.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7452
Test Plan: See CI tests pass.
Reviewed By: zhichao-cao
Differential Revision: D23979764
fbshipit-source-id: 6151570c2a9b931b0fbb3fe939a94b2bd1583cbe
Summary:
This PR merges the functionality of making the ColumnFamilyOptions, TableFactory, and DBOptions into Configurable into a single PR, resolving any merge conflicts
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5753
Reviewed By: ajkr
Differential Revision: D23385030
Pulled By: zhichao-cao
fbshipit-source-id: 8b977a7731556230b9b8c5a081b98e49ee4f160a
Summary:
gcc-4.8 returns error when using the constructor. Not sure if it's a compiler bug/limitation or code issue:
```
table/block_based/block_based_table_reader.cc:3183:67: error: use of deleted function ‘rocksdb::WritableFileStringStreamAdapter::WritableFileStringStreamAdapter(rocksdb::WritableFileStringStreamAdapter&&)’
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7358
Reviewed By: pdillinger
Differential Revision: D23577651
Pulled By: jay-zhuang
fbshipit-source-id: b0197e3d3538da61a6f3866410d88d2047fb9695
Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/7315
Test Plan:
`ASSERT_STATUS_CHECKED=1 make sst_dump_test && ./sst_dump_test`
And manually run `./sst_dump --file=*.sst` before and after the change.
Reviewed By: pdillinger
Differential Revision: D23361669
Pulled By: jay-zhuang
fbshipit-source-id: 5bf51a2a90ee35c8c679e5f604732ec2aef5949a
Summary:
In block-based table builder, the cut-over from buffered to unbuffered
mode involves sampling the buffered blocks and generating a dictionary.
There was a bug where `SstFileWriter` passed zero as the `target_file_size`
causing the cutover to happen immediately, so there were no samples
available for generating the dictionary.
This PR changes the meaning of `target_file_size == 0` to mean buffer
the whole file before cutting over. It also adds dictionary compression
support to `sst_dump --command=recompress` for easy evaluation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7323
Reviewed By: cheng-chang
Differential Revision: D23412158
Pulled By: ajkr
fbshipit-source-id: 3b232050e70ef3c2ee85a4b5f6fadb139c569873
