Summary:
If the block Cache is full with strict_capacity_limit=false,
then our CacheEntryStatsCollector could be immediately evicted on
release, so iterating through column families with shared block cache
could trigger re-scan for each CF. This change fixes that problem by
pinning the CacheEntryStatsCollector from InternalStats so that it's not
evicted.
I had originally thought that this object could participate in LRU like
everything else, but even though a re-load+re-scan only touches memory,
it can be orders of magnitude more expensive than other cache misses.
One service in Facebook has scans that take ~20s over 100GB block cache
that is mostly 4KB entries. (The up-side of this bug and https://github.com/facebook/rocksdb/issues/8369 is that
we had a natural experiment on the effect on some service metrics even
with block cache scans running continuously in the background--a kind
of worst case scenario. Metrics like latency were not affected enough
to trigger warnings.)
Other smaller fixes:
20s is already a sizable portion of 600s stats dump period, or 180s
default max age to force re-scan, so added logic to ensure that (for
each block cache) we don't spend more than 0.2% of our background thread
time scanning it. Nevertheless, "foreground" requests for cache entry
stats (calls to `db->GetMapProperty(DB::Properties::kBlockCacheEntryStats)`)
are permitted to consume more CPU.
Renamed field to cache_entry_stats_ to match code style.
This change is intended for patching in 6.21 release.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8385
Test Plan:
unit test expanded to cover new logic (detect regression),
some manual testing with db_bench
Reviewed By: ajkr
Differential Revision: D29042759
Pulled By: pdillinger
fbshipit-source-id: 236faa902397f50038c618f50fbc8cf3f277308c
Summary:
Recalculate the total size after generate new sst files.
New generated files might have different size as the previous time which
could cause the test failed.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8396
Test Plan:
```
gtest-parallel ./db_compaction_test
--gtest_filter=DBCompactionTest.ManualCompactionMax -r 1000 -w 100
```
Reviewed By: akankshamahajan15
Differential Revision: D29083299
Pulled By: jay-zhuang
fbshipit-source-id: 49d4bd619cefc0f9a1f452f8759ff4c2ba1b6fdb
Summary:
The subcompaction boundary picking logic does not currently guarantee
that all user keys that differ only by timestamp get processed by the same
subcompaction. This can cause issues with the `CompactionIterator` state
machine: for instance, one subcompaction that processes a subset of such KVs
might drop a tombstone based on the KVs it sees, while in reality the
tombstone might not have been eligible to be optimized out.
(See also https://github.com/facebook/rocksdb/issues/6645, which adjusted the way compaction inputs are picked for the
same reason.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8393
Test Plan: Ran `make check` and the crash test script with timestamps enabled.
Reviewed By: jay-zhuang
Differential Revision: D29071635
Pulled By: ltamasi
fbshipit-source-id: f6c72442122b4e581871e096fabe3876a9e8a5a6
Summary:
DBImpl::DumpStats is supposed to do this:
Dump DB stats to LOG
For each CF, dump CFStatsNoFileHistogram to LOG
For each CF, dump CFFileHistogram to LOG
Instead, due to a longstanding bug from 2017 (https://github.com/facebook/rocksdb/issues/2126), it would dump
CFStats, which includes both CFStatsNoFileHistogram and CFFileHistogram,
in both loops, resulting in near-duplicate output.
This fixes the bug.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8380
Test Plan: Manual inspection of LOG after db_bench
Reviewed By: jay-zhuang
Differential Revision: D29017535
Pulled By: pdillinger
fbshipit-source-id: 3010604c4a629a80347f129cd746ce9b0d0cbda6
Summary:
In the current logic, any IO Error with retryable flag == true will be handled by the special logic and in most cases, StartRecoverFromRetryableBGIOError will be called to do the auto resume. If the NoSpace error with retryable flag is set during WAL write, it is mapped as a hard error, which will trigger the auto recovery. During the recover process, if write continues and append to the WAL, the write process sees that bg_error is set to HardError and it calls WriteStatusCheck(), which calls SetBGError() with Status (not IOStatus). This will redirect to the regular SetBGError interface, in which recovery_error_ will be set to the corresponding error. With the recovery_error_ set, the auto resume thread created in StartRecoverFromRetryableBGIOError will keep failing as long as user keeps trying to write.
To fix this issue. All the NoSpace error (no matter retryable flag is set or not) will be redirect to the regular SetBGError, and RecoverFromNoSpace() will do the recovery job which calls SstFileManager::StartErrorRecovery().
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8376
Test Plan: make check and added the new testing case
Reviewed By: anand1976
Differential Revision: D29071828
Pulled By: zhichao-cao
fbshipit-source-id: 7171d7e14cc4620fdab49b7eff7a2fe9a89942c2
Summary:
This PR add support for Merge operation in Integrated BlobDB with base values(i.e DB::Put). Merged values can be retrieved through DB::Get, DB::MultiGet, DB::GetMergeOperands and Iterator operation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8292
Test Plan: Add new unit tests
Reviewed By: ltamasi
Differential Revision: D28415896
Pulled By: akankshamahajan15
fbshipit-source-id: e9b3478bef51d2f214fb88c31ed3c8d2f4a531ff
Summary:
Changed fprintf function to fputc in ApplyVersionEdit, and replaced null characters with whitespaces.
Added unit test in ldb_test.py - verifies that manifest_dump --verbose output is correct when keys and values containing null characters are inserted.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8378
Reviewed By: pdillinger
Differential Revision: D29034584
Pulled By: bjlemaire
fbshipit-source-id: 50833687a8a5f726e247c38457eadc3e6dbab862
Summary:
Currently, we either use the file system inode or a monotonically incrementing runtime ID as the block cache key prefix. However, if we use a monotonically incrementing runtime ID (in the case that the file system does not support inode id generation), in some cases, it cannot ensure uniqueness (e.g., we have secondary cache migrated from host to host). We use DbSessionID (20 bytes) + current file number (at most 10 bytes) as the new cache block key prefix when the secondary cache is enabled. So can accommodate scenarios such as transfer of cache state across hosts.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8360
Test Plan: add the test to lru_cache_test
Reviewed By: pdillinger
Differential Revision: D29006215
Pulled By: zhichao-cao
fbshipit-source-id: 6cff686b38d83904667a2bd39923cd030df16814
Summary:
Logically, subcompactions process a key range [start, end); however, the way
this is currently implemented is that the `CompactionIterator` for any given
subcompaction keeps processing key-values until it actually outputs a key that
is out of range, which is then discarded. Instead of doing this, the patch
introduces a new type of internal iterator called `ClippingIterator` which wraps
another internal iterator and "clips" its range of key-values so that any KVs
returned are strictly in the [start, end) interval. This does eliminate a (minor)
inefficiency by stopping processing in subcompactions exactly at the limit;
however, the main motivation is related to BlobDB: namely, we need this to be
able to measure the amount of garbage generated by a subcompaction
precisely and prevent off-by-one errors.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8327
Test Plan: `make check`
Reviewed By: siying
Differential Revision: D28761541
Pulled By: ltamasi
fbshipit-source-id: ee0e7229f04edabbc7bed5adb51771fbdc287f69
Summary:
In final polishing of https://github.com/facebook/rocksdb/issues/8297 (after most manual testing), I
broke my own caching layer by sanitizing an input parameter with
std::min(0, x) instead of std::max(0, x). I resisted unit testing the
timing part of the result caching because historically, these test
are either flaky or difficult to write, and this was not a correctness
issue. This bug is essentially unnoticeable with a small number
of column families but can explode background work with a
large number of column families.
This change fixes the logical error, removes some unnecessary related
optimization, and adds mock time/sleeps to the unit test to ensure we
can cache hit within the age limit.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8369
Test Plan: added time testing logic to existing unit test
Reviewed By: ajkr
Differential Revision: D28950892
Pulled By: pdillinger
fbshipit-source-id: e79cd4ff3eec68fd0119d994f1ed468c38026c3b
Summary:
Added the ability to cancel an in-progress range compaction by storing to an atomic "canceled" variable pointed to within the CompactRangeOptions structure.
Tested via two tests added to db_tests2.cc.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8351
Reviewed By: ajkr
Differential Revision: D28808894
Pulled By: ddevec
fbshipit-source-id: cb321361c9e23b084b188bb203f11c375a22c2dd
Summary:
This is a duplicate of https://github.com/facebook/rocksdb/issues/4948 by mzhaom to fix tests after rebase.
This change is a follow-up to https://github.com/facebook/rocksdb/issues/4927, which made this possible by allowing tombstone dropping/seqnum zeroing optimizations on the last key in the compaction. Now the `largest_seqno != 0` condition suffices to prevent snapshot release triggered compaction from entering an infinite loop.
The issues caused by the extraneous condition `level_and_file.second->num_deletions > 1` are:
- files could have `largest_seqno > 0` forever making it impossible to tell they cannot contain any covering keys
- it doesn't trigger compaction when there are many overwritten keys. Some MyRocks use case actually doesn't use Delete but instead calls Put with empty value to "delete" keys, so we'd like to be able to trigger compaction in this case too.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8357
Test Plan: - make check
Reviewed By: jay-zhuang
Differential Revision: D28855340
Pulled By: ajkr
fbshipit-source-id: a261b51eecafec492499e6d01e8e43112f801798
Summary:
I noticed ```openat``` system call with ```O_WRONLY``` flag and ```sync_file_range``` and ```truncate``` on WAL file when using ```rocksdb::DB::OpenForReadOnly``` by way of ```db_bench --readonly=true --benchmarks=readseq --use_existing_db=1 --num=1 ...```
Noticed in ```strace``` after seeing the last modification time of the WAL file change after each run (with ```--readonly=true```).
I think introduced by 7d7f14480e from https://github.com/facebook/rocksdb/pull/8122
I added a test to catch the WAL file being truncated and the modification time on it changing.
I am not sure if a mock filesystem with mock clock could be used to avoid having to sleep 1.1s.
The test could also check the set of files is the same and that the sizes are also unchanged.
Before:
```
[ RUN ] DBBasicTest.ReadOnlyReopenMtimeUnchanged
db/db_basic_test.cc:182: Failure
Expected equality of these values:
file_mtime_after_readonly_reopen
Which is: 1621611136
file_mtime_before_readonly_reopen
Which is: 1621611135
file is: 000010.log
[ FAILED ] DBBasicTest.ReadOnlyReopenMtimeUnchanged (1108 ms)
```
After:
```
[ RUN ] DBBasicTest.ReadOnlyReopenMtimeUnchanged
[ OK ] DBBasicTest.ReadOnlyReopenMtimeUnchanged (1108 ms)
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8313
Reviewed By: pdillinger
Differential Revision: D28656925
Pulled By: jay-zhuang
fbshipit-source-id: ea9e215cb53e7c830e76bc5fc75c45e21f12a1d6
Summary:
https://github.com/facebook/rocksdb/pull/8288 introduces a bug: SequenceIterWrapper should do next for seek key using internal key comparator rather than user comparator. Fix it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8328
Test Plan: Pass all existing tests
Reviewed By: ltamasi
Differential Revision: D28647263
fbshipit-source-id: 4081d684fd8a86d248c485ef8a1563c7af136447
Summary:
Error:
```
db/db_compaction_test.cc:5211:47: warning: The left operand of '*' is a garbage value
uint64_t total = (l1_avg_size + l2_avg_size * 10) * 10;
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8325
Test Plan: `$ make analyze`
Reviewed By: pdillinger
Differential Revision: D28620916
Pulled By: jay-zhuang
fbshipit-source-id: f6d58ab84eefbcc905cda45afb9522b0c6d230f8
Summary:
With Ribbon filter work and possible variance in actual bits
per key (or prefix; general term "entry") to achieve certain FP rates,
I've received a request to be able to track actual bits per key in
generated filters. This change adds a num_filter_entries table
property, which can be combined with filter_size to get bits per key
(entry).
This can vary from num_entries in at least these ways:
* Different versions of same key are only counted once in filters.
* With prefix filters, several user keys map to the same filter entry.
* A single filter can include both prefixes and user keys.
Note that FilterBlockBuilder::NumAdded() didn't do anything useful
except distinguish empty from non-empty.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8323
Test Plan: basic unit test included, others updated
Reviewed By: jay-zhuang
Differential Revision: D28596210
Pulled By: pdillinger
fbshipit-source-id: 529a111f3c84501e5a470bc84705e436ee68c376
Summary:
Fix a bug that for manual compaction, `max_compaction_bytes` is only
limit the SST files from input level, but not overlapped files on output
level.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8269
Test Plan: `make check`
Reviewed By: ajkr
Differential Revision: D28231044
Pulled By: jay-zhuang
fbshipit-source-id: 9d7d03004f30cc4b1b9819830141436907554b7c
Summary:
When a memtable is flushed, it will validate number of entries it reads, and compare the number with how many entries inserted into memtable. This serves as one sanity c\
heck against memory corruption. This change will also allow more counters to be added in the future for better validation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8288
Test Plan: Pass all existing tests
Reviewed By: ajkr
Differential Revision: D28369194
fbshipit-source-id: 7ff870380c41eab7f99eee508550dcdce32838ad
Summary:
The test want to make sure these's no compaction during `AddFile`
(between `DBImpl::AddFile:MutexLock` and `DBImpl::AddFile:MutexUnlock`)
but the mutex could be unlocked by `EnterUnbatched()`.
Move the lock start point after bumping the ingest file number.
Also fix the dead lock when ASSERT fails.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8307
Reviewed By: ajkr
Differential Revision: D28479849
Pulled By: jay-zhuang
fbshipit-source-id: b3c50f66aa5d5f59c5c27f815bfea189c4cd06cb
Summary:
This change gathers and publishes statistics about the
kinds of items in block cache. This is especially important for
profiling relative usage of cache by index vs. filter vs. data blocks.
It works by iterating over the cache during periodic stats dump
(InternalStats, stats_dump_period_sec) or on demand when
DB::Get(Map)Property(kBlockCacheEntryStats), except that for
efficiency and sharing among column families, saved data from
the last scan is used when the data is not considered too old.
The new information can be seen in info LOG, for example:
Block cache LRUCache@0x7fca62229330 capacity: 95.37 MB collections: 8 last_copies: 0 last_secs: 0.00178 secs_since: 0
Block cache entry stats(count,size,portion): DataBlock(7092,28.24 MB,29.6136%) FilterBlock(215,867.90 KB,0.888728%) FilterMetaBlock(2,5.31 KB,0.00544%) IndexBlock(217,180.11 KB,0.184432%) WriteBuffer(1,256.00 KB,0.262144%) Misc(1,0.00 KB,0%)
And also through DB::GetProperty and GetMapProperty (here using
ldb just for demonstration):
$ ./ldb --db=/dev/shm/dbbench/ get_property rocksdb.block-cache-entry-stats
rocksdb.block-cache-entry-stats.bytes.data-block: 0
rocksdb.block-cache-entry-stats.bytes.deprecated-filter-block: 0
rocksdb.block-cache-entry-stats.bytes.filter-block: 0
rocksdb.block-cache-entry-stats.bytes.filter-meta-block: 0
rocksdb.block-cache-entry-stats.bytes.index-block: 178992
rocksdb.block-cache-entry-stats.bytes.misc: 0
rocksdb.block-cache-entry-stats.bytes.other-block: 0
rocksdb.block-cache-entry-stats.bytes.write-buffer: 0
rocksdb.block-cache-entry-stats.capacity: 8388608
rocksdb.block-cache-entry-stats.count.data-block: 0
rocksdb.block-cache-entry-stats.count.deprecated-filter-block: 0
rocksdb.block-cache-entry-stats.count.filter-block: 0
rocksdb.block-cache-entry-stats.count.filter-meta-block: 0
rocksdb.block-cache-entry-stats.count.index-block: 215
rocksdb.block-cache-entry-stats.count.misc: 1
rocksdb.block-cache-entry-stats.count.other-block: 0
rocksdb.block-cache-entry-stats.count.write-buffer: 0
rocksdb.block-cache-entry-stats.id: LRUCache@0x7f3636661290
rocksdb.block-cache-entry-stats.percent.data-block: 0.000000
rocksdb.block-cache-entry-stats.percent.deprecated-filter-block: 0.000000
rocksdb.block-cache-entry-stats.percent.filter-block: 0.000000
rocksdb.block-cache-entry-stats.percent.filter-meta-block: 0.000000
rocksdb.block-cache-entry-stats.percent.index-block: 2.133751
rocksdb.block-cache-entry-stats.percent.misc: 0.000000
rocksdb.block-cache-entry-stats.percent.other-block: 0.000000
rocksdb.block-cache-entry-stats.percent.write-buffer: 0.000000
rocksdb.block-cache-entry-stats.secs_for_last_collection: 0.000052
rocksdb.block-cache-entry-stats.secs_since_last_collection: 0
Solution detail - We need some way to flag what kind of blocks each
entry belongs to, preferably without changing the Cache API.
One of the complications is that Cache is a general interface that could
have other users that don't adhere to whichever convention we decide
on for keys and values. Or we would pay for an extra field in the Handle
that would only be used for this purpose.
This change uses a back-door approach, the deleter, to indicate the
"role" of a Cache entry (in addition to the value type, implicitly).
This has the added benefit of ensuring proper code origin whenever we
recognize a particular role for a cache entry; if the entry came from
some other part of the code, it will use an unrecognized deleter, which
we simply attribute to the "Misc" role.
An internal API makes for simple instantiation and automatic
registration of Cache deleters for a given value type and "role".
Another internal API, CacheEntryStatsCollector, solves the problem of
caching the results of a scan and sharing them, to ensure scans are
neither excessive nor redundant so as not to harm Cache performance.
Because code is added to BlocklikeTraits, it is pulled out of
block_based_table_reader.cc into its own file.
This is a reformulation of https://github.com/facebook/rocksdb/issues/8276, without the type checking option
(could still be added), and with actual stat gathering.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8297
Test Plan: manual testing with db_bench, and a couple of basic unit tests
Reviewed By: ltamasi
Differential Revision: D28488721
Pulled By: pdillinger
fbshipit-source-id: 472f524a9691b5afb107934be2d41d84f2b129fb
Summary:
Some file systems (especially distributed FS) do not support reopening a file for writing. The ExternalSstFileIngestionJob calls ReopenWritableFile in order to sync the ingested file, which typically makes sense only on a local file system with a page cache (i.e Posix). So this change tries to sync the ingested file only if ReopenWritableFile doesn't return Status::NotSupported().
Tests:
Add a new unit test in external_sst_file_basic_test
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8296
Reviewed By: jay-zhuang
Differential Revision: D28420865
Pulled By: anand1976
fbshipit-source-id: 380e7f5ff95324997f7a59864a9ac96ebbd0100c
Summary:
This patch does two things:
1) Introduces some aliases in order to eliminate/prevent long-winded type names
w/r/t the internal table property collectors (see e.g.
`std::vector<std::unique_ptr<IntTblPropCollectorFactory>>`).
2) Makes it possible to apply only a subrange of table property collectors during
table building by turning `TableBuilderOptions::int_tbl_prop_collector_factories`
from a pointer to a `vector` into a range (i.e. a pair of iterators).
Rationale: I plan to introduce a BlobDB related table property collector, which
should only be applied during table creation if blob storage is enabled at the moment
(which can be changed dynamically). This change will make it possible to include/
exclude the BlobDB related collector as needed without having to introduce
a second `vector` of collectors in `ColumnFamilyData` with pretty much the same
contents.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8298
Test Plan: `make check`
Reviewed By: jay-zhuang
Differential Revision: D28430910
Pulled By: ltamasi
fbshipit-source-id: a81d28f2c59495865300f43deb2257d2e6977c8e
Summary:
As a part of tiered storage, writing tempeature information to manifest is needed so that after DB recovery, RocksDB still has the tiering information, to implement some further necessary functionalities.
Also fix some issues in simulated hybrid FS.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8284
Test Plan: Add a new unit test to validate that the information is indeed written and read back.
Reviewed By: zhichao-cao
Differential Revision: D28335801
fbshipit-source-id: 56aeb2e6ea090be0200181dd968c8a7278037def
Summary:
Defined the abstract interface for a secondary cache in include/rocksdb/secondary_cache.h, and updated LRUCacheOptions to take a std::shared_ptr<SecondaryCache>. An item is initially inserted into the LRU (primary) cache. When it ages out and evicted from memory, its inserted into the secondary cache. On a LRU cache miss and successful lookup in the secondary cache, the item is promoted to the LRU cache. Only support synchronous lookup currently. The secondary cache would be used to implement a persistent (flash cache) or compressed cache.
Tests:
Results from cache_bench and db_bench don't show any regression due to these changes.
cache_bench results before and after this change -
Command
```./cache_bench -ops_per_thread=10000000 -threads=1```
Before
```Complete in 40.688 s; QPS = 245774```
```Complete in 40.486 s; QPS = 246996```
```Complete in 42.019 s; QPS = 237989```
After
```Complete in 40.672 s; QPS = 245869```
```Complete in 44.622 s; QPS = 224107```
```Complete in 42.445 s; QPS = 235599```
db_bench results before this change, and with this change + https://github.com/facebook/rocksdb/issues/8213 and https://github.com/facebook/rocksdb/issues/8191 -
Commands
```./db_bench --benchmarks="fillseq,compact" -num=30000000 -key_size=32 -value_size=256 -use_direct_io_for_flush_and_compaction=true -db=/home/anand76/nvm_cache/db -partition_index_and_filters=true```
```./db_bench -db=/home/anand76/nvm_cache/db -use_existing_db=true -benchmarks=readrandom -num=30000000 -key_size=32 -value_size=256 -use_direct_reads=true -cache_size=1073741824 -cache_numshardbits=6 -cache_index_and_filter_blocks=true -read_random_exp_range=17 -statistics -partition_index_and_filters=true -threads=16 -duration=300```
Before
```
DB path: [/home/anand76/nvm_cache/db]
readrandom : 80.702 micros/op 198104 ops/sec; 54.4 MB/s (3708999 of 3708999 found)
```
```
DB path: [/home/anand76/nvm_cache/db]
readrandom : 87.124 micros/op 183625 ops/sec; 50.4 MB/s (3439999 of 3439999 found)
```
After
```
DB path: [/home/anand76/nvm_cache/db]
readrandom : 77.653 micros/op 206025 ops/sec; 56.6 MB/s (3866999 of 3866999 found)
```
```
DB path: [/home/anand76/nvm_cache/db]
readrandom : 84.962 micros/op 188299 ops/sec; 51.7 MB/s (3535999 of 3535999 found)
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8271
Reviewed By: zhichao-cao
Differential Revision: D28357511
Pulled By: anand1976
fbshipit-source-id: d1cfa236f00e649a18c53328be10a8062a4b6da2
Summary:
The functions will be used for remote compaction parameter
input and result.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8247
Test Plan: `make check`
Reviewed By: ajkr
Differential Revision: D28104680
Pulled By: jay-zhuang
fbshipit-source-id: c0a5178e6277125118384278efea2acbf90aa6cb
Summary:
Adds a new Cache::ApplyToAllEntries API that we expect to use
(in follow-up PRs) for efficiently gathering block cache statistics.
Notable features vs. old ApplyToAllCacheEntries:
* Includes key and deleter (in addition to value and charge). We could
have passed in a Handle but then more virtual function calls would be
needed to get the "fields" of each entry. We expect to use the 'deleter'
to identify the origin of entries, perhaps even more.
* Heavily tuned to minimize latency impact on operating cache. It
does this by iterating over small sections of each cache shard while
cycling through the shards.
* Supports tuning roughly how many entries to operate on for each
lock acquire and release, to control the impact on the latency of other
operations without excessive lock acquire & release. The right balance
can depend on the cost of the callback. Good default seems to be
around 256.
* There should be no need to disable thread safety. (I would expect
uncontended locks to be sufficiently fast.)
I have enhanced cache_bench to validate this approach:
* Reports a histogram of ns per operation, so we can look at the
ditribution of times, not just throughput (average).
* Can add a thread for simulated "gather stats" which calls
ApplyToAllEntries at a specified interval. We also generate a histogram
of time to run ApplyToAllEntries.
To make the iteration over some entries of each shard work as cleanly as
possible, even with resize between next set of entries, I have
re-arranged which hash bits are used for sharding and which for indexing
within a shard.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8225
Test Plan:
A couple of unit tests are added, but primary validation is manual, as
the primary risk is to performance.
The primary validation is using cache_bench to ensure that neither
the minor hashing changes nor the simulated stats gathering
significantly impact QPS or latency distribution. Note that adding op
latency histogram seriously impacts the benchmark QPS, so for a
fair baseline, we need the cache_bench changes (except remove simulated
stat gathering to make it compile). In short, we don't see any
reproducible difference in ops/sec or op latency unless we are gathering
stats nearly continuously. Test uses 10GB block cache with
8KB values to be somewhat realistic in the number of items to iterate
over.
Baseline typical output:
```
Complete in 92.017 s; Rough parallel ops/sec = 869401
Thread ops/sec = 54662
Operation latency (ns):
Count: 80000000 Average: 11223.9494 StdDev: 29.61
Min: 0 Median: 7759.3973 Max: 9620500
Percentiles: P50: 7759.40 P75: 14190.73 P99: 46922.75 P99.9: 77509.84 P99.99: 217030.58
------------------------------------------------------
[ 0, 1 ] 68 0.000% 0.000%
( 2900, 4400 ] 89 0.000% 0.000%
( 4400, 6600 ] 33630240 42.038% 42.038% ########
( 6600, 9900 ] 18129842 22.662% 64.700% #####
( 9900, 14000 ] 7877533 9.847% 74.547% ##
( 14000, 22000 ] 15193238 18.992% 93.539% ####
( 22000, 33000 ] 3037061 3.796% 97.335% #
( 33000, 50000 ] 1626316 2.033% 99.368%
( 50000, 75000 ] 421532 0.527% 99.895%
( 75000, 110000 ] 56910 0.071% 99.966%
( 110000, 170000 ] 16134 0.020% 99.986%
( 170000, 250000 ] 5166 0.006% 99.993%
( 250000, 380000 ] 3017 0.004% 99.996%
( 380000, 570000 ] 1337 0.002% 99.998%
( 570000, 860000 ] 805 0.001% 99.999%
( 860000, 1200000 ] 319 0.000% 100.000%
( 1200000, 1900000 ] 231 0.000% 100.000%
( 1900000, 2900000 ] 100 0.000% 100.000%
( 2900000, 4300000 ] 39 0.000% 100.000%
( 4300000, 6500000 ] 16 0.000% 100.000%
( 6500000, 9800000 ] 7 0.000% 100.000%
```
New, gather_stats=false. Median thread ops/sec of 5 runs:
```
Complete in 92.030 s; Rough parallel ops/sec = 869285
Thread ops/sec = 54458
Operation latency (ns):
Count: 80000000 Average: 11298.1027 StdDev: 42.18
Min: 0 Median: 7722.0822 Max: 6398720
Percentiles: P50: 7722.08 P75: 14294.68 P99: 47522.95 P99.9: 85292.16 P99.99: 228077.78
------------------------------------------------------
[ 0, 1 ] 109 0.000% 0.000%
( 2900, 4400 ] 793 0.001% 0.001%
( 4400, 6600 ] 34054563 42.568% 42.569% #########
( 6600, 9900 ] 17482646 21.853% 64.423% ####
( 9900, 14000 ] 7908180 9.885% 74.308% ##
( 14000, 22000 ] 15032072 18.790% 93.098% ####
( 22000, 33000 ] 3237834 4.047% 97.145% #
( 33000, 50000 ] 1736882 2.171% 99.316%
( 50000, 75000 ] 446851 0.559% 99.875%
( 75000, 110000 ] 68251 0.085% 99.960%
( 110000, 170000 ] 18592 0.023% 99.983%
( 170000, 250000 ] 7200 0.009% 99.992%
( 250000, 380000 ] 3334 0.004% 99.997%
( 380000, 570000 ] 1393 0.002% 99.998%
( 570000, 860000 ] 700 0.001% 99.999%
( 860000, 1200000 ] 293 0.000% 100.000%
( 1200000, 1900000 ] 196 0.000% 100.000%
( 1900000, 2900000 ] 69 0.000% 100.000%
( 2900000, 4300000 ] 32 0.000% 100.000%
( 4300000, 6500000 ] 10 0.000% 100.000%
```
New, gather_stats=true, 1 second delay between scans. Scans take about
1 second here so it's spending about 50% time scanning. Still the effect on
ops/sec and latency seems to be in the noise. Median thread ops/sec of 5 runs:
```
Complete in 91.890 s; Rough parallel ops/sec = 870608
Thread ops/sec = 54551
Operation latency (ns):
Count: 80000000 Average: 11311.2629 StdDev: 45.28
Min: 0 Median: 7686.5458 Max: 10018340
Percentiles: P50: 7686.55 P75: 14481.95 P99: 47232.60 P99.9: 79230.18 P99.99: 232998.86
------------------------------------------------------
[ 0, 1 ] 71 0.000% 0.000%
( 2900, 4400 ] 291 0.000% 0.000%
( 4400, 6600 ] 34492060 43.115% 43.116% #########
( 6600, 9900 ] 16727328 20.909% 64.025% ####
( 9900, 14000 ] 7845828 9.807% 73.832% ##
( 14000, 22000 ] 15510654 19.388% 93.220% ####
( 22000, 33000 ] 3216533 4.021% 97.241% #
( 33000, 50000 ] 1680859 2.101% 99.342%
( 50000, 75000 ] 439059 0.549% 99.891%
( 75000, 110000 ] 60540 0.076% 99.967%
( 110000, 170000 ] 14649 0.018% 99.985%
( 170000, 250000 ] 5242 0.007% 99.991%
( 250000, 380000 ] 3260 0.004% 99.995%
( 380000, 570000 ] 1599 0.002% 99.997%
( 570000, 860000 ] 1043 0.001% 99.999%
( 860000, 1200000 ] 471 0.001% 99.999%
( 1200000, 1900000 ] 275 0.000% 100.000%
( 1900000, 2900000 ] 143 0.000% 100.000%
( 2900000, 4300000 ] 60 0.000% 100.000%
( 4300000, 6500000 ] 27 0.000% 100.000%
( 6500000, 9800000 ] 7 0.000% 100.000%
( 9800000, 14000000 ] 1 0.000% 100.000%
Gather stats latency (us):
Count: 46 Average: 980387.5870 StdDev: 60911.18
Min: 879155 Median: 1033777.7778 Max: 1261431
Percentiles: P50: 1033777.78 P75: 1120666.67 P99: 1261431.00 P99.9: 1261431.00 P99.99: 1261431.00
------------------------------------------------------
( 860000, 1200000 ] 45 97.826% 97.826% ####################
( 1200000, 1900000 ] 1 2.174% 100.000%
Most recent cache entry stats:
Number of entries: 1295133
Total charge: 9.88 GB
Average key size: 23.4982
Average charge: 8.00 KB
Unique deleters: 3
```
Reviewed By: mrambacher
Differential Revision: D28295742
Pulled By: pdillinger
fbshipit-source-id: bbc4a552f91ba0fe10e5cc025c42cef5a81f2b95
Summary:
This change enables a couple of things:
- Different ConfigOptions can have different registry/factory associated with it, thereby allowing things like a "Test" ConfigOptions versus a "Production"
- The ObjectRegistry is created fewer times and can be re-used
The ConfigOptions can also be initialized/constructed from a DBOptions, in which case it will grab some of its settings (Env, Logger) from the DBOptions.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8166
Reviewed By: zhichao-cao
Differential Revision: D27657952
Pulled By: mrambacher
fbshipit-source-id: ae1d6200bb7ab127405cdeefaba43c7fe694dfdd
Summary:
The WBWI has two differing modes of operation dependent on the value
of the constructor parameter `overwrite_key`.
Currently, regardless of the parameter, neither mode performs as
expected when using Merge. This PR remedies this by correctly invoking
the appropriate Merge Operator before returning results from the WBWI.
Examples of issues that exist which are solved by this PR:
## Example 1 with `overwrite_key=false`
Currently, from an empty database, the following sequence:
```
Put('k1', 'v1')
Merge('k1', 'v2')
Get('k1')
```
Incorrectly yields `v2`, that is to say that the Merge behaves like a Put.
## Example 2 with o`verwrite_key=true`
Currently, from an empty database, the following sequence:
```
Put('k1', 'v1')
Merge('k1', 'v2')
Get('k1')
```
Incorrectly yields `ERROR: kMergeInProgress`.
## Example 3 with `overwrite_key=false`
Currently, with a database containing `('k1' -> 'v1')`, the following sequence:
```
Merge('k1', 'v2')
GetFromBatchAndDB('k1')
```
Incorrectly yields `v1,v2`
## Example 4 with `overwrite_key=true`
Currently, with a database containing `('k1' -> 'v1')`, the following sequence:
```
Merge('k1', 'v1')
GetFromBatchAndDB('k1')
```
Incorrectly yields `ERROR: kMergeInProgress`.
## Example 5 with `overwrite_key=false`
Currently, from an empty database, the following sequence:
```
Put('k1', 'v1')
Merge('k1', 'v2')
GetFromBatchAndDB('k1')
```
Incorrectly yields `v1,v2`
## Example 6 with `overwrite_key=true`
Currently, from an empty database, `('k1' -> 'v1')`, the following sequence:
```
Put('k1', 'v1')
Merge('k1', 'v2')
GetFromBatchAndDB('k1')
```
Incorrectly yields `ERROR: kMergeInProgress`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8135
Reviewed By: pdillinger
Differential Revision: D27657938
Pulled By: mrambacher
fbshipit-source-id: 0fbda6bbc66bedeba96a84786d90141d776297df
Summary:
From HISTORY.md release note:
- Allow `CompactionFilter`s to apply in more table file creation scenarios such as flush and recovery. For compatibility, `CompactionFilter`s by default apply during compaction. Users can customize this behavior by overriding `CompactionFilterFactory::ShouldFilterTableFileCreation()`.
- Removed unused structure `CompactionFilterContext`
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8243
Test Plan: added unit tests
Reviewed By: pdillinger
Differential Revision: D28088089
Pulled By: ajkr
fbshipit-source-id: 0799be7908e3b39fea09fc3f1ab00e13ad817fae
Summary:
Larger arena block size does provide the benefit of reducing allocation overhead, however it may cause other troubles. For example, allocator is more likely not to allocate them to physical memory and trigger page fault. Weighing the risk, we cap the arena block size to 1MB. Users can always use a larger value if they want.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7907
Test Plan: Run all existing tests
Reviewed By: pdillinger
Differential Revision: D26135269
fbshipit-source-id: b7f55afd03e6ee1d8715f90fa11b6c33944e9ea8
Summary:
Refactor kill point to one single class, rather than several extern variables. The intention was to drop unflushed data before killing to simulate some job, and I tried to a pointer to fault ingestion fs to the killing class, but it ended up with harder than I thought. Perhaps we'll need to do this in another way. But I thought the refactoring itself is good so I send it out.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8241
Test Plan: make release and run crash test for a while.
Reviewed By: anand1976
Differential Revision: D28078486
fbshipit-source-id: f9182c1455f52e6851c13f88a21bade63bcec45f
Summary:
The ImmutableCFOptions contained a bunch of fields that belonged to the ImmutableDBOptions. This change cleans that up by introducing an ImmutableOptions struct. Following the pattern of Options struct, this class inherits from the DB and CFOption structs (of the Immutable form).
Only one structural change (the ImmutableCFOptions::fs was changed to a shared_ptr from a raw one) is in this PR. All of the other changes involve moving the member variables from the ImmutableCFOptions into the ImmutableOptions and changing member variables or function parameters as required for compilation purposes.
Follow-on PRs may do a further clean-up of the code, such as renaming variables (such as "ImmutableOptions cf_options") and potentially eliminating un-needed function parameters (there is no longer a need to pass both an ImmutableDBOptions and an ImmutableOptions to a function).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8262
Reviewed By: pdillinger
Differential Revision: D28226540
Pulled By: mrambacher
fbshipit-source-id: 18ae71eadc879dedbe38b1eb8e6f9ff5c7147dbf
Summary:
Previously the shutdown process did not properly wait for all
`compaction_thread_limiter` tokens to be released before proceeding to
delete the DB's C++ objects. When this happened, we saw tests like
"DBCompactionTest.CompactionLimiter" flake with the following error:
```
virtual
rocksdb::ConcurrentTaskLimiterImpl::~ConcurrentTaskLimiterImpl():
Assertion `outstanding_tasks_ == 0' failed.
```
There is a case where a token can still be alive even after the shutdown
process has waited for BG work to complete. In particular, this happens
because the shutdown process only waits for flush/compaction scheduled/unscheduled counters to all
reach zero. These counters are decremented in `BackgroundCallCompaction()`
functions. However, tokens are released in `BGWork*Compaction()` functions, which
actually wrap the `BackgroundCallCompaction()` function.
A simple sleep could repro the race condition:
```
$ diff --git a/db/db_impl/db_impl_compaction_flush.cc
b/db/db_impl/db_impl_compaction_flush.cc
index 806bc548a..ba59efa89 100644
--- a/db/db_impl/db_impl_compaction_flush.cc
+++ b/db/db_impl/db_impl_compaction_flush.cc
@@ -2442,6 +2442,7 @@ void DBImpl::BGWorkCompaction(void* arg) {
static_cast<PrepickedCompaction*>(ca.prepicked_compaction);
static_cast_with_check<DBImpl>(ca.db)->BackgroundCallCompaction(
prepicked_compaction, Env::Priority::LOW);
+ sleep(1);
delete prepicked_compaction;
}
$ ./db_compaction_test --gtest_filter=DBCompactionTest.CompactionLimiter
db_compaction_test: util/concurrent_task_limiter_impl.cc:24: virtual rocksdb::ConcurrentTaskLimiterImpl::~ConcurrentTaskLimiterImpl(): Assertion `outstanding_tasks_ == 0' failed.
Received signal 6 (Aborted)
#0 /usr/local/fbcode/platform007/lib/libc.so.6(gsignal+0xcf) [0x7f02673c30ff] ?? ??:0
https://github.com/facebook/rocksdb/issues/1 /usr/local/fbcode/platform007/lib/libc.so.6(abort+0x134) [0x7f02673ac934] ?? ??:0
...
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8253
Test Plan: sleeps to expose race conditions
Reviewed By: akankshamahajan15
Differential Revision: D28168064
Pulled By: ajkr
fbshipit-source-id: 9e5167c74398d323e7975980c5cc00f450631160
Summary:
Previously we saw flakes on platforms like arm on CircleCI, such as the following:
```
Note: Google Test filter = DBTest.L0L1L2AndUpHitCounter
[==========] Running 1 test from 1 test case.
[----------] Global test environment set-up.
[----------] 1 test from DBTest
[ RUN ] DBTest.L0L1L2AndUpHitCounter
db/db_test.cc:5345: Failure
Expected: (TestGetTickerCount(options, GET_HIT_L0)) > (100), actual: 30 vs 100
[ FAILED ] DBTest.L0L1L2AndUpHitCounter (150 ms)
[----------] 1 test from DBTest (150 ms total)
[----------] Global test environment tear-down
[==========] 1 test from 1 test case ran. (150 ms total)
[ PASSED ] 0 tests.
[ FAILED ] 1 test, listed below:
[ FAILED ] DBTest.L0L1L2AndUpHitCounter
```
The test was totally non-deterministic, e.g., flush/compaction timing would affect how many files on each level. Furthermore, it depended heavily on platform-specific details, e.g., by having a 32KB memtable, it could become full with a very different number of entries depending on the platform.
This PR rewrites the test to build a deterministic LSM with one file per level.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8259
Reviewed By: mrambacher
Differential Revision: D28178100
Pulled By: ajkr
fbshipit-source-id: 0a03b26e8d23c29d8297c1bccb1b115dce33bdcd
Summary:
As the first part of the effort of having placing different files on different storage types, this change introduces several things:
(1) An experimental interface in FileSystem that specify temperature to a new file created.
(2) A test FileSystemWrapper, SimulatedHybridFileSystem, that simulates HDD for a file of "warm" temperature.
(3) A simple experimental feature ColumnFamilyOptions.bottommost_temperature. RocksDB would pass this value to FileSystem when creating any bottommost file.
(4) A db_bench parameter that applies the (2) and (3) to db_bench.
The motivation of the change is to introduce minimal changes that allow us to evolve tiered storage development.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8222
Test Plan:
./db_bench --benchmarks=fillrandom --write_buffer_size=2000000 -max_bytes_for_level_base=20000000 -level_compaction_dynamic_level_bytes --reads=100 -compaction_readahead_size=20000000 --reads=100000 -num=10000000
followed by
./db_bench --benchmarks=readrandom,stats --write_buffer_size=2000000 -max_bytes_for_level_base=20000000 -simulate_hybrid_fs_file=/tmp/warm_file_list -level_compaction_dynamic_level_bytes -compaction_readahead_size=20000000 --reads=500 --threads=16 -use_existing_db --num=10000000
and see results as expected.
Reviewed By: ajkr
Differential Revision: D28003028
fbshipit-source-id: 4724896d5205730227ba2f17c3fecb11261744ce
Summary:
Add `num_levels`, `is_bottommost`, and table file creation
`reason` to `FilterBuildingContext`, in anticipation of more powerful
Bloom-like filter support.
To support this, added `is_bottommost` and `reason` to
`TableBuilderOptions`, which allowed removing `reason` parameter from
`rocksdb::BuildTable`.
I attempted to remove `skip_filters` from `TableBuilderOptions`, because
filter construction decisions should arise from options, not one-off
parameters. I could not completely remove it because the public API for
SstFileWriter takes a `skip_filters` parameter, and translating this
into an option change would mean awkwardly replacing the table_factory
if it is BlockBasedTableFactory with new filter_policy=nullptr option.
I marked this public skip_filters option as deprecated because of this
oddity. (skip_filters on the read side probably makes sense.)
At least `skip_filters` is now largely hidden for users of
`TableBuilderOptions` and is no longer used for implementing the
optimize_filters_for_hits option. Bringing the logic for that option
closer to handling of FilterBuildingContext makes it more obvious that
hese two are using the same notion of "bottommost." (Planned:
configuration options for Bloom-like filters that generalize
`optimize_filters_for_hits`)
Recommended follow-up: Try to get away from "bottommost level" naming of
things, which is inaccurate (see
VersionStorageInfo::RangeMightExistAfterSortedRun), and move to
"bottommost run" or just "bottommost."
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8246
Test Plan:
extended an existing unit test to exercise and check various
filter building contexts. Also, existing tests for
optimize_filters_for_hits validate some of the "bottommost" handling,
which is now closely connected to FilterBuildingContext::is_bottommost
through TableBuilderOptions::is_bottommost
Reviewed By: mrambacher
Differential Revision: D28099346
Pulled By: pdillinger
fbshipit-source-id: 2c1072e29c24d4ac404c761a7b7663292372600a
Summary:
Greatly reduced the not-quite-copy-paste giant parameter lists
of rocksdb::NewTableBuilder, rocksdb::BuildTable,
BlockBasedTableBuilder::Rep ctor, and BlockBasedTableBuilder ctor.
Moved weird separate parameter `uint32_t column_family_id` of
TableFactory::NewTableBuilder into TableBuilderOptions.
Re-ordered parameters to TableBuilderOptions ctor, so that `uint64_t
target_file_size` is not randomly placed between uint64_t timestamps
(was easy to mix up).
Replaced a couple of fields of BlockBasedTableBuilder::Rep with a
FilterBuildingContext. The motivation for this change is making it
easier to pass along more data into new fields in FilterBuildingContext
(follow-up PR).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8240
Test Plan: ASAN make check
Reviewed By: mrambacher
Differential Revision: D28075891
Pulled By: pdillinger
fbshipit-source-id: fddb3dbb8260a0e8bdcbb51b877ebabf9a690d4f
Summary:
Don't call ```rocksdb_cache_disown_data()``` as it causes the memory allocated for ```shards_``` to be leaked.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8237
Reviewed By: jay-zhuang
Differential Revision: D28039061
Pulled By: anand1976
fbshipit-source-id: c3464efe2c006b93b4be87030116a12a124598c4
Summary:
Renaming ImmutableCFOptions::info_log and statistics to logger and stats. This is stage 2 in creating an ImmutableOptions class. It is necessary because the names match those in ImmutableOptions and have different types.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8227
Reviewed By: jay-zhuang
Differential Revision: D28000967
Pulled By: mrambacher
fbshipit-source-id: 3bf2aa04e8f1e8724d825b7deacf41080c14420b
Summary:
Add new C APIs to create the JemallocNodumpAllocator and set it on a Cache object.
`make test` passes with and without `DISABLE_JEMALLOC=1`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8178
Reviewed By: jay-zhuang
Differential Revision: D27944631
Pulled By: ajkr
fbshipit-source-id: 2531729aa285a8985c58f22f093c4d53029c4a7b
Summary:
This PR is a first step at attempting to clean up some of the Mutable/Immutable Options code. With this change, a DBOption and a ColumnFamilyOption can be reconstructed from their Mutable and Immutable equivalents, respectively.
readrandom tests do not show any performance degradation versus master (though both are slightly slower than the current 6.19 release).
There are still fields in the ImmutableCFOptions that are not CF options but DB options. Eventually, I would like to move those into an ImmutableOptions (= ImmutableDBOptions+ImmutableCFOptions). But that will be part of a future PR to minimize changes and disruptions.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8176
Reviewed By: pdillinger
Differential Revision: D27954339
Pulled By: mrambacher
fbshipit-source-id: ec6b805ba9afe6e094bffdbd76246c2d99aa9fad
Summary:
Add compaction API for secondary instance, which compact the files to a secondary DB path without installing to the LSM tree.
The API will be used to remote compaction.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8171
Test Plan: `make check`
Reviewed By: ajkr
Differential Revision: D27694545
Pulled By: jay-zhuang
fbshipit-source-id: 8ff3ec1bffdb2e1becee994918850c8902caf731
Peter Dillinger