Summary:
Currently, when we insert something into block cache, we say that the block cache capacity decreased by the size of the block. However, size of the block might be less than the actual memory used by this object. For example, 4.5KB block will actually use 8KB of memory. So even if we configure block cache to 10GB, our actually memory usage of block cache will be 20GB!
This problem showed up a lot in testing and just recently also showed up in MongoRocks production where we were using 30GB more memory than expected.
This diff will fix the problem. Instead of counting the block size, we will count memory used by the block. That way, a block cache configured to be 10GB will actually use only 10GB of memory.
I'm using non-portable function and I couldn't find info on portability on Google. However, it seems to work on Linux, which will cover majority of our use-cases.
Test Plan:
1. fill up mongo instance with 80GB of data
2. restart mongo with block cache size configured to 10GB
3. do a table scan in mongo
4. memory usage before the diff: 12GB. memory usage after the diff: 10.5GB
Reviewers: sdong, MarkCallaghan, rven, yhchiang
Reviewed By: yhchiang
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D40635
Summary: Replace force_bottommost_level_compaction in CompactRangeOption with an option that allow the user to (always skip, always compact, compact if compaction filter is present) the bottommost level for level based compaction.
Test Plan: make check
Reviewers: sdong, yhchiang, igor
Reviewed By: igor
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D40527
Summary:
Implementation of a table-level row cache.
It only caches point queries done through the `DB::Get` interface, queries done through the `Iterator` interface will completely skip the cache.
Supports snapshots and merge operations.
Test Plan: Ran `make valgrind_check commit-prereq`
Reviewers: igor, philipp, sdong
Reviewed By: sdong
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D39849
Summary:
Currently RocksDB silently ignores this issue and doesn't compress the data. Based on discussion, we agree that this is pretty bad because it can cause confusion for our users.
This patch fails DB::Open() if we don't support the compression that is specified in the options.
Test Plan: make check with LZ4 not present. If Snappy is not present all tests will just fail because Snappy is our default library. We should make Snappy the requirement, since without it our default DB::Open() fails.
Reviewers: sdong, MarkCallaghan, rven, yhchiang
Reviewed By: yhchiang
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D39687
Summary:
Add the funcion Cache.GetPinnedUsage() to return the memory size of entries
that are in use by the system (that is, all the entries not in the LRU list).
Test Plan:
Run ./cache_test and examine PinnedUsageTest.
Reviewers: tnovak, igor
Reviewed By: igor
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D40305
Summary:
This is https://reviews.facebook.net/D39999 but after introducing an option to force compaction the bottom most level
Changes in this patch
- Introduce force_bottommost_level_compaction to CompactRangeOptions that force compacting bottommost level during compaction
- Skip bottommost level compaction if we dont have a compaction filter and force_bottommost_level_compaction options is not set
Although tests pass on my machine but I suspect that there maybe some tests that I am not aware of that should use force_bottommost_level_compaction to pass in a deterministic way
Test Plan:
make check
adding new tests
Reviewers: igor, sdong, yhchiang
Reviewed By: yhchiang
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D40059
Summary:
This diff update DB::CompactRange to use RangeCompactionOptions instead of using multiple parameters
Old CompactRange is still available but deprecated
Test Plan:
make all check
make rocksdbjava
USE_CLANG=1 make all
OPT=-DROCKSDB_LITE make release
Reviewers: sdong, yhchiang, igor
Reviewed By: igor
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D40209
Summary:
Add an option in GetApproximateSize() so that the result will include estimated sizes in mem tables.
To implement it, implement an estimated count from the beginning to a key in skip list. The approach is to count to find the entry, how many Next() is issued from each level, and sum them with a weight that is <branching factor> ^ <level>.
Test Plan: Add a test case
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D40119
Summary:
This is part of an effort to better understand and optimize RocksDB stalls under high load. I added a feature to db_bench to periodically write QPS to CSV files. That way we can nicely see how our QPS changes in time (especially when DB is stalled) and can do a better job of evaluating our stall system (i.e. we want the QPS to be as constant as possible, as opposed to having bunch of stalls)
Cool part of CSV files is that we can easily graph them -- there are a bunch of tools available.
Test Plan:
Ran ./db_bench --report_interval_seconds=10 --benchmarks=fillrandom --num=10000000
and observed this in report.csv:
secs_elapsed,interval_qps
10,2725860
20,1980480
30,1863456
40,1454359
50,1460389
Reviewers: sdong, MarkCallaghan, rven, yhchiang
Reviewed By: yhchiang
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D40047
Summary:
We slow down data into the database to the rate of options.delayed_write_rate (a new option) with this patch.
The thread synchronization approach I take is to still synchronize write controller by DB mutex and GetDelay() is inside DB mutex. Try to minimize the frequency of getting time in GetDelay(). I verified it through db_bench and it seems to work
hard_rate_limit is deprecated.
options.delayed_write_rate is still not dynamically changeable. Need to work on it as a follow-up.
Test Plan: Add new unit tests in db_test
Reviewers: yhchiang, rven, kradhakrishnan, anthony, MarkCallaghan, igor
Reviewed By: igor
Subscribers: ikabiljo, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D36351
Summary:
[This is the resubmit of D39813. Tests were failing, so I reverted the diff. I found the bug and I'm now resubmitting]
If we don't do this, any calls to Entry() after WBWI mutation will result in undefined behavior. We need to re-fetch the offset from the skip list and regenerate the new pointer (because string's base pointer can change while mutating).
Test Plan: COMPILE_WITH_ASAN=1 make write_batch_with_index_test && ./write_batch_with_index_test
Reviewers: sdong
Reviewed By: sdong
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D39897
Summary: If we don't do this, any calls to Entry() after WBWI mutation will result in undefined behavior. We need to re-fetch the offset from the skip list and regenerate the new pointer (because string's base pointer can change while mutating).
Test Plan: COMPILE_WITH_ASAN=1 make write_batch_with_index_test && ./write_batch_with_index_test
Reviewers: sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D39813
Summary:
There is a hang during DB close in the following scenario:
a) a load with WAL disabled was done,
b) CancelAllBackgroundWork was called,
c) DB Close was called
This was because in that we will wait for a flush but we cannot do a
background flush because we have called CancelAllBackgroundWork which
marks the DB as shutting downn.
Test Plan: Added DBTest FlushOnDestroy
Reviewers: sdong
Reviewed By: sdong
Subscribers: yoshinorim, hermanlee4, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D39747
Summary:
Replacing the default value for compaction_filter_factory and compaction_filter_factory_v2 to be nullptr instead of DefaultCompactionFilterFactory / DefaultCompactionFilterFactoryV2
The reason for this is to be able to determine easily if we have compaction filter factory or not without depending on RTTI
Test Plan: make check
Reviewers: yoshinorim, ott, igor, sdong
Reviewed By: sdong
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D39693
Summary:
EventListener::OnFlushCompleted() now passes a structure instead
of a list of parameters. This minimizes the API change in the
future.
Test Plan:
listener_test
compact_files_test
example/compact_files_example
Reviewers: kradhakrishnan, sdong, IslamAbdelRahman, rven, igor
Reviewed By: rven, igor
Subscribers: IslamAbdelRahman, rven, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D39543
Summary: In DB::CompactRange(), change parameter "reduce_level" to "change_level". Users can compact all data to the last level if needed. By doing it, users can migrate the DB to options.level_compaction_dynamic_level_bytes=true.
Test Plan: Add a unit test for it.
Reviewers: yhchiang, anthony, kradhakrishnan, igor, rven
Reviewed By: rven
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D39099
Summary: Optimistic transactions supporting begin/commit/rollback semantics. Currently relies on checking the memtable to determine if there are any collisions at commit time. Not yet implemented would be a way of enuring the memtable has some minimum amount of history so that we won't fail to commit when the memtable is empty. You should probably start with transaction.h to get an overview of what is currently supported.
Test Plan: Added a new test, but still need to look into stress testing.
Reviewers: yhchiang, igor, rven, sdong
Reviewed By: sdong
Subscribers: adamretter, MarkCallaghan, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D33435
Summary:
For transactions, we are using the memtables to validate that there are no write conflicts. But after flushing, we don't have any memtables, and transactions could fail to commit. So we want to someone keep around some extra history to use for conflict checking. In addition, we want to provide a way to increase the size of this history if too many transactions fail to commit.
After chatting with people, it seems like everyone prefers just using Memtables to store this history (instead of a separate history structure). It seems like the best place for this is abstracted inside the memtable_list. I decide to create a separate list in MemtableListVersion as using the same list complicated the flush/installalflushresults logic too much.
This diff adds a new parameter to control how much memtable history to keep around after flushing. However, it sounds like people aren't too fond of adding new parameters. So I am making the default size of flushed+not-flushed memtables be set to max_write_buffers. This should not change the maximum amount of memory used, but make it more likely we're using closer the the limit. (We are now postponing deleting flushed memtables until the max_write_buffer limit is reached). So while we might use more memory on average, we are still obeying the limit set (and you could argue it's better to go ahead and use up memory now instead of waiting for a write stall to happen to test this limit).
However, if people are opposed to this default behavior, we can easily set it to 0 and require this parameter be set in order to use transactions.
Test Plan: Added a xfunc test to play around with setting different values of this parameter in all tests. Added testing in memtablelist_test and planning on adding more testing here.
Reviewers: sdong, rven, igor
Reviewed By: igor
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D37443
Summary:
sync_file_range is not always asyncronous and thus can block writes if we do this for WAL in the foreground thread. See more here: http://yoshinorimatsunobu.blogspot.com/2014/03/how-syncfilerange-really-works.html
Some users don't want us to call sync_file_range on WALs. Some other do.
Thus, I'm adding a separate option wal_bytes_per_sync to control calling
sync_file_range on WAL files. bytes_per_sync will apply only to table
files now.
Test Plan: no more sync_file_range for WAL as evidenced by strace
Reviewers: yhchiang, rven, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D38253
Summary:
When new capacity is larger than existing capacity, simply update the capacity to the new valie
When new capacity is less than existing capacity, but more than the usage, simply update the capacity to new value
When new capacity is less than the existing capacity and existing usage both, try to purge entries in LRU if feasible to make usage < capacity
Test Plan: Created unit tests in cache_test.cc
Reviewers: sdong, rven, yhchiang, igor
Reviewed By: igor
Subscribers: dhruba
Differential Revision: https://reviews.facebook.net/D37527
Summary:
Currently users have no idea a key is add, delete or merge from TablePropertiesCollector call back. Add a new function to add it.
Also refactor the codes so that
(1) make table property collector and internal table property collector two separate data structures with the later one now exposed
(2) table builders only receive internal table properties
Test Plan: Add cases in table_properties_collector_test to cover both of old and new ways of using TablePropertiesCollector.
Reviewers: yhchiang, igor.sugak, rven, igor
Reviewed By: rven, igor
Subscribers: meyering, yoshinorim, maykov, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D35373
Summary:
With this change, we use L1 and up to store compaction outputs in universal compaction.
The compaction pick logic stays the same. Outputs are stored in the largest "level" as possible.
If options.num_levels=1, it behaves all the same as now.
Test Plan:
1) convert most of existing unit tests for universal comapaction to include the option of one level and multiple levels.
2) add a unit test to cover parallel compaction in universal compaction and run it in one level and multiple levels
3) add unit test to migrate from multiple level setting back to one level setting
4) add a unit test to insert keys to trigger multiple rounds of compactions and verify results.
Reviewers: rven, kradhakrishnan, yhchiang, igor
Reviewed By: igor
Subscribers: meyering, leveldb, MarkCallaghan, dhruba
Differential Revision: https://reviews.facebook.net/D34539
Summary: Most of the approach is copied from WebSQL's MySQL branch. It's nice that we can do this without touching core RocksDB code.
Test Plan: Compiles and runs. Didn't test flashback code, as I don't have flashback device and most if it is c/p
Reviewers: MarkCallaghan, sdong
Reviewed By: sdong
Subscribers: rven, lgalanis, kradhakrishnan, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D35391
Summary: Update HISTORY.md for 3.10.0
Test Plan: no code chagne.
Reviewers: sdong, rven, igor
Reviewed By: igor
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D35871
Summary:
To understand the bug read t5943287 and check out the new test in column_family_test (ReadDroppedColumnFamily), iter 0.
RocksDB contract allowes you to read a drop column family as long as there is a live reference. However, since our iteration ignores dropped column families, AddLiveFiles() didn't mark files of a dropped column families as live. So we deleted them.
In this patch I no longer ignore dropped column families in the iteration. I think this behavior was confusing and it also led to this bug. Now if an iterator client wants to ignore dropped column families, he needs to do it explicitly.
Test Plan: Added a new unit test that is failing on master. Unit test succeeds now.
Reviewers: sdong, rven, yhchiang
Reviewed By: yhchiang
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D32535
Summary: WriteBatch and WriteBatchWithIndex now both inherit from a common abstract base class. This makes it easier to write code that is agnostic toward the implementation of the particular write batch. In particular, I plan on utilizing this abstraction to allow transactions to support using either implementation of a write batch.
Test Plan: modified existing WriteBatchWithIndex tests to test new functions. Running all tests.
Reviewers: igor, rven, yhchiang, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D34017
Summary: It is no longer used by the implementation, so we should also remove it from the public API.
Test Plan: make check
Reviewers: sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D34971
Summary:
When having fixed max_bytes_for_level_base, the ratio of size of largest level and the second one can range from 0 to the multiplier. This makes LSM tree frequently irregular and unpredictable. It can also cause poor space amplification in some cases.
In this improvement (proposed by Igor Kabiljo), we introduce a parameter option.level_compaction_use_dynamic_max_bytes. When turning it on, RocksDB is free to pick a level base in the range of (options.max_bytes_for_level_base/options.max_bytes_for_level_multiplier, options.max_bytes_for_level_base] so that real level ratios are close to options.max_bytes_for_level_multiplier.
Test Plan: New unit tests and pass tests suites including valgrind.
Reviewers: MarkCallaghan, rven, yhchiang, igor, ikabiljo
Reviewed By: ikabiljo
Subscribers: yoshinorim, ikabiljo, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D31437
Summary:
This diff fixes a bug introduced by D28521. Read-only backup engine can delete a backup that is later than the latest -- we never check the condition.
I also added a bunch of logging that will help with debugging cases like this in the future.
See more discussion at t6218248.
Test Plan: Added a unit test that was failing before the change. Also, see new LOG file contents: https://phabricator.fb.com/P19738984
Reviewers: benj, sanketh, sumeet, yhchiang, rven, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D33897
Summary: Remember whole key or prefix filtering on/off in SST files. If user opens the DB with a different setting that cannot be satisfied while reading the SST file, ignore the bloom filter.
Test Plan: Add a unit test for it
Reviewers: yhchiang, igor, rven
Reviewed By: rven
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D32889
Summary:
Get() now doesn't make use of bloom filter if it is prefix based. Add the check.
Didn't touch block based bloom filter. I can't fully reason whether it is correct to do that. But it's straight-forward to for full bloom filter.
Test Plan:
make all check
Add a test case in DBTest
Reviewers: rven, yhchiang, igor
Reviewed By: igor
Subscribers: MarkCallaghan, leveldb, dhruba, yoshinorim
Differential Revision: https://reviews.facebook.net/D31941
Summary: Add CappedFixTransform, which is the same as fixed length prefix extractor, except that when slice is shorter than the fixed length, it will use the full key.
Test Plan:
Add a test case for
db_test
options_test
and a new test
Reviewers: yhchiang, rven, igor
Reviewed By: igor
Subscribers: MarkCallaghan, leveldb, dhruba, yoshinorim
Differential Revision: https://reviews.facebook.net/D31887
Summary: This was a feature request by osquery. See task t5617758
Test Plan: compiles and memenv_test runs
Reviewers: yhchiang, rven, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D32115
Summary:
This diff adds BlockBasedTable format_version = 2. New format version brings better compressed block format for these compressions:
1) Zlib -- encode decompressed size in compressed block header
2) BZip2 -- encode decompressed size in compressed block header
3) LZ4 and LZ4HC -- instead of doing memcpy of size_t encode size as varint32. memcpy is very bad because the DB is not portable accross big/little endian machines or even platforms where size_t might be 8 or 4 bytes.
It does not affect format for snappy.
If you write a new database with format_version = 2, it will not be readable by RocksDB versions before 3.10. DB::Open() will return corruption in that case.
Test Plan:
Added a new test in db_test.
I will also run db_bench and verify VSIZE when block_cache == 1GB
Reviewers: yhchiang, rven, MarkCallaghan, dhruba, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D31461
Summary:
Since https://reviews.facebook.net/D16119, we ignore partial tailing writes. Because of that, we no longer need skip_log_error_on_recovery.
The documentation says "Skip log corruption error on recovery (If client is ok with losing most recent changes)", while the option actually ignores any corruption of the WAL (not only just the most recent changes). This is very dangerous and can lead to DB inconsistencies. This was originally set up to ignore partial tailing writes, which we now do automatically (after D16119). I have digged up old task t2416297 which confirms my findings.
Test Plan: There was actually no tests that verified correct behavior of skip_log_error_on_recovery.
Reviewers: yhchiang, rven, dhruba, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D30603
Summary:
Add support to allow nested config for block-based table factory. The format looks like this:
"write_buffer_size=1024;block_based_table_factory={block_size=4k};max_write_buffer_num=2"
Test Plan: unit test
Reviewers: yhchiang, rven, igor, ljin, jonahcohen
Reviewed By: jonahcohen
Subscribers: jonahcohen, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D29223
Summary:
There are two versions of FindObsoleteFiles():
* full scan, which is executed every 6 hours (and it's terribly slow)
* no full scan, which is executed every time a background process finishes and iterator is deleted
This diff is optimizing the second case (no full scan). Here's what we do before the diff:
* Get the list of obsolete files (files with ref==0). Some files in obsolete_files set might actually be live.
* Get the list of live files to avoid deleting files that are live.
* Delete files that are in obsolete_files and not in live_files.
After this diff:
* The only files with ref==0 that are still live are files that have been part of move compaction. Don't include moved files in obsolete_files.
* Get the list of obsolete files (which exclude moved files).
* No need to get the list of live files, since all files in obsolete_files need to be deleted.
I'll post the benchmark results, but you can get the feel of it here: https://reviews.facebook.net/D30123
This depends on D30123.
P.S. We should do full scan only in failure scenarios, not every 6 hours. I'll do this in a follow-up diff.
Test Plan:
One new unit test. Made sure that unit test fails if we don't have a `if (!f->moved)` safeguard in ~Version.
make check
Big number of compactions and flushes:
./db_stress --threads=30 --ops_per_thread=20000000 --max_key=10000 --column_families=20 --clear_column_family_one_in=10000000 --verify_before_write=0 --reopen=15 --max_background_compactions=10 --max_background_flushes=10 --db=/fast-rocksdb-tmp/db_stress --prefixpercent=0 --iterpercent=0 --writepercent=75 --db_write_buffer_size=2000000
Reviewers: yhchiang, rven, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D30249
Summary:
We now release the mutex before copying the files in the case
of the trivial move. This path does not use the compaction job.
Test Plan: DBTest.LevelCompactionThirdPath
Reviewers: yhchiang, igor, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D30381
Summary:
Allow Level-style compaction to place files in different paths
This diff provides the code for task 4854591. We now support level-compaction
to place files in different paths by specifying them in db_paths along with
the minimum level for files to store in that path.
Test Plan: ManualLevelCompactionOutputPathId in db_test.cc
Reviewers: yhchiang, MarkCallaghan, dhruba, yoshinorim, sdong
Reviewed By: sdong
Subscribers: yoshinorim, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D29799
Summary:
This diff changes compile to optimize for native platform by default. This will automatically turn on crc32 optimizations for modern processors, which greatly improves rocksdb's performance.
I also did some more changes to compilation documentation.
Test Plan:
compile with `make`, observe -march=native
compile with `PORTABLE=1 make`, observe no -march=native
Reviewers: sdong, rven, yhchiang, MarkCallaghan
Reviewed By: MarkCallaghan
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D30225
Summary:
Currently, blocks which have more than one reference (ie referenced by something other than cache itself) are evicted from cache. This doesn't make much sense:
- blocks are still in RAM, so the RAM usage reported by the cache is incorrect
- if the same block is needed by another iterator, it will be loaded and decompressed again
This diff changes the reference counting scheme a bit. Previously, if the cache contained the block, this was accounted for in its refcount. After this change, the refcount is only used to track external references. There is a boolean flag which indicates whether or not the block is contained in the cache.
This diff also changes how LRU list is used. Previously, both hashtable and the LRU list contained all blocks. After this change, the LRU list contains blocks with the refcount==0, ie those which can be evicted from the cache.
Note that this change still allows for cache to grow beyond its capacity. This happens when all blocks are pinned (ie refcount>0). This is consistent with the current behavior. The cache's insert function never fails. I spent lots of time trying to make table_reader and other places work with the insert which might failed. It turned out to be pretty hard. It might really destabilize some customers, so finally, I decided against doing this.
table_cache_remove_scan_count_limit option will be unneeded after this change, but I will remove it in the following diff, if this one gets approved
Test Plan: Ran tests, made sure they pass
Reviewers: sdong, ljin
Differential Revision: https://reviews.facebook.net/D25503
Igor Canadi