Summary:
Although we've been tracking SST unique IDs in the DB manifest
unconditionally, checking has been opt-in and with an extra pass at DB::Open
time. This changes the behavior of `verify_sst_unique_id_in_manifest` to
check unique ID against manifest every time an SST file is opened through
table cache (normal DB operations), replacing the explicit pass over files
at DB::Open time. This change also enables the option by default and
removes the "EXPERIMENTAL" designation.
One possible criticism is that the option no longer ensures the integrity
of a DB at Open time. This is far from an all-or-nothing issue. Verifying
the IDs of all SST files hardly ensures all the data in the DB is readable.
(VerifyChecksum is supposed to do that.) Also, with
max_open_files=-1 (default, extremely common), all SST files are
opened at DB::Open time anyway.
Implementation details:
* `VerifySstUniqueIdInManifest()` functions are the extra/explicit pass
that is now removed.
* Unit tests that manipulate/corrupt table properties have to opt out of
this check, because that corrupts the "actual" unique id. (And even for
testing we don't currently have a mechanism to set "no unique id"
in the in-memory file metadata for new files.)
* A lot of other unit test churn relates to (a) default checking on, and
(b) checking on SST open even without DB::Open (e.g. on flush)
* Use `FileMetaData` for more `TableCache` operations (in place of
`FileDescriptor`) so that we have access to the unique_id whenever
we might need to open an SST file. **There is the possibility of
performance impact because we can no longer use the more
localized `fd` part of an `FdWithKeyRange` but instead follow the
`file_metadata` pointer. However, this change (possible regression)
is only done for `GetMemoryUsageByTableReaders`.**
* Removed a completely unnecessary constructor overload of
`TableReaderOptions`
Possible follow-up:
* Verification only happens when opening through table cache. Are there
more places where this should happen?
* Improve error message when there is a file size mismatch vs. manifest
(FIXME added in the appropriate place).
* I'm not sure there's a justification for `FileDescriptor` to be distinct from
`FileMetaData`.
* I'm skeptical that `FdWithKeyRange` really still makes sense for
optimizing some data locality by duplicating some data in memory, but I
could be wrong.
* An unnecessary overload of NewTableReader was recently added, in
the public API nonetheless (though unusable there). It should be cleaned
up to put most things under `TableReaderOptions`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10532
Test Plan:
updated unit tests
Performance test showing no significant difference (just noise I think):
`./db_bench -benchmarks=readwhilewriting[-X10] -num=3000000 -disable_wal=1 -bloom_bits=8 -write_buffer_size=1000000 -target_file_size_base=1000000`
Before: readwhilewriting [AVG 10 runs] : 68702 (± 6932) ops/sec
After: readwhilewriting [AVG 10 runs] : 68239 (± 7198) ops/sec
Reviewed By: jay-zhuang
Differential Revision: D38765551
Pulled By: pdillinger
fbshipit-source-id: a827a708155f12344ab2a5c16e7701c7636da4c2
Summary:
Delete range logic is moved from `DBIter` to `MergingIterator`, and `MergingIterator` will seek to the end of a range deletion if possible instead of scanning through each key and check with `RangeDelAggregator`.
With the invariant that a key in level L (consider memtable as the first level, each immutable and L0 as a separate level) has a larger sequence number than all keys in any level >L, a range tombstone `[start, end)` from level L covers all keys in its range in any level >L. This property motivates optimizations in iterator:
- in `Seek(target)`, if level L has a range tombstone `[start, end)` that covers `target.UserKey`, then for all levels > L, we can do Seek() on `end` instead of `target` to skip some range tombstone covered keys.
- in `Next()/Prev()`, if the current key is covered by a range tombstone `[start, end)` from level L, we can do `Seek` to `end` for all levels > L.
This PR implements the above optimizations in `MergingIterator`. As all range tombstone covered keys are now skipped in `MergingIterator`, the range tombstone logic is removed from `DBIter`. The idea in this PR is similar to https://github.com/facebook/rocksdb/issues/7317, but this PR leaves `InternalIterator` interface mostly unchanged. **Credit**: the cascading seek optimization and the sentinel key (discussed below) are inspired by [Pebble](https://github.com/cockroachdb/pebble/blob/master/merging_iter.go) and suggested by ajkr in https://github.com/facebook/rocksdb/issues/7317. The two optimizations are mostly implemented in `SeekImpl()/SeekForPrevImpl()` and `IsNextDeleted()/IsPrevDeleted()` in `merging_iterator.cc`. See comments for each method for more detail.
One notable change is that the minHeap/maxHeap used by `MergingIterator` now contains range tombstone end keys besides point key iterators. This helps to reduce the number of key comparisons. For example, for a range tombstone `[start, end)`, a `start` and an `end` `HeapItem` are inserted into the heap. When a `HeapItem` for range tombstone start key is popped from the minHeap, we know this range tombstone becomes "active" in the sense that, before the range tombstone's end key is popped from the minHeap, all the keys popped from this heap is covered by the range tombstone's internal key range `[start, end)`.
Another major change, *delete range sentinel key*, is made to `LevelIterator`. Before this PR, when all point keys in an SST file are iterated through in `MergingIterator`, a level iterator would advance to the next SST file in its level. In the case when an SST file has a range tombstone that covers keys beyond the SST file's last point key, advancing to the next SST file would lose this range tombstone. Consequently, `MergingIterator` could return keys that should have been deleted by some range tombstone. We prevent this by pretending that file boundaries in each SST file are sentinel keys. A `LevelIterator` now only advance the file iterator once the sentinel key is processed.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10449
Test Plan:
- Added many unit tests in db_range_del_test
- Stress test: `./db_stress --readpercent=5 --prefixpercent=19 --writepercent=20 -delpercent=10 --iterpercent=44 --delrangepercent=2`
- Additional iterator stress test is added to verify against iterators against expected state: https://github.com/facebook/rocksdb/issues/10538. This is based on ajkr's previous attempt https://github.com/facebook/rocksdb/pull/5506#issuecomment-506021913.
```
python3 ./tools/db_crashtest.py blackbox --simple --write_buffer_size=524288 --target_file_size_base=524288 --max_bytes_for_level_base=2097152 --compression_type=none --max_background_compactions=8 --value_size_mult=33 --max_key=5000000 --interval=10 --duration=7200 --delrangepercent=3 --delpercent=9 --iterpercent=25 --writepercent=60 --readpercent=3 --prefixpercent=0 --num_iterations=1000 --range_deletion_width=100 --verify_iterator_with_expected_state_one_in=1
```
- Performance benchmark: I used a similar setup as in the blog [post](http://rocksdb.org/blog/2018/11/21/delete-range.html) that introduced DeleteRange, "a database with 5 million data keys, and 10000 range tombstones (ignoring those dropped during compaction) that were written in regular intervals after 4.5 million data keys were written". As expected, the performance with this PR depends on the range tombstone width.
```
# Setup:
TEST_TMPDIR=/dev/shm ./db_bench_main --benchmarks=fillrandom --writes=4500000 --num=5000000
TEST_TMPDIR=/dev/shm ./db_bench_main --benchmarks=overwrite --writes=500000 --num=5000000 --use_existing_db=true --writes_per_range_tombstone=50
# Scan entire DB
TEST_TMPDIR=/dev/shm ./db_bench_main --benchmarks=readseq[-X5] --use_existing_db=true --num=5000000 --disable_auto_compactions=true
# Short range scan (10 Next())
TEST_TMPDIR=/dev/shm/width-100/ ./db_bench_main --benchmarks=seekrandom[-X5] --use_existing_db=true --num=500000 --reads=100000 --seek_nexts=10 --disable_auto_compactions=true
# Long range scan(1000 Next())
TEST_TMPDIR=/dev/shm/width-100/ ./db_bench_main --benchmarks=seekrandom[-X5] --use_existing_db=true --num=500000 --reads=2500 --seek_nexts=1000 --disable_auto_compactions=true
```
Avg over of 10 runs (some slower tests had fews runs):
For the first column (tombstone), 0 means no range tombstone, 100-10000 means width of the 10k range tombstones, and 1 means there is a single range tombstone in the entire DB (width is 1000). The 1 tombstone case is to test regression when there's very few range tombstones in the DB, as no range tombstone is likely to take a different code path than with range tombstones.
- Scan entire DB
| tombstone width | Pre-PR ops/sec | Post-PR ops/sec | ±% |
| ------------- | ------------- | ------------- | ------------- |
| 0 range tombstone |2525600 (± 43564) |2486917 (± 33698) |-1.53% |
| 100 |1853835 (± 24736) |2073884 (± 32176) |+11.87% |
| 1000 |422415 (± 7466) |1115801 (± 22781) |+164.15% |
| 10000 |22384 (± 227) |227919 (± 6647) |+918.22% |
| 1 range tombstone |2176540 (± 39050) |2434954 (± 24563) |+11.87% |
- Short range scan
| tombstone width | Pre-PR ops/sec | Post-PR ops/sec | ±% |
| ------------- | ------------- | ------------- | ------------- |
| 0 range tombstone |35398 (± 533) |35338 (± 569) |-0.17% |
| 100 |28276 (± 664) |31684 (± 331) |+12.05% |
| 1000 |7637 (± 77) |25422 (± 277) |+232.88% |
| 10000 |1367 |28667 |+1997.07% |
| 1 range tombstone |32618 (± 581) |32748 (± 506) |+0.4% |
- Long range scan
| tombstone width | Pre-PR ops/sec | Post-PR ops/sec | ±% |
| ------------- | ------------- | ------------- | ------------- |
| 0 range tombstone |2262 (± 33) |2353 (± 20) |+4.02% |
| 100 |1696 (± 26) |1926 (± 18) |+13.56% |
| 1000 |410 (± 6) |1255 (± 29) |+206.1% |
| 10000 |25 |414 |+1556.0% |
| 1 range tombstone |1957 (± 30) |2185 (± 44) |+11.65% |
- Microbench does not show significant regression: https://gist.github.com/cbi42/59f280f85a59b678e7e5d8561e693b61
Reviewed By: ajkr
Differential Revision: D38450331
Pulled By: cbi42
fbshipit-source-id: b5ef12e8d8c289ed2e163ccdf277f5039b511fca
Summary:
Some APIs for getting live files, which are used by Checkpoint
and BackupEngine, can optionally trigger and wait for a flush. These
would deadlock when used on a read-only DB. Here we fix that by assuming
the user wants the overall operation to succeed and is OK without
flushing (because the DB is read-only).
Follow-up work: the same or other issues can be hit by directly invoking
some DB functions that are clearly not appropriate for read-only
instance, but are not covered by overrides in DBImplReadOnly and
CompactedDBImpl. These should be fixed to avoid similar problems on
accidental misuse. (Long term, it would be nice to have a DBReadOnly
class without those members, like BackupEngineReadOnly.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10569
Test Plan: tests updated to catch regression (hang before the fix)
Reviewed By: riversand963
Differential Revision: D38995759
Pulled By: pdillinger
fbshipit-source-id: f5f8bc7123e13cb45bd393dd974d7d6eda20bc68
Summary:
Timer has a limitation that it cannot re-register a task with the same name,
because the cancel only mark the task as invalid and wait for the Timer thread
to clean it up later, before the task is cleaned up, the same task name cannot
be added. Which makes the task option update likely to fail, which basically
cancel and re-register the same task name. Change the periodic task name to a
random unique id and store it in periodic_task_scheduler.
Also refactor the `periodic_work` to `periodic_task` to make each job function
as a `task`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10379
Test Plan: unittests
Reviewed By: ajkr
Differential Revision: D38000615
Pulled By: jay-zhuang
fbshipit-source-id: e4135f9422e3b53aaec8eda54f4e18ce633a279e
Summary:
The patch adds a new API `GetEntity` that can be used to perform
wide-column point lookups. It also extends the `Get` code path and
the `MemTable` / `MemTableList` and `Version` / `GetContext` logic
accordingly so that wide-column entities can be served from both
memtables and SSTs. If the result of a lookup is a wide-column entity
(`kTypeWideColumnEntity`), it is passed to the application in deserialized
form; if it is a plain old key-value (`kTypeValue`), it is presented as a
wide-column entity with a single default (anonymous) column.
(In contrast, regular `Get` returns plain old key-values as-is, and
returns the value of the default column for wide-column entities, see
https://github.com/facebook/rocksdb/issues/10483 .)
The result of `GetEntity` is a self-contained `PinnableWideColumns` object.
`PinnableWideColumns` contains a `PinnableSlice`, which either stores the
underlying data in its own buffer or holds on to a cache handle. It also contains
a `WideColumns` instance, which indexes the contents of the `PinnableSlice`,
so applications can access the values of columns efficiently.
There are several pieces of functionality which are currently not supported
for wide-column entities: there is currently no `MultiGetEntity` or wide-column
iterator; also, `Merge` and `GetMergeOperands` are not supported, and there
is no `GetEntity` implementation for read-only and secondary instances.
We plan to implement these in future PRs.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10540
Test Plan: `make check`
Reviewed By: akankshamahajan15
Differential Revision: D38847474
Pulled By: ltamasi
fbshipit-source-id: 42311a34ccdfe88b3775e847a5e2a5296e002b5b
Summary:
This PR avoids allocations and copies for the result of `GetMergeOperands()` when the average operand size is at least 256 bytes and the total operands size is at least 32KB. The `GetMergeOperands()` already included `PinnableSlice` but was calling `PinSelf()` (i.e., allocating and copying) for each operand. When this optimization takes effect, we instead call `PinSlice()` to skip that allocation and copy. Resources are pinned in order for the `PinnableSlice` to point to valid memory even after `GetMergeOperands()` returns.
The pinned resources include a referenced `SuperVersion`, a `MergingContext`, and a `PinnedIteratorsManager`. They are bundled into a `GetMergeOperandsState`. We use `SharedCleanablePtr` to share that bundle among all `PinnableSlice`s populated by `GetMergeOperands()`. That way, the last `PinnableSlice` to be `Reset()` will cleanup the bundle, including unreferencing the `SuperVersion`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10458
Test Plan:
- new DB level test
- measured benefit/regression in a number of memtable scenarios
Setup command:
```
$ ./db_bench -benchmarks=mergerandom -merge_operator=StringAppendOperator -num=$num -writes=16384 -key_size=16 -value_size=$value_sz -compression_type=none -write_buffer_size=1048576000
```
Benchmark command:
```
./db_bench -threads=$threads -use_existing_db=true -avoid_flush_during_recovery=true -write_buffer_size=1048576000 -benchmarks=readrandomoperands -merge_operator=StringAppendOperator -num=$num -duration=10
```
Worst regression is when a key has many tiny operands:
- Parameters: num=1 (implying 16384 operands per key), value_sz=8, threads=1
- `GetMergeOperands()` latency increases 682 micros -> 800 micros (+17%)
The regression disappears into the noise (<1% difference) if we remove the `Reset()` loop and the size counting loop. The former is arguably needed regardless of this PR as the convention in `Get()` and `MultiGet()` is to `Reset()` the input `PinnableSlice`s at the start. The latter could be optimized to count the size as we accumulate operands rather than after the fact.
Best improvement is when a key has large operands and high concurrency:
- Parameters: num=4 (implying 4096 operands per key), value_sz=2KB, threads=32
- `GetMergeOperands()` latency decreases 11492 micros -> 437 micros (-96%).
Reviewed By: cbi42
Differential Revision: D38336578
Pulled By: ajkr
fbshipit-source-id: 48146d127e04cb7f2d4d2939a2b9dff3aba18258
Summary:
TL;DR: due to a recent change, if you drop a column family,
often that DB will no longer fsync after writing new SST files
to remaining or new column families, which could lead to data
loss on power loss.
More bug detail:
The intent of https://github.com/facebook/rocksdb/issues/10049 was to Close FSDirectory objects at
DB::Close time rather than waiting for DB object destruction.
Unfortunately, it also closes shared FSDirectory objects on
DropColumnFamily (& destroy remaining handles), which can lead
to use-after-Close on FSDirectory shared with remaining column
families. Those "uses" are only Fsyncs (or redundant Closes). In
the default Posix filesystem, an Fsync on a closed FSDirectory is a
quiet no-op. Consequently (under most configurations), if you drop
a column family, that DB will no longer fsync after writing new SST
files to column families sharing the same directory (true under most
configurations).
More fix detail:
Basically, this removes unnecessary Close ops on destroying
ColumnFamilyData. We let `shared_ptr` take care of calling the
destructor at the right time. If the intent was to require Close be
called before destroying FSDirectory, that was not made clear by the
author of FileSystem and was not at all enforced by https://github.com/facebook/rocksdb/issues/10049, which
could have added `assert(fd_ == -1)` to `~PosixDirectory()` but did
not. To keep this fix simple, we relax the unit test for https://github.com/facebook/rocksdb/issues/10049 to allow
timely destruction of FSDirectory to suffice as Close (in
CountedFileSystem). Added a TODO to revisit that.
Also in this PR:
* Added a TODO to share FSDirectory instances between DB and its column
families. (Already shared among column families.)
* Made DB::Close attempt to close all its open FSDirectory objects even
if there is a failure in closing one. Also code clean-up around this
logic.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10460
Test Plan:
add an assert to check for use-after-Close. With that
existing tests can detect the misuse. With fix, tests pass (except noted
relaxing of unit test for https://github.com/facebook/rocksdb/issues/10049)
Reviewed By: ajkr
Differential Revision: D38357922
Pulled By: pdillinger
fbshipit-source-id: d42079cadbedf0a969f03389bf586b3b4e1f9137
Summary:
## Problem Summary
RocksDB will acquire the global mutex of db instance for every time when user calls `Write`. When RocksDB schedules a lot of compaction jobs, it will compete the mutex with write thread and it will hurt the write performance.
## Problem Solution:
I want to use log_write_mutex to replace the global mutex in most case so that we do not acquire it in write-thread unless there is a write-stall event or a write-buffer-full event occur.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7516
Test Plan:
1. make check
2. CI
3. COMPILE_WITH_TSAN=1 make db_stress
make crash_test
make crash_test_with_multiops_wp_txn
make crash_test_with_multiops_wc_txn
make crash_test_with_atomic_flush
Reviewed By: siying
Differential Revision: D36908702
Pulled By: riversand963
fbshipit-source-id: 59b13881f4f5c0a58fd3ca79128a396d9cd98efe
Summary:
`PeriodicWorkScheduler` is a global singleton, which were used to store per-instance setting `record_seqno_time_cadence_`. Move that to db_impl.h which is per-instance.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10383
Reviewed By: siying
Differential Revision: D37928009
Pulled By: jay-zhuang
fbshipit-source-id: e517754f4a9db98798ac04f72033d4b517f734e9
Summary:
Using the Sequence number to time mapping to decide if a key is hot or not in
compaction and place it in the corresponding level.
Note: the feature is not complete, level compaction will run indefinitely until
all penultimate level data is cold and small enough to not trigger compaction.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10370
Test Plan:
CI
* Run basic db_bench for universal compaction manually
Reviewed By: siying
Differential Revision: D37892338
Pulled By: jay-zhuang
fbshipit-source-id: 792bbd91b1ccc2f62b5d14c53118434bcaac4bbe
Summary:
Which will be used for tiered storage to preclude hot data from
compacting to the cold tier (the last level).
Internally, adding seqno to time mapping. A periodic_task is scheduled
to record the current_seqno -> current_time in certain cadence. When
memtable flush, the mapping informaiton is stored in sstable property.
During compaction, the mapping information are merged and get the
approximate time of sequence number, which is used to determine if a key
is recently inserted or not and preclude it from the last level if it's
recently inserted (within the `preclude_last_level_data_seconds`).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10338
Test Plan: CI
Reviewed By: siying
Differential Revision: D37810187
Pulled By: jay-zhuang
fbshipit-source-id: 6953be7a18a99de8b1cb3b162d712f79c2b4899f
Summary:
The patch builds on https://github.com/facebook/rocksdb/pull/9915 and adds
a new API called `PutEntity` that can be used to write a wide-column entity
to the database. The new API is added to both `DB` and `WriteBatch`. Note
that currently there is no way to retrieve these entities; more precisely, all
read APIs (`Get`, `MultiGet`, and iterator) return `NotSupported` when they
encounter a wide-column entity that is required to answer a query. Read-side
support (as well as other missing functionality like `Merge`, compaction filter,
and timestamp support) will be added in later PRs.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10242
Test Plan: `make check`
Reviewed By: riversand963
Differential Revision: D37369748
Pulled By: ltamasi
fbshipit-source-id: 7f5e412359ed7a400fd80b897dae5599dbcd685d
Summary:
https://github.com/facebook/rocksdb/issues/9984 changes the behavior of RocksDB: if logger creation failed during `SanitizeOptions()`,
`DB::Open()` will fail. However, since `SanitizeOptions()` is called in `DBImpl::DBImpl()`, we cannot
directly expose the error to caller without some additional work.
This is a first version proposal which:
- Adds a new member `init_logger_creation_s` to `DBImpl` to store the result of init logger creation
- Checks the error during `DB::Open()` and return it to caller if non-ok
This is not very ideal. We can alternatively move the logger creation logic out of the `SanitizeOptions()`.
Since `SanitizeOptions()` is used in other places, we need to check whether this change breaks anything
in case other callers of `SanitizeOptions()` assumes that a logger should be created.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10223
Test Plan: make check
Reviewed By: pdillinger
Differential Revision: D37321717
Pulled By: riversand963
fbshipit-source-id: 58042358a86369d606549dd9938933dd47591c4b
Summary:
So that DBImpl::RecoverLogFiles do not have to deal with implementation
details of WalFilter.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10214
Test Plan: make check
Reviewed By: ajkr
Differential Revision: D37299122
Pulled By: riversand963
fbshipit-source-id: acf1a80f1ef75da393d375f55968b2f3ac189816
Summary:
`FlushWAL(true /* sync */)` is used internally and for manual WAL sync. It had a bug when used together with `track_and_verify_wals_in_manifest` where the synced size tracked in MANIFEST was larger than the number of bytes actually synced.
The bug could be repro'd almost immediately with the following crash test command: `python3 tools/db_crashtest.py blackbox --simple --write_buffer_size=524288 --max_bytes_for_level_base=2097152 --target_file_size_base=524288 --duration=3600 --interval=10 --sync_fault_injection=1 --disable_wal=0 --checkpoint_one_in=1000 --max_key=10000 --value_size_mult=33`.
An example error message produced by the above command is shown below. The error sometimes arose from the checkpoint and other times arose from the main stress test DB.
```
Corruption: Size mismatch: WAL (log number: 119) in MANIFEST is 27938 bytes , but actually is 27859 bytes on disk.
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10185
Test Plan:
- repro unit test
- the above crash test command no longer finds the error. It does find a different error after a while longer such as "Corruption: WAL file 481 required by manifest but not in directory list"
Reviewed By: riversand963
Differential Revision: D37200993
Pulled By: ajkr
fbshipit-source-id: 98e0071c1a89f4d009888512ed89f9219779ae5f
Summary:
A consequence of https://github.com/facebook/rocksdb/issues/9990 was requiring a non-empty DB ID to generate
new SST files. But if the DB ID is not tracked in the manifest and the IDENTITY file
is somehow truncated to 0 bytes, then an empty DB ID would be assigned, leading
to crash. This change ensures a non-empty DB ID is assigned and set in the
IDENTITY file.
Also,
* Some light refactoring to clean up the logic
* (I/O efficiency) If the ID is tracked in the manifest and already matches the
IDENTITY file, don't needlessly overwrite the file.
* (Debugging) Log the DB ID to info log on open, because sometimes IDENTITY
can change if DB is moved around (though it would be unusual for info log to
be copied/moved without IDENTITY file)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10173
Test Plan: unit tests expanded/updated
Reviewed By: ajkr
Differential Revision: D37176545
Pulled By: pdillinger
fbshipit-source-id: a9b414cd35bfa33de48af322a36c24538d50bef1
Summary:
In RocksDB, keys are associated with (internal) sequence numbers which denote when the keys are written
to the database. Sequence numbers in different RocksDB instances are unrelated, thus not comparable.
It is nice if we can associate sequence numbers with their corresponding actual timestamps. One thing we can
do is to support user-defined timestamp, which allows the applications to specify the format of custom timestamps
and encode a timestamp with each key. More details can be found at https://github.com/facebook/rocksdb/wiki/User-defined-Timestamp-%28Experimental%29.
This PR provides a different but complementary approach. We can associate rocksdb snapshots (defined in
https://github.com/facebook/rocksdb/blob/7.2.fb/include/rocksdb/snapshot.h#L20) with **user-specified** timestamps.
Since a snapshot is essentially an object representing a sequence number, this PR establishes a bi-directional mapping between sequence numbers and timestamps.
In the past, snapshots are usually taken by readers. The current super-version is grabbed, and a `rocksdb::Snapshot`
object is created with the last published sequence number of the super-version. You can see that the reader actually
has no good idea of what timestamp to assign to this snapshot, because by the time the `GetSnapshot()` is called,
an arbitrarily long period of time may have already elapsed since the last write, which is when the last published
sequence number is written.
This observation motivates the creation of "timestamped" snapshots on the write path. Currently, this functionality is
exposed only to the layer of `TransactionDB`. Application can tell RocksDB to create a snapshot when a transaction
commits, effectively associating the last sequence number with a timestamp. It is also assumed that application will
ensure any two snapshots with timestamps should satisfy the following:
```
snapshot1.seq < snapshot2.seq iff. snapshot1.ts < snapshot2.ts
```
If the application can guarantee that when a reader takes a timestamped snapshot, there is no active writes going on
in the database, then we also allow the user to use a new API `TransactionDB::CreateTimestampedSnapshot()` to create
a snapshot with associated timestamp.
Code example
```cpp
// Create a timestamped snapshot when committing transaction.
txn->SetCommitTimestamp(100);
txn->SetSnapshotOnNextOperation();
txn->Commit();
// A wrapper API for convenience
Status Transaction::CommitAndTryCreateSnapshot(
std::shared_ptr<TransactionNotifier> notifier,
TxnTimestamp ts,
std::shared_ptr<const Snapshot>* ret);
// Create a timestamped snapshot if caller guarantees no concurrent writes
std::pair<Status, std::shared_ptr<const Snapshot>> snapshot = txn_db->CreateTimestampedSnapshot(100);
```
The snapshots created in this way will be managed by RocksDB with ref-counting and potentially shared with
other readers. We provide the following APIs for readers to retrieve a snapshot given a timestamp.
```cpp
// Return the timestamped snapshot correponding to given timestamp. If ts is
// kMaxTxnTimestamp, then we return the latest timestamped snapshot if present.
// Othersise, we return the snapshot whose timestamp is equal to `ts`. If no
// such snapshot exists, then we return null.
std::shared_ptr<const Snapshot> TransactionDB::GetTimestampedSnapshot(TxnTimestamp ts) const;
// Return the latest timestamped snapshot if present.
std::shared_ptr<const Snapshot> TransactionDB::GetLatestTimestampedSnapshot() const;
```
We also provide two additional APIs for stats collection and reporting purposes.
```cpp
Status TransactionDB::GetAllTimestampedSnapshots(
std::vector<std::shared_ptr<const Snapshot>>& snapshots) const;
// Return timestamped snapshots whose timestamps fall in [ts_lb, ts_ub) and store them in `snapshots`.
Status TransactionDB::GetTimestampedSnapshots(
TxnTimestamp ts_lb,
TxnTimestamp ts_ub,
std::vector<std::shared_ptr<const Snapshot>>& snapshots) const;
```
To prevent the number of timestamped snapshots from growing infinitely, we provide the following API to release
timestamped snapshots whose timestamps are older than or equal to a given threshold.
```cpp
void TransactionDB::ReleaseTimestampedSnapshotsOlderThan(TxnTimestamp ts);
```
Before shutdown, RocksDB will release all timestamped snapshots.
Comparison with user-defined timestamp and how they can be combined:
User-defined timestamp persists every key with a timestamp, while timestamped snapshots maintain a volatile
mapping between snapshots (sequence numbers) and timestamps.
Different internal keys with the same user key but different timestamps will be treated as different by compaction,
thus a newer version will not hide older versions (with smaller timestamps) unless they are eligible for garbage collection.
In contrast, taking a timestamped snapshot at a certain sequence number and timestamp prevents all the keys visible in
this snapshot from been dropped by compaction. Here, visible means (seq < snapshot and most recent).
The timestamped snapshot supports the semantics of reading at an exact point in time.
Timestamped snapshots can also be used with user-defined timestamp.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9879
Test Plan:
```
make check
TEST_TMPDIR=/dev/shm make crash_test_with_txn
```
Reviewed By: siying
Differential Revision: D35783919
Pulled By: riversand963
fbshipit-source-id: 586ad905e169189e19d3bfc0cb0177a7239d1bd4
Summary:
The default implementation of Close() function in Directory/FSDirectory classes returns `NotSupported` status. However, we don't want operations that worked in older versions to begin failing after upgrading when run on FileSystems that have not implemented Directory::Close() yet. So we require the upper level that calls Close() function should properly handle "NotSupported" status instead of treating it as an error status.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10127
Reviewed By: ajkr
Differential Revision: D36971112
Pulled By: littlepig2013
fbshipit-source-id: 100f0e6ad1191e1acc1ba6458c566a11724cf466
Summary:
As pointed out by [https://github.com/facebook/rocksdb/pull/8351#discussion_r645765422](https://github.com/facebook/rocksdb/pull/8351#discussion_r645765422), check `manual_compaction_paused` and `manual_compaction_canceled` can be reduced by setting `*canceled` to be true in `DisableManualCompaction()` and `*canceled` to be false in the last time calling `EnableManualCompaction()`.
Changed Tests: The origin `DBTest2.PausingManualCompaction1` uses a callback function to increase `manual_compaction_paused` and the origin CompactionJob/CompactionIterator with `manual_compaction_paused` can detect this. I changed the callback function so that it sets `*canceled` as true if `canceled` is not `nullptr` (to notify CompactionJob/CompactionIterator the compaction has been canceled).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10070
Test Plan: This change does not introduce new features, but some slight difference in compaction implementation. Run the same manual compaction unit tests as before (e.g., PausingManualCompaction[1-4], CancelManualCompaction[1-2], CancelManualCompactionWithListener in db_test2, and db_compaction_test).
Reviewed By: ajkr
Differential Revision: D36949133
Pulled By: littlepig2013
fbshipit-source-id: c5dc4c956fbf8f624003a0f5ad2690240063a821
Summary:
With this change, when a given read timestamp is smaller than the column-family's full_history_ts_low, Get(), MultiGet() and iterators APIs will return Status::InValidArgument().
Test plan
```
$COMPILE_WITH_ASAN=1 make -j24 all
$./db_with_timestamp_basic_test --gtest_filter=DBBasicTestWithTimestamp.UpdateFullHistoryTsLow
$ make -j24 check
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10109
Reviewed By: riversand963
Differential Revision: D36901126
Pulled By: jowlyzhang
fbshipit-source-id: 255feb1a66195351f06c1d0e42acb1ff74527f86
Summary:
Currently, the DB directory file descriptor is left open until the deconstruction process (`DB::Close()` does not close the file descriptor). To verify this, comment out the lines between `db_ = nullptr` and `db_->Close()` (line 512, 513, 514, 515 in ldb_cmd.cc) to leak the ``db_'' object, build `ldb` tool and run
```
strace --trace=open,openat,close ./ldb --db=$TEST_TMPDIR --ignore_unknown_options put K1 V1 --create_if_missing
```
There is one directory file descriptor that is not closed in the strace log.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10049
Test Plan: Add a new unit test DBBasicTest.DBCloseAllDirectoryFDs: Open a database with different WAL directory and three different data directories, and all directory file descriptors should be closed after calling Close(). Explicitly call Close() after a directory file descriptor is not used so that the counter of directory open and close should be equivalent.
Reviewed By: ajkr, hx235
Differential Revision: D36722135
Pulled By: littlepig2013
fbshipit-source-id: 07bdc2abc417c6b30997b9bbef1f79aa757b21ff
Summary:
`db_impl.alive_log_files_` is used to track the WAL size in `db_impl.logs_`.
Get the `LogFileNumberSize` obj in `alive_log_files_` the same time as `log_writer` to keep them consistent.
For this issue, it's not safe to do `deque::reverse_iterator::operator*` and `deque::pop_front()` concurrently,
so remove the tail cache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10086
Test Plan:
```
# on Windows
gtest-parallel ./db_test --gtest_filter=DBTest.FileCreationRandomFailure -r 1000 -w 100
```
Reviewed By: riversand963
Differential Revision: D36822373
Pulled By: jay-zhuang
fbshipit-source-id: 5e738051dfc7bcf6a15d85ba25e6365df6b6a6af
Summary:
In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't
flush the data from WAL to L0 for all column families if possible. As a
result, not all column families can increase their log_numbers, and
min_log_number_to_keep won't change.
For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change.
If we persist a new MANIFEST with
advanced log_numbers for some column families, then during a second
crash after persisting the MANIFEST, RocksDB will see some column
families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail.
As a solution, RocksDB will persist the new MANIFEST after successfully syncing the new WAL.
If a future recovery starts from the new MANIFEST, then it means the new WAL is successfully synced. Due to the sentinel empty write batch at the beginning, kPointInTimeRecovery of WAL is guaranteed to go after this point.
If future recovery starts from the old MANIFEST, it means the writing the new MANIFEST failed. We won't have the "SST ahead of WAL" error.
Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9922
Test Plan:
1. Update unit tests to fail without this change
2. make crast_test -j
Branch with unit test and no fix https://github.com/facebook/rocksdb/pull/9942 to keep track of unit test (without fix)
Reviewed By: riversand963
Differential Revision: D36043701
Pulled By: akankshamahajan15
fbshipit-source-id: 5760970db0a0920fb73d3c054a4155733500acd9
Summary:
This variable is actually not being used for anything meaningful, thus remove it.
This can make https://github.com/facebook/rocksdb/issues/7516 slightly simpler by reducing the amount of state that must be made lock-free.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10078
Test Plan: make check
Reviewed By: ajkr
Differential Revision: D36779817
Pulled By: riversand963
fbshipit-source-id: ffb0d9ad6149616917ae5e02bb28102cb90fc406
Summary:
This PR adds timestamp support to a read only DB instance opened as `DBImplReadOnly`. A follow up PR will add the same support to `CompactedDBImpl`.
With this, read only database has these timestamp related APIs:
`ReadOptions.timestamp` : read should return the latest data visible to this specified timestamp
`Iterator::timestamp()` : returns the timestamp associated with the key, value
`DB:Get(..., std::string* timestamp)` : returns the timestamp associated with the key, value in `timestamp`
Test plan (on devserver):
```
$COMPILE_WITH_ASAN=1 make -j24 all
$./db_with_timestamp_basic_test --gtest_filter=DBBasicTestWithTimestamp.ReadOnlyDB*
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/10004
Reviewed By: riversand963
Differential Revision: D36434422
Pulled By: jowlyzhang
fbshipit-source-id: 5d949e65b1ffb845758000e2b310fdd4aae71cfb
Summary:
Start tracking SST unique id in MANIFEST, which is used to verify with
SST properties to make sure the SST file is not overwritten or
misplaced. A DB option `try_verify_sst_unique_id` is introduced to
enable/disable the verification, if enabled, it opens all SST files
during DB-open to read the unique_id from table properties (default is
false), so it's recommended to use it with `max_open_files = -1` to
pre-open the files.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9990
Test Plan: unittests, format-compatible test, mini-crash
Reviewed By: anand1976
Differential Revision: D36381863
Pulled By: jay-zhuang
fbshipit-source-id: 89ea2eb6b35ed3e80ead9c724eb096083eaba63f
Summary:
Right now we still don't fully use std::numeric_limits but use a macro, mainly for supporting VS 2013. Right now we only support VS 2017 and up so it is not a problem. The code comment claims that MinGW still needs it. We don't have a CI running MinGW so it's hard to validate. since we now require C++17, it's hard to imagine MinGW would still build RocksDB but doesn't support std::numeric_limits<>.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9954
Test Plan: See CI Runs.
Reviewed By: riversand963
Differential Revision: D36173954
fbshipit-source-id: a35a73af17cdcae20e258cdef57fcf29a50b49e0
Summary:
Left HISTORY.md and unit tests.
Added a new unit test to repro the corruption scenario that this PR fixes, and HISTORY.md line for that.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9906
Reviewed By: riversand963
Differential Revision: D35940093
Pulled By: ajkr
fbshipit-source-id: 9816f99e1ce405ba36f316beb4f6378c37c8c86b
Summary:
1) In case of non-TransactionDB and avoid_flush_during_recovery = true, RocksDB won't
flush the data from WAL to L0 for all column families if possible. As a
result, not all column families can increase their log_numbers, and
min_log_number_to_keep won't change.
2) For transaction DB (.allow_2pc), even with the flush, there may be old WAL files that it must not delete because they can contain data of uncommitted transactions and min_log_number_to_keep won't change.
If we persist a new MANIFEST with
advanced log_numbers for some column families, then during a second
crash after persisting the MANIFEST, RocksDB will see some column
families' log_numbers larger than the corrupted wal, and the "column family inconsistency" error will be hit, causing recovery to fail.
As a solution,
1. the corrupted WALs whose numbers are larger than the
corrupted wal and smaller than the new WAL will be moved to archive folder.
2. Currently, RocksDB DB::Open() may creates and writes to two new MANIFEST files even before recovery succeeds. This PR buffers the edits in a structure and writes to a new MANIFEST after recovery is successful
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9634
Test Plan:
1. Added new unit tests
2. make crast_test -j
Reviewed By: riversand963
Differential Revision: D34463666
Pulled By: akankshamahajan15
fbshipit-source-id: e233d3af0ed4e2028ca0cf051e5a334a0fdc9d19
Summary:
Options `preserve_deletes` and `iter_start_seqnum` have been removed since 7.0.
This PR removes dead code related to these two removed options.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9825
Test Plan: make check
Reviewed By: akankshamahajan15
Differential Revision: D35517950
Pulled By: riversand963
fbshipit-source-id: 86282ce5ec4087acb94a06a42a1b6d55b1715482
Summary:
For write-prepared/write-unprepared transactions,
GetCommitTimeWriteBatch() can be used only if the transaction is started
with `TransactionOptions::use_only_the_last_commit_time_batch_for_recovery` set
to true. Otherwise, it is possible that multiple uncommitted versions of the
same key exist in the database. During bottommost compaction, RocksDB may
set the sequence numbers of both to zero once they become committed, causing
output SST file to have two identical internal keys.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9794
Test Plan:
make check
pay special attention to the following
```
transaction_test --gtest_filter=MySQLStyleTransactionTest/MySQLStyleTransactionTest.TransactionStressTest/*
```
Reviewed By: lth
Differential Revision: D35327214
Pulled By: riversand963
fbshipit-source-id: 3bae00a28359c10e96e4c6f676d20de5610d8a0f
Summary:
In https://github.com/facebook/rocksdb/issues/9659, when `DisableManualCompaction()` is issued, the foreground
manual compaction thread does not have to wait background compaction
thread to finish. Which could be a problem that the user re-enable
manual compaction with `EnableManualCompaction()`, it may re-enable the
BG compaction which supposed be cancelled.
This patch makes the FG compaction wait on
`manual_compaction_state.done`, which either be set by BG compaction or
Unschedule callback. Then when FG manual compaction thread returns, it
should not have BG compaction running. So shared_ptr is no longer needed
for `manual_compaction_state`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9694
Test Plan: a StressTest and unittest
Reviewed By: ajkr
Differential Revision: D34885472
Pulled By: jay-zhuang
fbshipit-source-id: e6476175b43e8c59cd49f5c09241036a0716c274
Summary:
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9686
According to https://www.cplusplus.com/reference/deque/deque/back/,
"
The container is accessed (neither the const nor the non-const versions modify the container).
The last element is potentially accessed or modified by the caller. Concurrently accessing or modifying other elements is safe.
"
Also according to https://www.cplusplus.com/reference/deque/deque/pop_front/,
"
The container is modified.
The first element is modified. Concurrently accessing or modifying other elements is safe (although see iterator validity above).
"
In RocksDB, we never pop the last element of `DBImpl::alive_log_files_`. We have been
exploiting this fact and the above two properties when ensuring correctness when
`DBImpl::alive_log_files_` may be accessed concurrently. Specifically, it can be accessed
in the write path when db mutex is released. Sometimes, the log_mute_ is held. It can also be accessed in `FindObsoleteFiles()`
when db mutex is always held. It can also be accessed
during recovery when db mutex is also held.
Given the fact that we never pop the last element of alive_log_files_, we currently do not
acquire additional locks when accessing it in `WriteToWAL()` as follows
```
alive_log_files_.back().AddSize(log_entry.size());
```
This is problematic.
Check source code of deque.h
```
back() _GLIBCXX_NOEXCEPT
{
__glibcxx_requires_nonempty();
...
}
pop_front() _GLIBCXX_NOEXCEPT
{
...
if (this->_M_impl._M_start._M_cur
!= this->_M_impl._M_start._M_last - 1)
{
...
++this->_M_impl._M_start._M_cur;
}
...
}
```
`back()` will actually call `__glibcxx_requires_nonempty()` first.
If `__glibcxx_requires_nonempty()` is enabled and not an empty macro,
it will call `empty()`
```
bool empty() {
return this->_M_impl._M_finish == this->_M_impl._M_start;
}
```
You can see that it will access `this->_M_impl._M_start`, racing with `pop_front()`.
Therefore, TSAN will actually catch the bug in this case.
To be able to use TSAN on our library and unit tests, we should always coordinate
concurrent accesses to STL containers properly.
We need to pass information about db mutex and log mutex into `WriteToWAL()`, otherwise
it's impossible to know which mutex to acquire inside the function.
To fix this, we can catch the tail of `alive_log_files_` by reference, so that we do not have to call `back()` in `WriteToWAL()`.
Reviewed By: pdillinger
Differential Revision: D34780309
fbshipit-source-id: 1def9821f0c437f2736c6a26445d75890377889b
Summary:
https://github.com/facebook/rocksdb/issues/9625 didn't change the unschedule condition which was waiting for the background thread to clean-up the compaction.
make sure we only unschedule the task when it's scheduled.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9659
Reviewed By: ajkr
Differential Revision: D34651820
Pulled By: jay-zhuang
fbshipit-source-id: 23f42081b15ec8886cd81cbf131b116e0c74dc2f
Summary:
Timer crash when multiple DB instances doing heavy DB open and close
operations concurrently. Which is caused by adding a timer task with
smaller timestamp than the current running task. Fix it by moving the
getting new task timestamp part within timer mutex protection.
And other fixes:
- Disallow adding duplicated function name to timer
- Fix a minor memory leak in timer when a running task is cancelled
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9656
Reviewed By: ajkr
Differential Revision: D34626296
Pulled By: jay-zhuang
fbshipit-source-id: 6b6d96a5149746bf503546244912a9e41a0c5f6b
Summary:
As disscussed in (https://github.com/facebook/rocksdb/issues/9223), Here added a new API named DB::OpenAndTrimHistory, this API will open DB and trim data to the timestamp specofied by **trim_ts** (The data with newer timestamp than specified trim bound will be removed). This API should only be used at a timestamp-enabled db instance recovery.
And this PR implemented a new iterator named HistoryTrimmingIterator to support trimming history with a new API named DB::OpenAndTrimHistory. HistoryTrimmingIterator wrapped around the underlying InternalITerator such that keys whose timestamps newer than **trim_ts** should not be returned to the compaction iterator while **trim_ts** is not null.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9410
Reviewed By: ltamasi
Differential Revision: D34410207
Pulled By: riversand963
fbshipit-source-id: e54049dc234eccd673244c566b15df58df5a6236
Summary:
PR https://github.com/facebook/rocksdb/issues/9557 introduced a race condition between manual compaction
foreground thread and background compaction thread.
This PR adds the ability to really unschedule manual compaction from
thread-pool queue by differentiate tag name for manual compaction and
other tasks.
Also fix an issue that db `close()` didn't cancel the manual compaction thread.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9625
Test Plan: unittest not hang
Reviewed By: ajkr
Differential Revision: D34410811
Pulled By: jay-zhuang
fbshipit-source-id: cb14065eabb8cf1345fa042b5652d4f788c0c40c
Summary:
We found a case of cacheline bouncing due to writers locking/unlocking `mutex_` and readers accessing `block_cache_tracer_`. We discovered it only after the issue was fixed by https://github.com/facebook/rocksdb/issues/9462 shifting the `DBImpl` members such that `mutex_` and `block_cache_tracer_` were naturally placed in separate cachelines in our regression testing setup. This PR forces the cacheline alignment of `mutex_` so we don't accidentally reintroduce the problem.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9637
Reviewed By: riversand963
Differential Revision: D34502233
Pulled By: ajkr
fbshipit-source-id: 46aa313b7fe83e80c3de254e332b6fb242434c07
Summary:
The following sequence of events can cause silent data loss for write-committed
transactions.
```
Time thread 1 bg flush
| db->Put("a")
| txn = NewTxn()
| txn->Put("b", "v")
| txn->Prepare() // writes only to 5.log
| db->SwitchMemtable() // memtable 1 has "a"
| // close 5.log,
| // creates 8.log
| trigger flush
| pick memtable 1
| unlock db mutex
| write new sst
| txn->ctwb->Put("gtid", "1") // writes 8.log
| txn->Commit() // writes to 8.log
| // writes to memtable 2
| compute min_log_number_to_keep_2pc, this
| will be 8 (incorrect).
|
| Purge obsolete wals, including 5.log
|
V
```
At this point, writes of txn exists only in memtable. Close db without flush because db thinks the data in
memtable are backed by log. Then reopen, the writes are lost except key-value pair {"gtid"->"1"},
only the commit marker of txn is in 8.log
The reason lies in `PrecomputeMinLogNumberToKeep2PC()` which calls `FindMinPrepLogReferencedByMemTable()`.
In the above example, when bg flush thread tries to find obsolete wals, it uses the information
computed by `PrecomputeMinLogNumberToKeep2PC()`. The return value of `PrecomputeMinLogNumberToKeep2PC()`
depends on three components
- `PrecomputeMinLogNumberToKeepNon2PC()`. This represents the WAL that has unflushed data. As the name of this method suggests, it does not account for 2PC. Although the keys reside in the prepare section of a previous WAL, the column family references the current WAL when they are actually inserted into the memtable during txn commit.
- `prep_tracker->FindMinLogContainingOutstandingPrep()`. This represents the WAL with a prepare section but the txn hasn't committed.
- `FindMinPrepLogReferencedByMemTable()`. This represents the WAL on which some memtables (mutable and immutable) depend for their unflushed data.
The bug lies in `FindMinPrepLogReferencedByMemTable()`. Originally, this function skips checking the column families
that are being flushed, but the unit test added in this PR shows that they should not be. In this unit test, there is
only the default column family, and one of its memtables has unflushed data backed by a prepare section in 5.log.
We should return this information via `FindMinPrepLogReferencedByMemTable()`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9571
Test Plan:
```
./transaction_test --gtest_filter=*/TransactionTest.SwitchMemtableDuringPrepareAndCommit_WC/*
make check
```
Reviewed By: siying
Differential Revision: D34235236
Pulled By: riversand963
fbshipit-source-id: 120eb21a666728a38dda77b96276c6af72b008b1
Summary:
Fix `DisableManualCompaction()` has to wait scheduled manual compaction to start the execution to cancel the job.
When a manual compaction in thread-pool queue is cancel, set the job is_canceled to true and clean the resource.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9557
Test Plan: added unittest that will hang without the change
Reviewed By: ajkr
Differential Revision: D34214910
Pulled By: jay-zhuang
fbshipit-source-id: 89dbaee78ddf26eb13ce862c2b15f4a098b36a78
Summary:
ajkr reminded me that we have a rule of not including per-kv related data in `WriteOptions`.
Namely, `WriteOptions` should not include information about "what-to-write", but should just
include information about "how-to-write".
According to this rule, `WriteOptions::timestamp` (experimental) is clearly a violation. Therefore,
this PR removes `WriteOptions::timestamp` for compliance.
After the removal, we need to pass timestamp info via another set of APIs. This PR proposes a set
of overloaded functions `Put(write_opts, key, value, ts)`, `Delete(write_opts, key, ts)`, and
`SingleDelete(write_opts, key, ts)`. Planned to add `Write(write_opts, batch, ts)`, but its complexity
made me reconsider doing it in another PR (maybe).
For better checking and returning error early, we also add a new set of APIs to `WriteBatch` that take
extra `timestamp` information when writing to `WriteBatch`es.
These set of APIs in `WriteBatchWithIndex` are currently not supported, and are on our TODO list.
Removed `WriteBatch::AssignTimestamps()` and renamed `WriteBatch::AssignTimestamp()` to
`WriteBatch::UpdateTimestamps()` since this method require that all keys have space for timestamps
allocated already and multiple timestamps can be updated.
The constructor of `WriteBatch` now takes a fourth argument `default_cf_ts_sz` which is the timestamp
size of the default column family. This will be used to allocate space when calling APIs that do not
specify a column family handle.
Also, updated `DB::Get()`, `DB::MultiGet()`, `DB::NewIterator()`, `DB::NewIterators()` methods, replacing
some assertions about timestamp to returning Status code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/8946
Test Plan:
make check
./db_bench -benchmarks=fillseq,fillrandom,readrandom,readseq,deleterandom -user_timestamp_size=8
./db_stress --user_timestamp_size=8 -nooverwritepercent=0 -test_secondary=0 -secondary_catch_up_one_in=0 -continuous_verification_interval=0
Make sure there is no perf regression by running the following
```
./db_bench_opt -db=/dev/shm/rocksdb -use_existing_db=0 -level0_stop_writes_trigger=256 -level0_slowdown_writes_trigger=256 -level0_file_num_compaction_trigger=256 -disable_wal=1 -duration=10 -benchmarks=fillrandom
```
Before this PR
```
DB path: [/dev/shm/rocksdb]
fillrandom : 1.831 micros/op 546235 ops/sec; 60.4 MB/s
```
After this PR
```
DB path: [/dev/shm/rocksdb]
fillrandom : 1.820 micros/op 549404 ops/sec; 60.8 MB/s
```
Reviewed By: ltamasi
Differential Revision: D33721359
Pulled By: riversand963
fbshipit-source-id: c131561534272c120ffb80711d42748d21badf09
Summary:
* Add more micro-benchmark tests
* Expose an API in DBImpl for waiting for compactions (still not visible to the user)
* Add argument name for ribbon_bench
* remove benchmark run from CI, as it runs too long.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9436
Test Plan: CI
Reviewed By: riversand963
Differential Revision: D33777836
Pulled By: jay-zhuang
fbshipit-source-id: c05de3bc082cc05b5d019f00b324e774bf4bbd96
Summary:
As (https://github.com/facebook/rocksdb/issues/9210) discussed, the **full_history_ts_low** is a member of CompactRangeOptions currently, which means a CF's fullHistoryTsLow is advanced only when users submit a CompactRange request.
However, users may want to advance the fllHistoryTsLow without an immediate compact.
This merge make IncreaseFullHistoryTsLow to a public API so users can advance each CF's fullHistoryTsLow seperately.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9221
Reviewed By: akankshamahajan15
Differential Revision: D33201106
Pulled By: riversand963
fbshipit-source-id: 9cb1d013ba93260f72e16353e693ffee167b47ee
Summary:
The failure looked like this:
```
utilities/backupable/backupable_db_test.cc:3161: Failure
Value of: db_chroot_env_->FileExists(prev_manifest_path).IsNotFound()
Actual: false
Expected: true
```
The failure could be coerced consistently with the following patch:
```
diff --git a/db/db_impl/db_impl_compaction_flush.cc b/db/db_impl/db_impl_compaction_flush.cc
index 80410f671..637636791 100644
--- a/db/db_impl/db_impl_compaction_flush.cc
+++ b/db/db_impl/db_impl_compaction_flush.cc
@@ -2772,6 +2772,8 @@ void DBImpl::BackgroundCallFlush(Env::Priority thread_pri) {
if (job_context.HaveSomethingToClean() ||
job_context.HaveSomethingToDelete() || !log_buffer.IsEmpty()) {
mutex_.Unlock();
+ bg_cv_.SignalAll();
+ sleep(1);
TEST_SYNC_POINT("DBImpl::BackgroundCallFlush:FilesFound");
// Have to flush the info logs before bg_flush_scheduled_--
// because if bg_flush_scheduled_ becomes 0 and the lock is
```
The cause was a familiar problem, which is manual flush/compaction may
return before files they obsoleted are removed. The solution is just to
wait for "scheduled" work to complete, which includes all phases
including cleanup.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9327
Test Plan:
after this PR, even the above patch to coerce the bug cannot
cause the test to fail.
Reviewed By: riversand963
Differential Revision: D33252208
Pulled By: ajkr
fbshipit-source-id: 720a7eaca58c7247d221911fffe3d5e1dbf581e9
Summary:
This change standardizes on a new 16-byte cache key format for
block cache (incl compressed and secondary) and persistent cache (but
not table cache and row cache).
The goal is a really fast cache key with practically ideal stability and
uniqueness properties without external dependencies (e.g. from FileSystem).
A fixed key size of 16 bytes should enable future optimizations to the
concurrent hash table for block cache, which is a heavy CPU user /
bottleneck, but there appears to be measurable performance improvement
even with no changes to LRUCache.
This change replaces a lot of disjointed and ugly code handling cache
keys with calls to a simple, clean new internal API (cache_key.h).
(Preserving the old cache key logic under an option would be very ugly
and likely negate the performance gain of the new approach. Complete
replacement carries some inherent risk, but I think that's acceptable
with sufficient analysis and testing.)
The scheme for encoding new cache keys is complicated but explained
in cache_key.cc.
Also: EndianSwapValue is moved to math.h to be next to other bit
operations. (Explains some new include "math.h".) ReverseBits operation
added and unit tests added to hash_test for both.
Fixes https://github.com/facebook/rocksdb/issues/7405 (presuming a root cause)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9126
Test Plan:
### Basic correctness
Several tests needed updates to work with the new functionality, mostly
because we are no longer relying on filesystem for stable cache keys
so table builders & readers need more context info to agree on cache
keys. This functionality is so core, a huge number of existing tests
exercise the cache key functionality.
### Performance
Create db with
`TEST_TMPDIR=/dev/shm ./db_bench -bloom_bits=10 -benchmarks=fillrandom -num=3000000 -partition_index_and_filters`
And test performance with
`TEST_TMPDIR=/dev/shm ./db_bench -readonly -use_existing_db -bloom_bits=10 -benchmarks=readrandom -num=3000000 -duration=30 -cache_index_and_filter_blocks -cache_size=250000 -threads=4`
using DEBUG_LEVEL=0 and simultaneous before & after runs.
Before ops/sec, avg over 100 runs: 121924
After ops/sec, avg over 100 runs: 125385 (+2.8%)
### Collision probability
I have built a tool, ./cache_bench -stress_cache_key to broadly simulate host-wide cache activity
over many months, by making some pessimistic simplifying assumptions:
* Every generated file has a cache entry for every byte offset in the file (contiguous range of cache keys)
* All of every file is cached for its entire lifetime
We use a simple table with skewed address assignment and replacement on address collision
to simulate files coming & going, with quite a variance (super-Poisson) in ages. Some output
with `./cache_bench -stress_cache_key -sck_keep_bits=40`:
```
Total cache or DBs size: 32TiB Writing 925.926 MiB/s or 76.2939TiB/day
Multiply by 9.22337e+18 to correct for simulation losses (but still assume whole file cached)
```
These come from default settings of 2.5M files per day of 32 MB each, and
`-sck_keep_bits=40` means that to represent a single file, we are only keeping 40 bits of
the 128-bit cache key. With file size of 2\*\*25 contiguous keys (pessimistic), our simulation
is about 2\*\*(128-40-25) or about 9 billion billion times more prone to collision than reality.
More default assumptions, relatively pessimistic:
* 100 DBs in same process (doesn't matter much)
* Re-open DB in same process (new session ID related to old session ID) on average
every 100 files generated
* Restart process (all new session IDs unrelated to old) 24 times per day
After enough data, we get a result at the end:
```
(keep 40 bits) 17 collisions after 2 x 90 days, est 10.5882 days between (9.76592e+19 corrected)
```
If we believe the (pessimistic) simulation and the mathematical generalization, we would need to run a billion machines all for 97 billion days to expect a cache key collision. To help verify that our generalization ("corrected") is robust, we can make our simulation more precise with `-sck_keep_bits=41` and `42`, which takes more running time to get enough data:
```
(keep 41 bits) 16 collisions after 4 x 90 days, est 22.5 days between (1.03763e+20 corrected)
(keep 42 bits) 19 collisions after 10 x 90 days, est 47.3684 days between (1.09224e+20 corrected)
```
The generalized prediction still holds. With the `-sck_randomize` option, we can see that we are beating "random" cache keys (except offsets still non-randomized) by a modest amount (roughly 20x less collision prone than random), which should make us reasonably comfortable even in "degenerate" cases:
```
197 collisions after 1 x 90 days, est 0.456853 days between (4.21372e+18 corrected)
```
I've run other tests to validate other conditions behave as expected, never behaving "worse than random" unless we start chopping off structured data.
Reviewed By: zhichao-cao
Differential Revision: D33171746
Pulled By: pdillinger
fbshipit-source-id: f16a57e369ed37be5e7e33525ace848d0537c88f
Summary:
Pull Request resolved: https://github.com/facebook/rocksdb/pull/9266
This diff adds a new tag `CommitWithTimestamp`. Currently, there is no API to trigger writing
this tag to WAL, thus it is unavailable to users.
This is an ongoing effort to add user-defined timestamp support to write-committed transactions.
This diff also indicates all column families that may potentially participate in the same
transaction must either disable timestamp or have the same timestamp format, since
`CommitWithTimestamp` tag is followed by a single byte-array denoting the commit
timestamp of the transaction. We will enforce this checking in a future diff. We keep this
diff small.
Reviewed By: ltamasi
Differential Revision: D31721350
fbshipit-source-id: e1450811443647feb6ca01adec4c8aaae270ffc6
