Summary:
Given that index value is a BlockHandle, which is basically an <offset, size> pair we can apply delta encoding on the values. The first value at each index restart interval encoded the full BlockHandle but the rest encode only the size. Refer to IndexBlockIter::DecodeCurrentValue for the detail of the encoding. This reduces the index size which helps using the block cache more efficiently. The feature is enabled with using format_version 4.
The feature comes with a bit of cpu overhead which should be paid back by the higher cache hits due to smaller index block size.
Results with sysbench read-only using 4k blocks and using 16 index restart interval:
Format 2:
19585 rocksdb read-only range=100
Format 3:
19569 rocksdb read-only range=100
Format 4:
19352 rocksdb read-only range=100
Pull Request resolved: https://github.com/facebook/rocksdb/pull/3983
Differential Revision: D8361343
Pulled By: maysamyabandeh
fbshipit-source-id: f882ee082322acac32b0072e2bdbb0b5f854e651
Summary:
Do not consider the range tombstone sentinel key as causing 2 adjacent
sstables in a level to overlap. When a range tombstone's end key is the
largest key in an sstable, the sstable's end key is so to a "sentinel"
value that is the smallest key in the next sstable with a sequence
number of kMaxSequenceNumber. This "sentinel" is guaranteed to not
overlap in internal-key space with the next sstable. Unfortunately,
GetOverlappingFiles uses user-keys to determine overlap and was thus
considering 2 adjacent sstables in a level to overlap if they were
separated by this sentinel key. This in turn would cause compactions to
be larger than necessary.
Note that this conflicts with
https://github.com/facebook/rocksdb/pull/2769 and cases
`DBRangeDelTest.CompactionTreatsSplitInputLevelDeletionAtomically` to
fail.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4050
Differential Revision: D8844423
Pulled By: ajkr
fbshipit-source-id: df3f9f1db8f4cff2bff77376b98b83c2ae1d155b
Summary:
https://github.com/facebook/rocksdb/pull/3881 fixed a bug where PinnableSlice pin mmap files which could be deleted with background compaction. This is however a non-issue for ReadOnlyDB when there is no compaction running and max_open_files is -1. This patch reenables the pinning feature for that case.
Closes https://github.com/facebook/rocksdb/pull/4053
Differential Revision: D8662546
Pulled By: maysamyabandeh
fbshipit-source-id: 402962602eb0f644e17822748332999c3af029fd
Summary:
Currently it is not possible to change bloom filter config without restart the db, which is causing a lot of operational complexity for users.
This PR aims to make it possible to dynamically change bloom filter config.
Closes https://github.com/facebook/rocksdb/pull/3601
Differential Revision: D7253114
Pulled By: miasantreble
fbshipit-source-id: f22595437d3e0b86c95918c484502de2ceca120c
Summary:
Allow users to rate limit background work based on read bytes, written bytes, or sum of read and written bytes. Support these by changing the RateLimiter API, so no additional options were needed.
Closes https://github.com/facebook/rocksdb/pull/2433
Differential Revision: D5216946
Pulled By: ajkr
fbshipit-source-id: aec57a8357dbb4bfde2003261094d786d94f724e
Summary:
Previously we returned NotSupported when ingesting files into a database containing any range deletions. This diff adds the support.
- Flush if any memtable contains range deletions overlapping the to-be-ingested file
- Place to-be-ingested file before any level that contains range deletions overlapping it.
- Added support for `Version` to return iterators over range deletions in a given level. Previously, we piggybacked getting range deletions onto `Version`'s `Get()` / `AddIterator()` functions by passing them a `RangeDelAggregator*`. But file ingestion needs to get iterators over range deletions, not populate an aggregator (since the aggregator does collapsing and doesn't expose the actual ranges).
Closes https://github.com/facebook/rocksdb/pull/2370
Differential Revision: D5127648
Pulled By: ajkr
fbshipit-source-id: 816faeb9708adfa5287962bafdde717db56e3f1a
Summary:
Move some files under util/ to new directories env/, monitoring/ options/ and cache/
Closes https://github.com/facebook/rocksdb/pull/2090
Differential Revision: D4833681
Pulled By: siying
fbshipit-source-id: 2fd8bef
Summary:
When we introduced range deletion block, TableCache::Get() and TableCache::NewIterator() each did two table cache lookups, one for range deletion block iterator and another for getting the table reader to which the Get()/NewIterator() is delegated. This extra cache lookup was very CPU-intensive (about 10% overhead in a read-heavy benchmark). We can avoid it by reusing the Cache::Handle created for range deletion block iterator to get the file reader.
Closes https://github.com/facebook/rocksdb/pull/1537
Differential Revision: D4201167
Pulled By: ajkr
fbshipit-source-id: d33ffd8
Summary:
Previously we used TableCache::NewIterator() for multiple purposes (data
block iterator and range deletion iterator), and returned non-ok status in
the data block iterator. In one case where the caller only used the range
deletion block iterator (9e7cf3469b/db/version_set.cc (L965-L973)),
we didn't check/free the data block iterator containing non-ok status, which
caused a valgrind error.
So, this diff decouples creation of data block and range deletion block iterators,
and updates the callers accordingly. Both functions can return non-ok status
in an InternalIterator. Since the non-ok status is returned in an iterator that the
callers will definitely use, it should be more usable/less error-prone.
Closes https://github.com/facebook/rocksdb/pull/1513
Differential Revision: D4181423
Pulled By: ajkr
fbshipit-source-id: 835b8f5
Summary:
This diff introduces RangeDelAggregator, which takes ownership of iterators
provided to it via AddTombstones(). The tombstones are organized in a two-level
map (snapshot stripe -> begin key -> tombstone). Tombstone creation avoids data
copy by holding Slices returned by the iterator, which remain valid thanks to pinning.
For compaction, we create a hierarchical range tombstone iterator with structure
matching the iterator over compaction input data. An aggregator based on that
iterator is used by CompactionIterator to determine which keys are covered by
range tombstones. In case of merge operand, the same aggregator is used by
MergeHelper. Upon finishing each file in the compaction, relevant range tombstones
are added to the output file's range tombstone metablock and file boundaries are
updated accordingly.
To check whether a key is covered by range tombstone, RangeDelAggregator::ShouldDelete()
considers tombstones in the key's snapshot stripe. When this function is used outside of
compaction, it also checks newer stripes, which can contain covering tombstones. Currently
the intra-stripe check involves a linear scan; however, in the future we plan to collapse ranges
within a stripe such that binary search can be used.
RangeDelAggregator::AddToBuilder() adds all range tombstones in the table's key-range
to a new table's range tombstone meta-block. Since range tombstones may fall in the gap
between files, we may need to extend some files' key-ranges. The strategy is (1) first file
extends as far left as possible and other files do not extend left, (2) all files extend right
until either the start of the next file or the end of the last range tombstone in the gap,
whichever comes first.
One other notable change is adding release/move semantics to ScopedArenaIterator
such that it can be used to transfer ownership of an arena-allocated iterator, similar to
how unique_ptr is used for malloc'd data.
Depends on D61473
Test Plan: compaction_iterator_test, mock_table, end-to-end tests in D63927
Reviewers: sdong, IslamAbdelRahman, wanning, yhchiang, lightmark
Reviewed By: lightmark
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D62205
Summary: There's no reference to ImmutableCFOptions elsewhere in /include/rocksdb. ImmutableCFOptions was introduced in this commit (5665e5e285) but later its reference in /include/rocksdb/table.h is removed.
Test Plan:
make all check
Reviewers: IslamAbdelRahman, sdong, yhchiang
Reviewed By: yhchiang
Subscribers: yhchiang, andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D63177
Summary: In T8216281 we decided to disable prefetching the index and filter during opening table handlers during startup (max_open_files = -1).
Test Plan: Rely on `IndexAndFilterBlocksOfNewTableAddedToCache` to guarantee L0 indexes and filters are still cached and change `PinL0IndexAndFilterBlocksTest` to make sure other levels are not cached (maybe add one more test to test we don't cache other levels?)
Reviewers: sdong, andrewkr
Reviewed By: andrewkr
Subscribers: andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D59913
Summary:
Add an option `iterator_readahead_size` to `ReadOptions` to enable
configurable readahead for iterators similar to the corresponding
option for compaction.
Test Plan:
```
make commit_prereq
```
Reviewers: kumar.rangarajan, ott, igor, sdong
Reviewed By: sdong
Subscribers: yiwu, andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D55419
Summary:
When a block based table file is opened, if prefetch_index_and_filter is true, it will prefetch the index and filter blocks, putting them into the block cache.
What this feature adds: when a L0 block based table file is opened, if pin_l0_filter_and_index_blocks_in_cache is true in the options (and prefetch_index_and_filter is true), then the filter and index blocks aren't released back to the block cache at the end of BlockBasedTableReader::Open(). Instead the table reader takes ownership of them, hence pinning them, ie. the LRU cache will never push them out. Meanwhile in the table reader, further accesses will not hit the block cache, thus avoiding lock contention.
Test Plan:
'export TEST_TMPDIR=/dev/shm/ && DISABLE_JEMALLOC=1 OPT=-g make all valgrind_check -j32' is OK.
I didn't run the Java tests, I don't have Java set up on my devserver.
Reviewers: sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D56133
Summary:
When a block based table file is opened, if prefetch_index_and_filter is true, it will prefetch the index and filter blocks, putting them into the block cache.
What this feature adds: when a L0 block based table file is opened, if pin_l0_filter_and_index_blocks_in_cache is true in the options (and prefetch_index_and_filter is true), then the filter and index blocks aren't released back to the block cache at the end of BlockBasedTableReader::Open(). Instead the table reader takes ownership of them, hence pinning them, ie. the LRU cache will never push them out. Meanwhile in the table reader, further accesses will not hit the block cache, thus avoiding lock contention.
When the table reader is destroyed, it releases the pinned blocks (if there were any). This has to happen before the cache is destroyed, so I had to introduce a TableReader::Close(), to guarantee the order of destruction.
Test Plan:
Added two unit tests for this. Existing unit tests run fine (default is pin_l0_filter_and_index_blocks_in_cache=false).
DISABLE_JEMALLOC=1 OPT=-g make all valgrind_check -j32
Mac: OK.
Linux: with D55287 patched in it's OK.
Reviewers: sdong
Reviewed By: sdong
Subscribers: andrewkr, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D54801
Summary:
When Get() or NewIterator() trigger file loads, skip caching the filter block if
(1) optimize_filters_for_hits is set and (2) the file is on the bottommost
level. Also skip checking filters under the same conditions, which means that
for a preloaded file or a file that was trivially-moved to the bottom level, its
filter block will eventually expire from the cache.
- added parameters/instance variables in various places in order to propagate the config ("skip_filters") from version_set to block_based_table_reader
- in BlockBasedTable::Rep, this optimization prevents filter from being loaded when the file is opened simply by setting filter_policy = nullptr
- in BlockBasedTable::Get/BlockBasedTable::NewIterator, this optimization prevents filter from being used (even if it was loaded already) by setting filter = nullptr
Test Plan:
updated unit test:
$ ./db_test --gtest_filter=DBTest.OptimizeFiltersForHits
will also run 'make check'
Reviewers: sdong, igor, paultuckfield, anthony, rven, kradhakrishnan, IslamAbdelRahman, yhchiang
Reviewed By: yhchiang
Subscribers: leveldb
Differential Revision: https://reviews.facebook.net/D51633
Summary:
Separate a new class InternalIterator from class Iterator, when the look-up is done internally, which also means they operate on key with sequence ID and type.
This change will enable potential future optimizations but for now InternalIterator's functions are still the same as Iterator's.
At the same time, separate the cleanup function to a separate class and let both of InternalIterator and Iterator inherit from it.
Test Plan: Run all existing tests.
Reviewers: igor, yhchiang, anthony, kradhakrishnan, IslamAbdelRahman, rven
Reviewed By: rven
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D48549
Summary:
ReadaheadRandomAccessFile acts as a transparent layer on top of RandomAccessFile. When a Read() request is issued, it issues a much bigger request to the OS and caches the result. When a new request comes in and we already have the data cached, it doesn't have to issue any requests to the OS.
We add ReadaheadRandomAccessFile layer only when file is read during compactions.
D45105 was incorrectly closed by Phabricator because I committed it to a separate branch (not master), so I'm resubmitting the diff.
Test Plan: make check
Reviewers: MarkCallaghan, sdong
Reviewed By: sdong
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D45123
Summary:
DBTest.GetProperty was failing occasionally (see task #8131266). The reason was
that the test closed the database before the compaction was done. When the test
reopened the database, RocksDB would schedule a compaction which in turn
created table readers and lead the test to fail the assertion that
rocksdb.estimate-table-readers-mem is 0. In most cases, GetIntProperty() of
rocksdb.estimate-table-readers-mem happened before the compaction created the
table readers, hiding the problem. This patch changes the
WaitForFlushMemTable() to WaitForCompact(). WaitForFlushMemTable() is not
necessary because it is already being called a couple of lines before without
any insertions in-between.
Test Plan:
Insert `usleep(10000);` just after `Reopen(options);` on line 2333 to make the issue more likely, then run:
make db_test && while ./db_test --gtest_filter=DBTest.GetProperty; do true; done
Reviewers: rven, yhchiang, anthony, igor, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D45603
Summary: Currently compaction inputs share the same file descriptor and table reader as other foreground threads. It makes fadvise works less predictable. Add options.new_table_reader_for_compaction_inputs to enforce to create a new file descriptor and new table reader for it.
Test Plan: Add the option.
Reviewers: rven, anthony, kradhakrishnan, IslamAbdelRahman, igor, yhchiang
Reviewed By: igor
Subscribers: igor, MarkCallaghan, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D43311
Summary: In internal stats, remember read latency histogram, if statistics is enabled. It can be retrieved from DB::GetProperty() with "rocksdb.dbstats" property, if it is enabled.
Test Plan: Manually run db_bench and prints out "rocksdb.dbstats" by hand and make sure it prints out as expected
Reviewers: igor, IslamAbdelRahman, rven, kradhakrishnan, anthony, yhchiang
Reviewed By: yhchiang
Subscribers: MarkCallaghan, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D44193
Summary: Measure read latency histogram and put in statistics. Compaction inputs are excluded from it when possible (unfortunately usually no possible as we usually take table reader from table cache.
Test Plan:
Run db_bench and it shows the stats, like:
rocksdb.sst.read.micros statistics Percentiles :=> 50 : 1.238522 95 : 2.529740 99 : 3.912180
Reviewers: kradhakrishnan, rven, anthony, IslamAbdelRahman, MarkCallaghan, yhchiang
Reviewed By: yhchiang
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D43275
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:
Intead of passing callback function pointer and its arg on Table::Get()
interface, passing GetContext. This makes the interface cleaner and
possible better perf. Also adding a fast pass for SaveValue()
Test Plan: make all check
Reviewers: igor, yhchiang, sdong
Reviewed By: sdong
Subscribers: leveldb
Differential Revision: https://reviews.facebook.net/D24057
Summary:
As a preparation to support updating some options dynamically, I'd like
to first introduce ImmutableOptions, which is a subset of Options that
cannot be changed during the course of a DB lifetime without restart.
ColumnFamily will keep both Options and ImmutableOptions. Any component
below ColumnFamily should only take ImmutableOptions in their
constructor. Other options should be taken from APIs, which will be
allowed to adjust dynamically.
I am yet to make changes to memtable and other related classes to take
ImmutableOptions in their ctor. That can be done in a seprate diff as
this one is already pretty big.
Test Plan: make all check
Reviewers: yhchiang, igor, sdong
Reviewed By: sdong
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D22545
Summary:
Add a DB Property "rocksdb.estimate-table-readers-mem" to return estimated memory usage by all loaded table readers, other than allocated from block cache.
Refactor the property codes to allow getting property from a version, with DB mutex not acquired.
Test Plan: Add several checks of this new property in existing codes for various cases.
Reviewers: yhchiang, ljin
Reviewed By: ljin
Subscribers: xjin, igor, leveldb
Differential Revision: https://reviews.facebook.net/D20733
Summary:
This patch adds a target size parameter in options.db_paths and universal compaction will base it to determine which DB path to place a new file.
Level-style stays the same.
Test Plan: Add new unit tests
Reviewers: ljin, yhchiang
Reviewed By: yhchiang
Subscribers: MarkCallaghan, dhruba, igor, leveldb
Differential Revision: https://reviews.facebook.net/D19869
Summary:
In this patch, we allow RocksDB to support multiple DB paths internally.
No user interface is supported yet so this patch is silent to users.
Test Plan: make all check
Reviewers: igor, haobo, ljin, yhchiang
Reviewed By: yhchiang
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D18921
Summary:
As discussed in our internal group, we don't get much use of seek compaction at the moment, while it's making code more complicated and slower in some cases.
This diff removes seek compaction and (hopefully) all code that was introduced to support seek compaction.
There is one test case that relied on didIO information. I'll try to find another way to implement it.
Test Plan: make check
Reviewers: sdong, haobo, yhchiang, ljin, dhruba
Reviewed By: ljin
Subscribers: leveldb
Differential Revision: https://reviews.facebook.net/D19161
Summary:
We added multiple fields to FileMetaData recently and are planning to add more.
This refactoring separate the minimum information for accessing the file. This object is copyable (FileMetaData is not copyable since the ref counter). I hope this refactoring can enable further improvements:
(1) use it to design a more efficient data structure to speed up read queries.
(2) in the future, when we add information of storage level, we can easily do the encoding, instead of enlarge this structure, which might expand memory work set for file meta data.
The definition is same as current EncodedFileMetaData used in two level iterator, so now the logic in two level iterator is easier to understand.
Test Plan: make all check
Reviewers: haobo, igor, ljin
Reviewed By: ljin
Subscribers: leveldb, dhruba, yhchiang
Differential Revision: https://reviews.facebook.net/D18933
Summary:
In this patch, try to allocate the whole iterator tree starting from DBIter from an arena
1. ArenaWrappedDBIter is created when serves as the entry point of an iterator tree, with an arena in it.
2. Add an option to create iterator from arena for following iterators: DBIter, MergingIterator, MemtableIterator, all mem table's iterators, all table reader's iterators and two level iterator.
3. MergeIteratorBuilder is created to incrementally build the tree of internal iterators. It is passed to mem table list and version set and add iterators to it.
Limitations:
(1) Only DB::NewIterator() without tailing uses the arena. Other cases, including readonly DB and compactions are still from malloc
(2) Two level iterator itself is allocated in arena, but not iterators inside it.
Test Plan: make all check
Reviewers: ljin, haobo
Reviewed By: haobo
Subscribers: leveldb, dhruba, yhchiang, igor
Differential Revision: https://reviews.facebook.net/D18513
Summary: It is a regression valgrind bug caused by using FileMetaData as index block handle. One of the fields of FileMetaData is not initialized after being contructed and copied, but I'm not able to find which one. Also, I realized that it's not a good idea to use FileMetaData as in TwoLevelIterator::InitDataBlock(), a copied FileMetaData can be compared with the one in version set byte by byte, but the refs can be changed. Also comparing such a large structure is slightly more expensive. Use a simpler structure instead
Test Plan:
Run the failing valgrind test (Harness.RandomizedLongDB)
make all check
Reviewers: igor, haobo, ljin
Reviewed By: igor
CC: yhchiang, leveldb
Differential Revision: https://reviews.facebook.net/D17409
Summary:
This is the part that was not finished when doing the Options.max_num_files=-1 feature. For iterating non level0 SST files (which was done using two level iterator), table cache is not bypassed. With this patch, the leftover feature is done.
Test Plan: make all check; change Options.max_num_files=-1 in one of the tests to cover the codes.
Reviewers: haobo, igor, dhruba, ljin, yhchiang
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D17001
Summary: Provide a public API for users to access the table properties for each SSTable.
Test Plan: Added a unit tests to test the function correctness under differnet conditions.
Reviewers: haobo, dhruba, sdong
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D16083
Summary:
Lots of code expects Options on construction/function call. My original idea was to split Options argument into ColumnFamilyOptions and DBOptions (the latter only if needed). However, this will require huge code changes very deep in the stack.
The better idea is to have ColumnFamilyData hold both ColumnFamilyOptions and Options. ColumnFamilyData::Options would be constructed from DBOptions (same for each column family) and ColumnFamilyOptions (different for each column family)
Now when we construct a class or call any method that requires Options, we can just push him ColumnFamilyData::Options and be sure that it's using column-family-specific settings.
Test Plan: make check
Reviewers: dhruba, haobo, kailiu, sdong
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15927
Summary:
Adapting table cache to column families is interesting. We want table cache to be global LRU, so if some column families are use not as often as others, we want them to be evicted from cache. However, current TableCache object also constructs tables on its own. If table is not found in the cache, TableCache automatically creates new table. We want each column family to be able to specify different table factory.
To solve the problem, we still have a single LRU, but we provide the LRUCache object to TableCache on construction. We have one TableCache per column family, but the underyling cache is shared by all TableCache objects.
This allows us to have a global LRU, but still be able to support different table factories for different column families. Also, in the future it will also be able to support different directories for different column families.
Test Plan: make check
Reviewers: dhruba, haobo, kailiu, sdong
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15915
Summary:
Previous I am too ambitious to hide every detail about table factory
to internal api. However, we cannot pass the compilatoin for external
users since we use table factory as the shared_ptr, which requires
the definition of table factory's destructor.
Test Plan: make check;
Reviewers: sdong, haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15861
Summary: In PlainTable, use one single byte to represent 8 bytes of internal bytes, if seqID = 0 and it is value type (which should be common for bottom most files). It is to save 7 bytes for uncompressed cases.
Test Plan: make all check
Reviewers: haobo, dhruba, kailiu
Reviewed By: haobo
CC: igor, leveldb
Differential Revision: https://reviews.facebook.net/D15489
Summary:
In some use cases, table readers for all live files should always be cached. In that case, there will be an opportunity to avoid the table cache look-up while Get() and NewIterator().
We define options.max_open_files = -1 to be the mode that table readers for live files will always be kept. In that mode, table readers are cached in FileMetaData (with a reference count hold in table cache). So that when executing table_cache.Get() and table_cache.newInterator(), LRU cache checking can be by-passed, to reduce latency.
Test Plan: add a test case in db_test
Reviewers: haobo, kailiu
Reviewed By: haobo
CC: dhruba, igor, leveldb
Differential Revision: https://reviews.facebook.net/D15039
