- 10 10月, 2015 5 次提交
-
-
由 krad 提交于
Summary: The parser will help parse the output in order to send meaningful notifications. Test Plan: Manual testing Reviewers: CC: leveldb@ Task ID: #6968635 Blame Rev:
-
由 agiardullo 提交于
Summary: Support for Transaction::CreateSnapshotOnNextOperation(). This is to fix a write-conflict race-condition that Yoshinori was running into when testing MyRocks with LinkBench. Test Plan: New tests Reviewers: yhchiang, spetrunia, rven, igor, yoshinorim, sdong Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48099
-
由 agiardullo 提交于
Summary: MyRocks reported some perfomance issues when inserting many keys into a transaction due to the cost of inserting new keys into WriteBatchWithIndex. Frequently, they don't even need the keys to be indexed as they don't need to read them back. DisableIndexing() can be used to avoid the cost of indexing. I also plan on eventually investigating if we can improve WriteBatchWithIndex performance. But even if we improved the perf here, it is still beneficial to be able to disable the indexing all together for large transactions. Test Plan: unit test Reviewers: igor, rven, yoshinorim, spetrunia, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48471
-
由 sdong 提交于
Summary: Pass column family ID through TablePropertiesCollectorFactory::CreateTablePropertiesCollector() so that users can identify which column family this file is for and handle it differently. Test Plan: Add unit test scenarios in tests related to table properties collectors to verify the information passed in is correct. Reviewers: rven, yhchiang, anthony, kradhakrishnan, igor, IslamAbdelRahman Reviewed By: IslamAbdelRahman Subscribers: yoshinorim, leveldb, dhruba Differential Revision: https://reviews.facebook.net/D48411
-
由 sdong 提交于
Summary: Now based on environment, DBTest.AggregatedTableProperties has a possibility of issuing a L0->L1 compaction after reopening and the results are not what we expected. We tune the L0 compaction trigger to make it less likely to happen. Test Plan: I can't repro the failure but I think the change is better. Just run the test and make sure it passes. Reviewers: kradhakrishnan, yhchiang, igor Reviewed By: igor Subscribers: leveldb, dhruba Differential Revision: https://reviews.facebook.net/D48423
-
- 09 10月, 2015 8 次提交
-
-
由 sdong 提交于
Summary: With this commit, we add a new format in manifest when adding a new file. Now path ID and need-compaction hint are first two customized fields. Test Plan: Add a test case in version_edit_test to verify the encoding and decoding logic. Add a unit test in db_test to verify need compaction is persistent after DB restarting. Reviewers: kradhakrishnan, anthony, IslamAbdelRahman, yhchiang, rven, igor Reviewed By: igor Subscribers: javigon, leveldb, dhruba Differential Revision: https://reviews.facebook.net/D48123
-
由 Igor Canadi 提交于
Install snappy headers to standard locations using yum, so that build…
-
由 Vivek Nelamangala 提交于
Install snappy headers to standard locations using yum, so that build_tools/build_detect_platform sets -DSNAPPY flag to g++ . Current jars of rocksdb do no have snappy compression avaliable .
-
由 Igor Canadi 提交于
Summary: I see a lot of users compiling RocksDB with `make` or `make all` and then using those binaries in production. They end up running debug builds :( This diff makes it harder for them: 1. I added an explicit warning to INSTALL.md 2. When you compile with `make all`, your resulting library will be librocksdb_debug.a 3. I also print out a warning when you compile in debug mode. Hopefully should be enough :) Test Plan: none Reviewers: rven, yhchiang, kradhakrishnan, anthony, sdong, dhruba, IslamAbdelRahman Reviewed By: IslamAbdelRahman Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48093
-
由 sdong 提交于
Summary: Add the column family ID to compaction filter context, so it is easier for compaction filter to apply different logic for different column families. Test Plan: Add a unit test to verify the column family ID passed is correct. Reviewers: rven, yhchiang, igor Reviewed By: igor Subscribers: yoshinorim, leveldb, dhruba Differential Revision: https://reviews.facebook.net/D48357
-
由 Yueh-Hsuan Chiang 提交于
Summary: Change RocksDB version to 4.1 Test Plan: no code change. Reviewers: sdong, anthony, IslamAbdelRahman, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48387
-
由 krad 提交于
Summary: Enabling CI crash test jobs Test Plan: Manual testing Reviewers: CC: leveldb@ Task ID: #8635140 Blame Rev:
-
由 Igor Canadi 提交于
Summary: Two changes: 1. remove *V2 filter stuff. we deprecated that a while ago 2. clarify what happens when user sets max_subcompactions to bigger than 1 Test Plan: none Reviewers: yhchiang, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47871
-
- 08 10月, 2015 8 次提交
-
-
由 Islam AbdelRahman 提交于
Summary: Add 2 new counters BLOCK_CACHE_BYTES_WRITE, BLOCK_CACHE_BYTES_READ to keep track of how many bytes were written to the cache and how many bytes that we read from cache Test Plan: make check Reviewers: sdong, yhchiang, igor Reviewed By: igor Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D48195
-
由 Igor Canadi 提交于
Summary: unity_test will compile db_test with unity.a instead of librocksdb.a. This will test both the compilation and some small amount of runtime. Test Plan: This is a test :) Reviewers: kradhakrishnan Reviewed By: kradhakrishnan Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48297
-
由 dyniusz 提交于
Summary: hit and miss bloom filter stats for memtable and SST stats added to perf_context struct key matches and prefix matches combined into one stat Test Plan: unit test veryfing the functionality added, see BloomStatsTest in db_test.cc for details Reviewers: yhchiang, igor, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47859
-
由 Igor Canadi 提交于
Summary: If a platform doesn't have ROCKSDB_FALLOCATE_PRESENT, then compiler complains: util/env_posix.cc:354:8: error: private field 'allow_fallocate_' is not used [-Werror,-Wunused-private-field] This was caught by travis. Test Plan: compiles with ROCKSDB_FALLOCATE_PRESENT. Reviewers: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48327
-
由 Igor Canadi 提交于
Summary: Travis is complaining against using {} to initialize KVMap: https://travis-ci.org/facebook/rocksdb/jobs/84132600 db/compaction_job_test.cc:526:26: error: chosen constructor is explicit in copy-initialization RunCompaction({files}, {}); This diff should fix it Test Plan: travis Reviewers: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48309
-
由 Lakshmi Narayanan 提交于
Summary: Added boolean variable to guard fallocate() calls. Set to false to prevent space leaks when tests fail. Test Plan: Compliles Set to false and ran log device tests Reviewers: sdong, lovro, igor Reviewed By: igor Subscribers: dhruba Differential Revision: https://reviews.facebook.net/D48027
-
由 Igor Canadi 提交于
Summary: This diff avoids failing travis test when clang-format fails. We should keep the test green even if clang-format fails. It would be good to look at travis output for pull requests to make sure they're following the good code style. However, we broke travis test 3 times today because of formatting issues. We should keep travis test green if the only thing wrong is the formatting. Test Plan: none Reviewers: sdong, yhchiang, anthony Reviewed By: anthony Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48291
-
由 Igor Canadi 提交于
Summary: Since Andres' internship is over, I took over https://reviews.facebook.net/D42555 and rebased and simplified it a bit. The behavior in this diff is a bit simpler than in D42555: * only merge operators are passed through FilterMergeValue(). If fitler function returns true, the merge operator is ignored * compaction filter is *not* called on: 1) results of merge operations and 2) base values that are getting merged with merge operands (the second case was also true in previous diff) Do we also need a compaction filter to get called on merge results? Test Plan: make && make check Reviewers: lovro, tnovak, rven, yhchiang, sdong Reviewed By: sdong Subscribers: noetzli, kolmike, leveldb, dhruba, sdong Differential Revision: https://reviews.facebook.net/D47847
-
- 07 10月, 2015 7 次提交
-
-
由 krad 提交于
Summary: Unity CI job is broken. Disabling. Test Plan: Manual test Reviewers: CC: leveldb@ Task ID: #8632483 Blame Rev:
-
由 dyniusz 提交于
Summary: Handle SST files with both ".sst" and ".ldb" suffix. This enables user to migrate from leveldb to rocksdb. Test Plan: Added unit test with DB operating on SSTs with names schema. See db/dc_test.cc:SSTsWithLdbSuffixHandling for details Reviewers: yhchiang, sdong, igor Reviewed By: igor Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48003
-
由 Siying Dong 提交于
FIx up build options
-
由 Igor Canadi 提交于
Mmap reads should not return error if reading past file
-
由 Dmitri Smirnov 提交于
Summary: This mirrors https://reviews.facebook.net/D45645 Currently, mmap returns IOError when user tries to read data past the end of the file. This diff changes the behavior. Now, we return just the bytes that we can, and report the size we returned via a Slice result. This is consistent with non-mmap behavior and also pread() system call.
-
由 Dmitri Smirnov 提交于
-
由 krad 提交于
Summary: Sandcastle does not raise a task if assignee is specified. Removing the assignee will force the system to use the oncall. Test Plan: Manual Reviewers: CC: leveldb@ Task ID: #6968635 Blame Rev:
-
- 06 10月, 2015 4 次提交
-
-
由 Islam AbdelRahman 提交于
Summary: Refactor dump_tool and undump_tool so that it's possible to use them with customized options for example setting a specific comparator similar to what Dragon is doing with the LdbTool https://phabricator.fb.com/diffusion/FBCODE/browse/master/dragon/tools/Ldb.cpp Test Plan: compiles used it to dump / undump a dragon shard Reviewers: sdong, yhchiang, igor Reviewed By: igor Subscribers: dhruba, adsharma Differential Revision: https://reviews.facebook.net/D47853
-
由 Igor Canadi 提交于
Summary: The function GetBoundaryKeys() returns the smallest key from the first file and largest key from the last file. This is good for any level >0, but it's not correct for level 0. In level 0, files can overlap, so we need to check all files for boundary keys. This bug can cause wrong value for bottommost_level in compaction (value of true, although correct is false), which means we can set sequence numbers to 0 even if the key is not the oldest one in the database. Herman reported corruption while testing MyRocks. Fortunately, the patch that added the bug was not released yet. Test Plan: added a new test to compaction_picker_test. Reviewers: hermanlee4, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48201
-
由 Igor Canadi 提交于
Summary: Latest travis failed because of corruption test TableFileIndexData: https://travis-ci.org/facebook/rocksdb/jobs/83732558 This diff makes the test more explicit: 1. create two files 2. corrupt the second's file index 3. expect to get only 5000 keys when range scanning Test Plan: the test is still passing :) Reviewers: sdong, rven, yhchiang, kradhakrishnan, IslamAbdelRahman, anthony Reviewed By: anthony Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48183
-
由 Igor Canadi 提交于
Summary: Fixes the valgrind warning "Conditional jump or move depends on uninitialised value(s)" Test Plan: valgrind test, no more warning Reviewers: sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D48177
-
- 03 10月, 2015 5 次提交
-
-
由 Yueh-Hsuan Chiang 提交于
Summary: RocksDBOptionsParser now supports CompressionType and the following pointer-typed options in RocksDBOptionParser for sanity check: prefix_extractor table_factory comparator compaction_filter compaction_filter_factory merge_operator memtable_factory In the RocksDB Options file, only high level information about pointer-typed options are serialized, and those information is only used for verification / sanity check purpose. Test Plan: added more tests in options_test Reviewers: igor, IslamAbdelRahman, sdong, anthony Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47925
-
由 Igor Canadi 提交于
Summary: To support a new MongoDB capability, we need to make sure that we don't do any IO for a short period of time. For background, see: * https://jira.mongodb.org/browse/SERVER-20704 * https://jira.mongodb.org/browse/SERVER-18899 To implement that, I add a new API calls PauseBackgroundWork() and ContinueBackgroundWork() which reuse the capability we already have in place for RefitLevel() function. Test Plan: Added a new test in db_test. Made sure that test fails when PauseBackgroundWork() is commented out. Reviewers: IslamAbdelRahman, sdong Reviewed By: sdong Subscribers: dhruba, leveldb Differential Revision: https://reviews.facebook.net/D47901
-
由 krad 提交于
Summary: Added following capability (1) Ability to do configurable reports using args (2) Support to generating a task on failure (3) Ability to configure email trigger condition Test Plan: Manual testing Reviewers: CC:leveldb@ Task ID: #6968635 Blame Rev:
-
由 Igor Canadi 提交于
Improve test running scripts
-
由 Dmitri Smirnov 提交于
Introduce proper command line arguments so we can control the script Add appveyor support Add an ability to run all other (non db_tests) test executables in parallel Use .NET HashSet instead of empty valued hashtable to improve the looks TODO: Some of the tests do not use GTests and need to improve log parsing
-
- 01 10月, 2015 3 次提交
-
-
由 Igor Canadi 提交于
New amalgamation target
-
由 Evan Shaw 提交于
This commit adds two new targets to the Makefile: rocksdb.cc and rocksdb.h These files, when combined with the c.h header, are a self-contained RocksDB source distribution called an amalgamation. (The name comes from SQLite's, which is similar in concept.) The main benefit of an amalgamation is that it's very easy to drop into a new project. It also compiles faster compared to compiling individual source files and potentially gives the compiler more opportunity to make optimizations since it can see all functions at once. rocksdb.cc and rocksdb.h are generated by a new script, amalgamate.py. A detailed description of how amalgamate.py works is in a comment at the top of the file. There are also some small changes to existing files to enable the amalgamation: * Use quotes for includes in unity build * Fix an old header inclusion in util/xfunc.cc * Move some includes outside ifdef in util/env_hdfs.cc * Separate out tool sources in Makefile so they won't be included in unity.cc * Unity build now produces a static library Closes #733
-
由 Igor Canadi 提交于
Improve CI build and fix Windows build breakage
-