Commit Graph

4124 Commits

Author SHA1 Message Date
agiardullo
03b08ba9a9 Return MergeInProgress when fetching from transactions or WBWI with overwrite_key
Summary:
WriteBatchWithIndex::GetFromBatchAndDB only works correctly for overwrite_key=false.  Transactions use overwrite_key=true (since WriteBatchWithIndex::GetIteratorWithBase only works when overwrite_key=true).  So currently, Transactions could return incorrectly merged results when calling Get/GetForUpdate().

Until a permanent fix can be put in place, Transaction::Get[ForUpdate] and WriteBatchWithIndex::GetFromBatch[AndDB] will now return MergeInProgress if the most recent write to a key in the batch is a Merge.

Test Plan: more tests

Reviewers: sdong, yhchiang, rven, igor

Reviewed By: igor

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47817
2015-09-30 11:14:42 -07:00
Islam AbdelRahman
c29af48d3e Add max_file_opening_threads to db_bench
Summary: Add an option to db_bench for max_file_opening_threads

Test Plan: compile and run db_bench

Reviewers: sdong, yhchiang, igor

Reviewed By: igor

Subscribers: dhruba, paultuckfield

Differential Revision: https://reviews.facebook.net/D47811
2015-09-30 09:51:31 -07:00
Yueh-Hsuan Chiang
da1cf8a9bc Add a missing check for deprecated options in options_helper.cc
Summary: Add a missing check for deprecated options in options_helper.cc

Test Plan: options_test

Reviewers: sdong, anthony, IslamAbdelRahman, igor

Reviewed By: igor

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47793
2015-09-29 17:58:00 -07:00
agiardullo
5a51fa907b Fix accidental object copy in transactions
Summary: Should have used auto& instead of auto.  Also needed to change the code a bit due to const correctness.

Test Plan: unit tests

Reviewers: sdong, igor, yoshinorim, yhchiang

Reviewed By: yhchiang

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47787
2015-09-29 17:36:56 -07:00
Yueh-Hsuan Chiang
1e73b11af7 Better handling of deprecated options in RocksDBOptionsParser
Summary:
Previously, we treat deprecated options as normal options in
RocksDBOptionsParser.  However, these deprecated options should
not be verified and serialized.

Test Plan: options_test

Reviewers: igor, sdong, IslamAbdelRahman, anthony

Reviewed By: anthony

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47775
2015-09-29 17:13:02 -07:00
Yueh-Hsuan Chiang
a8b8295d18 Fixed a compile warning in options_test.cc under clang
Summary:
Fixed the following compile warning in options_test.cc under clang
util/options_test.cc:94:12: error: 'Skip' overrides a member function but is not marked 'override' [-Werror,-Winconsistent-missing-override]
    Status Skip(uint64_t n) {
           ^
./include/rocksdb/env.h:368:18: note: overridden virtual function is here
  virtual Status Skip(uint64_t n) = 0;
                 ^

Test Plan: options_test

Reviewers: igor, sdong, anthony, IslamAbdelRahman

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47763
2015-09-29 14:59:02 -07:00
Yueh-Hsuan Chiang
74b100ac17 RocksDB Options file format and its serialization / deserialization.
Summary:
This patch defines the format of RocksDB options file, which
follows the INI file format, and implements functions for its
serialization and deserialization.  An example RocksDB options
file can be found in examples/rocksdb_option_file_example.ini.

A typical RocksDB options file has three sections, which are
Version, DBOptions, and more than one CFOptions.  The RocksDB
options file in general follows the basic INI file format
with the following extensions / modifications:
 * Escaped characters
   We escaped the following characters:
    - \n -- line feed - new line
    - \r -- carriage return
    - \\ -- backslash \
    - \: -- colon symbol :
    - \# -- hash tag #
 * Comments
   We support # style comments.  Comments can appear at the ending
   part of a line.
 * Statements
   A statement is of the form option_name = value.
   Each statement contains a '=', where extra white-spaces
   are supported. However, we don't support multi-lined statement.
   Furthermore, each line can only contain at most one statement.
 * Section
   Sections are of the form [SecitonTitle "SectionArgument"],
   where section argument is optional.
 * List
   We use colon-separated string to represent a list.
   For instance, n1:n2:n3:n4 is a list containing four values.

Below is an example of a RocksDB options file:

[Version]
  rocksdb_version=4.0.0
  options_file_version=1.0
[DBOptions]
  max_open_files=12345
  max_background_flushes=301
[CFOptions "default"]
[CFOptions "the second column family"]
[CFOptions "the third column family"]

Test Plan: Added many tests in options_test.cc

Reviewers: igor, IslamAbdelRahman, sdong, anthony

Reviewed By: anthony

Subscribers: maykov, dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D46059
2015-09-29 14:42:40 -07:00
Igor Canadi
75134f7562 Merge pull request #741 from yuslepukhin/test_appveyor_baseline
Improve CI build and build switches
2015-09-29 13:16:40 -07:00
Dmitri Smirnov
2e7506d82c Improve CI build and build switches
Add an optimized build config switch for faster test runs
  Change compiler options to introduce more opitmizations and be more inline with MS internal switches.
  Make appveyor build to utilize all the avaiable cores on the VM (parallel)
  Introduce new appveyor configuration for daily test runs as it would take too long
  to run db_test after each checkin even in paralell.
  With some exclusions we make it in 38 minutes. We currently fail to install ramdisk during the build.
  Add a powershell script to faicilitate paralell run for db_test cases.
2015-09-29 12:22:48 -07:00
Igor Canadi
1eff1834b2 Remove non-existing functions. Closes #680
Summary: See https://github.com/facebook/rocksdb/issues/680

Test Plan: compiles

Reviewers: sdong, yhchiang, anthony

Reviewed By: anthony

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47721
2015-09-29 09:34:42 -07:00
Islam AbdelRahman
16d1ba7001 Clear SyncPoint Trace in DeleteSchedulerTests
Summary: DeleteSchedulerTests is running the same test with different rates, After the first iteraton sync points become useless because ClearTrace was not being called

Test Plan: Run the test

Reviewers: sdong, yhchiang, igor

Reviewed By: igor

Subscribers: dhruba

Differential Revision: https://reviews.facebook.net/D47709
2015-09-28 15:58:50 -07:00
Yueh-Hsuan Chiang
30f74fa964 Make CompactionJobStatsTest.UniversalCompactionTest more robust
Summary:
CompactionJobStatsTest.UniversalCompactionTest assumes compaction
kicks in when the number of L0 files equals to the compaction trigger.
However, in some case, the compaction might not catch up the write
speed and thus compaction might not kick in until the number of L0 files
is GREATER than the compaction trigger.

This patch tries to fix this corner case by making the Put thread wait
for a potential compaction whenever it flushes.

Test Plan: ./compaction_job_stats_test

Reviewers: sdong, anthony, IslamAbdelRahman, igor

Subscribers: dhruba

Differential Revision: https://reviews.facebook.net/D47589
2015-09-28 13:55:53 -07:00
agiardullo
afe0dc539b SingleDelete support for Transactions
Summary: Transactional SingleDelete is needed for MyRocks.  Note: This diff requires D47529.

Test Plan: Added some new tests in this diff as well as more tests added in D47529

Reviewers: rven, sdong, igor, yhchiang

Reviewed By: yhchiang

Subscribers: yoshinorim, dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47535
2015-09-28 12:14:26 -07:00
Yueh-Hsuan Chiang
a263002a36 Fixed a tsan warning in db_stress.cc
Summary:
Fixed the following tsan warning in db_stress.cc

  WARNING: ThreadSanitizer: data race (pid=3163194)
  Read of size 8 at 0x7fd1797cb518 by thread T32:
    #0 VerifyDb tools/db_stress.cc:1731 (db_stress+0x000000040674)
    #1 rocksdb::StressTest::ThreadBody(void*) tools/db_stress.cc:1191 (db_stress+0x0000000625a9)
    #2 StartThreadWrapper util/env_posix.cc:1648 (db_stress+0x00000028bbbd)

  Previous write of size 8 at 0x7fd1797cb518 by thread T31:
    #0 VerifyDb tools/db_stress.cc:1726 (db_stress+0x00000004072a)
    #1 rocksdb::StressTest::ThreadBody(void*) tools/db_stress.cc:1191 (db_stress+0x0000000625a9)
    #2 StartThreadWrapper util/env_posix.cc:1648 (db_stress+0x00000028bbbd)

The cause is that in VerifyDb(), the static local const variable long max_key
can be read and written at the same time.  This patch fixed it by making it
non-static.

Test Plan: db_stress

Reviewers: igor, sdong, IslamAbdelRahman, anthony

Reviewed By: anthony

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47703
2015-09-28 12:06:43 -07:00
Yueh-Hsuan Chiang
e4861e7d68 Fixed a compile error in util/arena.h
Summary:
Fixed the compile error in util/arena.h caused by not
including TLB related header.

Test Plan: make db_stress

Reviewers: igor, sdong, anthony, IslamAbdelRahman

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47697
2015-09-28 11:51:32 -07:00
Yueh-Hsuan Chiang
0fdb4f1688 Fixed a compile warning in util/arena.cc when hugetlb is not supported.
Summary:
Fixed the following compile warning when hugetlb is not supported.

    ./util/arena.h:102:10: error: private field 'hugetlb_size_' is not used [-Werror,-Wunused-private-field]
      size_t hugetlb_size_ = 0;
             ^
    1 error generated.

Test Plan: make db_stress

Reviewers: igor, sdong, anthony, IslamAbdelRahman

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47691
2015-09-28 11:43:04 -07:00
Igor Canadi
94ac8826c4 Merge pull request #737 from mlin/readonly-syncwal
Override DBImplReadOnly::SyncWAL() to return NotSupported.
2015-09-26 10:34:54 -07:00
Mike Lin
60fa9cf0b5 Override DBImplReadOnly::SyncWAL() to return NotSupported. Previously, calling it caused program abort. 2015-09-25 21:25:30 -07:00
krad
7df348b405 Minor fix to CI job definition
Summary: Added email notification for a few jobs and fixed a bug in
no_compression script

Test Plan: Manual test

Reviewers:

CC: leveldb@

Task ID: #6968635

Blame Rev:
2015-09-25 14:54:51 -07:00
Igor Canadi
dac3f22b77 Fix the test failure
Summary: AllocateFromHugePage() can return nullptr, and then we need to try to allocate the block with AllocateNewBlock()

Test Plan: arena_test

Reviewers: sdong

Reviewed By: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47607
2015-09-25 13:55:11 -07:00
Yueh-Hsuan Chiang
63e0f86797 Fixed a bug which causes rocksdb.flush.write.bytes stat is always zero
Summary: Fixed a bug which causes rocksdb.flush.write.bytes stat is always zero

Test Plan: augment existing db_test

Reviewers: sdong, anthony, IslamAbdelRahman, igor

Reviewed By: igor

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47595
2015-09-25 13:34:49 -07:00
Igor Canadi
7ee445dd68 Fix the compile warning
Summary: clang is a bit confused, see here: https://travis-ci.org/facebook/rocksdb/jobs/82214750

Test Plan: travis CI

Reviewers: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47601
2015-09-25 13:17:19 -07:00
Igor Canadi
174e2be5f0 Merge pull request #735 from jsteemann/fix-potential-leak-in-allocate
prevent potential memleaks in Arena::Allocate*()
2015-09-25 12:35:15 -07:00
jsteemann
aa58958d38 prevent potential memleaks in Arena::Allocate*()
The previous memory allocation procedures tried to allocate memory
via `new` or `mmap` and inserted the pointer to the memory into an
std::vector afterwards. In case `new` or `mmap` threw or returned
a nullptr, no memory was leaking. If `new` or `mmap` worked ok, the
following `vector::push_back` could still fail and throw an exception.
In this case, the memory just allocated was leaked.

The fix is to reserve space in the target memory pointer block
beforehand. If this throws, then no memory is allocated nor leaked.
If the reserve works but the actual allocation fails, still no
memory is leaked, only the target vector will have space for at
least one more element than actually required (but this may be
reused for the next allocation)
2015-09-25 21:23:40 +02:00
agiardullo
25fd743d75 Fix SingleDelete support in WriteBatchWithIndex
Summary: Fixed some  bugs in using SingleDelete on a WriteBatchWithIndex and added some tests.

Test Plan: new tests

Reviewers: sdong, yhchiang, rven, kradhakrishnan, IslamAbdelRahman, igor

Reviewed By: igor

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47529
2015-09-25 12:23:07 -07:00
Yueh-Hsuan Chiang
b6aa3f962d Fixed a memory leak issue in DBTest.UnremovableSingleDelete
Summary: Fixed a memory leak issue in DBTest.UnremovableSingleDelete

Test Plan: valgrind --error-exitcode=2 --leak-check=full ./db_test --gtest_filter="*UnremovableSingleDelete*"

Reviewers: sdong, anthony, IslamAbdelRahman, igor

Reviewed By: igor

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47583
2015-09-25 12:07:32 -07:00
Igor Canadi
7b7b5d9f18 [minor] Reuse SleepingBackgroundTask
Summary: As title

Test Plan: make check

Reviewers: yhchiang, sdong

Reviewed By: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D46983
2015-09-25 10:29:44 -07:00
krad
e01f32cc3c Parameterizing email id
Summary: Removing the leveldb@fb.com and making it a command line argument

Test Plan: Manual test

Reviewers:

CC: leveldb@

Task ID: #6968635

Blame Rev:
2015-09-24 15:18:36 -07:00
Mayank Pundir
c58bac701c Fix valgrind failure due to memory leaks
Summary: Test cases for IsBottommostLevel function create FileMetaData objects which were not getting deleted in the destructor.

Test Plan: Valgrind check on compaction_picker_test

Reviewers: yhchiang, igor, sdong

Subscribers: rven, kradhakrishnan, IslamAbdelRahman, dhruba, anthony

Differential Revision: https://reviews.facebook.net/D47463
2015-09-23 17:41:42 -07:00
sdong
a70d08ec07 Fix the bug of using freed memory introduced by recent plain table reader patch
Summary: Recent patch introduced a bug that if non-mmap mode is used, in prefix encoding case, there is a resizing of cur_key_ within the same prefix, we still read prefix from the released buffer. It fails ASAN tests and this commit fixes it.

Test Plan: Run the ASAN tests for the failing test case.

Reviewers: IslamAbdelRahman, yhchiang, anthony, igor, kradhakrishnan, rven

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D47457
2015-09-23 16:16:26 -07:00
krad
628216fc1f Simplifying valgrind testing
Summary: Our valgrind testing is buggy and cumbersome in terms of locating the
error. It originates from the fact we accumulate all output for the tests. It is
extremely hard to locate the point of error.

The communication between valgrind and the script is not sturdy. We are
experiencing bugs.

Simplifying to stop on first error.

Test Plan: Run manually

Reviewers: sdong igor

CC: leveldb@

Task ID: #6968635

Blame Rev:
2015-09-23 15:08:58 -07:00
Assaf Sela
4805fa0eae Remove ldb HexToString method's usage of sscanf
Summary:
Fix hex2String performance issues by removing sscanf dependency.
Also fixed some edge case handling (odd length, bad input).

Test Plan: Created a test file which called old and new implementation, and validated results are the same. I'll paste results in the phabricator diff.

Reviewers: igor, rven, anthony, IslamAbdelRahman, kradhakrishnan, yhchiang, sdong

Reviewed By: sdong

Subscribers: thatsafunnyname, leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D46785
2015-09-23 14:25:46 -07:00
Islam AbdelRahman
f03b5c987b Add experimental DB::AddFile() to plug sst files into empty DB
Summary:
This is an initial version of bulk load feature

This diff allow us to create sst files, and then bulk load them later, right now the restrictions for loading an sst file are
(1) Memtables are empty
(2) Added sst files have sequence number = 0, and existing values in database have sequence number = 0
(3) Added sst files values are not overlapping

Test Plan: unit testing

Reviewers: igor, ott, sdong

Reviewed By: sdong

Subscribers: leveldb, ott, dhruba

Differential Revision: https://reviews.facebook.net/D39081
2015-09-23 12:42:43 -07:00
Yueh-Hsuan Chiang
3fdb6e5234 Fixed old lint errors in db/filename.cc
Summary: Fixed old lint errors in db/filename.cc

Test Plan: make

Reviewers: igor, sdong, anthony, IslamAbdelRahman

Reviewed By: IslamAbdelRahman

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47445
2015-09-23 12:39:16 -07:00
Yueh-Hsuan Chiang
b349d22786 Fixed old lint errors in db/filename.h
Summary: Fixed old lint errors in db/filename.h

Test Plan: make

Reviewers: igor, sdong, anthony, IslamAbdelRahman

Reviewed By: IslamAbdelRahman

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47439
2015-09-23 12:22:44 -07:00
sdong
df34aea331 PlainTableReader to support non-mmap mode
Summary:
PlainTableReader now only allows mmap-mode. Add the support to non-mmap mode for more flexibility.
Refactor the codes to move all logic of reading data to PlainTableKeyDecoder, and consolidate the calls to Read() call and ReadVarint32() call. Implement the calls for both of mmap and non-mmap case seperately. For non-mmap mode, make copy of keys in several places when we need to move the buffer after reading the keys.

Test Plan: Add the mode of non-mmap case in plain_table_db_test. Run it in valgrind mode too.

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D47187
2015-09-23 11:41:07 -07:00
sdong
d746eaad5e RandomAccessFileReader should not inherit RandomAccessFile
Summary: RandomAccessFileReader unnecessarily inherited RandomAccessFile, which can introduce unnecessarily extra costs. Remove it.

Test Plan: Run all existing tests

Reviewers: yhchiang, anthony, igor, kradhakrishnan, rven, IslamAbdelRahman

Reviewed By: IslamAbdelRahman

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D47409
2015-09-23 11:00:41 -07:00
krad
03dd8f3ca7 Fixing punit job description
Summary: The job errors with log print step. Removing the step and re-enabling
the test case.

Test Plan: Manual test run

Reviewers:

CC: levledb@

Task ID: #6968635

Blame Rev:
2015-09-23 07:35:29 -07:00
sdong
d0c31641d2 Internal stats WAL file synced to match meaning of the stats of the same name
Summary: https://reviews.facebook.net/D23343 changed WAL sync bytes to extra fsync. This change does the same for internal stats.

Test Plan: Run all existing unit tests and verify results in db_bench.

Reviewers: anthony, rven, igor, MarkCallaghan, kradhakrishnan, yhchiang

Reviewed By: yhchiang

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D47349
2015-09-22 14:23:11 -07:00
Siying Dong
48b4497f75 Merge pull request #730 from yuslepukhin/fix_write_batch_win_const_expr
Fix Windows constexpr issue and '#ifdef' column_family_test in Release.
2015-09-22 11:08:10 -07:00
krad
a6c22e3e41 Disabling parallel test CI job
Summary: Parallel test CI is failing. Reducing noise by temporarily disabling
the job

Test Plan: Manual

Reviewers:

CC: leveldb@

Task ID: #6968635

Blame Rev:
2015-09-22 10:37:56 -07:00
Dmitri Smirnov
489a3e95d4 Re-work to support size_t max constant for 32/64-bit. 2015-09-22 10:34:21 -07:00
Yueh-Hsuan Chiang
ff57c6511d [RocksJava] Fix test failure of InfoLogLevelTest
Summary:
In patch https://reviews.facebook.net/D47067, we change the
log level of the initial database information to header level.
As a result, even when the InfoLogLevel is set to Fatal, the
LOG file of a newly opened rocksdb instance will not be empty.
However, the current InfoLogLevelTest expect it should be empty.

This patch fixes this issue by enabling InfoLogLevelTest to
ignore the Log header.

Test Plan: make jtest

Reviewers: fyrz, anthony, IslamAbdelRahman, sdong, adamretter

Reviewed By: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D47229
2015-09-21 18:26:21 -07:00
sdong
f1b9f804e9 Add a mode to always pick the oldest file to compact for each level
Summary:
Add options.compaction_pri, which specifies the policy about which file to compact first.
kCompactionPriByLargestSeq will compact oldest files first.
Verified the behavior in db_bench but did not write unit tests yet. Also need to make it settable through option string and dynamically changeable.

Test Plan: Will write unit tests

Reviewers: igor, rven, anthony, kradhakrishnan, IslamAbdelRahman, yhchiang, MarkCallaghan

Reviewed By: yhchiang

Subscribers: leveldb, dhruba

Differential Revision: https://reviews.facebook.net/D45951
2015-09-21 17:21:59 -07:00
Dmitri Smirnov
5e8f0a66db Use port::constant for std::muneric_limtis<>::max() 2015-09-21 16:56:47 -07:00
Dmitri Smirnov
2754ec9994 Fix Windows constexpr issue and '#ifdef' column_family_test in Release. 2015-09-21 16:21:01 -07:00
krad
dd2e1eeb35 Disabling log running jobs
Summary: Long running jobs are blocking the queue due to lack of adequate
machines in the CI

Test Plan: Manual testing

Reviewers:

CC: leveldb@

Task ID: #6968635

Blame Rev:
2015-09-21 12:58:33 -07:00
Igor Canadi
199744f4c4 Merge pull request #728 from jsteemann/fix-missing-include-header
add missing header required for std::function
2015-09-21 19:50:00 +02:00
Igor Canadi
0e65693f17 Merge pull request #727 from jsteemann/micro-optimization
Micro optimizations
2015-09-21 19:37:18 +02:00
krad
4d6eb52d1a Fix to CI jobs.
Summary: The approach to serialize long scripts as string is flawed. Instead
adding args to invoke the long scripts.

Test Plan: Manual test

Reviewers:

CC: leveldb@

Task ID: #6968635

Blame Rev:
2015-09-19 17:33:06 -07:00