Summary:
Transaction Log Iterator did not move to the next file in the series if there was a write batch at the end of the currentFile.
The solution is if the last seq no. of the current file is < RequestedSeqNo. Assume the first seqNo. of the next file has to satisfy the request.
Also major refactoring around the code. Moved opening the logreader to a seperate function, got rid of goto.
Test Plan: added a unit test for it.
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb, emayanke
Differential Revision: https://reviews.facebook.net/D10029
Summary: The crash_test depends on db_stress to work with pre-existing dir
Test Plan: make db_stress; Run db_stress with 'destroy_db_initially=0'
Reviewers: vamsi, dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D10041
Summary: For sanity w.r.t. the way we split up the reopens equally among the ops/thread
Test Plan: make db_stress; db_stress --ops_per_thread=10 --reopens=10 => error
Reviewers: vamsi, dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D10023
Summary: Renames in the Makefile. It will be used like this in third-party.
Test Plan: make
Reviewers: dhruba, sheki, heyongqiang, haobo
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9633
Summary:
TableAndFile was a struct used earlier to delete the file as we did not have std::unique_ptr in the codebase.
With Chip introducing C++11 hotness like std::unique_ptr we can do away with the struct.
Test Plan: make all check
Reviewers: haobo, heyongqiang
Reviewed By: heyongqiang
CC: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D9975
Summary:
1. The stock LRUCache nukes itself whenever the working set (the total number of entries not released by client at a certain time) is bigger than the cache capacity.
See https://our.dev.facebook.com/intern/tasks/?t=2252281
2. There's a bug in shard calculation leading to segmentation fault when only one shard is needed.
Test Plan: make check
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb, zshao, sheki
Differential Revision: https://reviews.facebook.net/D9927
Summary:
When I run db_crashtest, I am seeing lot of warnings that say db_stress completed
before it was killed. To fix that I made ops per thread a very large value so that it keeps
running until it is killed.
I also set #reopens to 0. Since we are killing the process anyway, the 'simulated crash'
that happens during reopen may not add additional value.
I usually see 10-25K ops happening before the kill. So I increased max_key from 100 to
1000 so that we use more distinct keys.
Test Plan:
Ran a few times.
Revert Plan: OK
Task ID: #
Reviewers: emayanke
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9909
Summary:
During recovery, last_updated_manifest number was not set if there were no records in the Write-ahead log.
Now check for the recovered manifest also and set last_updated_manifest file to the max value.
Test Plan: unit test
Reviewers: heyongqiang
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9891
Summary:
1. SetBackgroundThreads was not thread safe
2. queue_size_ does not seem necessary
3. moved condition signal after shared state change. Even though the original
order is in practice ok (because the mutex is still held), it looks fishy
and non-intuitive.
Test Plan: make check
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb, zshao
Differential Revision: https://reviews.facebook.net/D9825
Summary: The script runs and kills the stress test periodically. Default values have been used in the script now. Should I make this a part of the Makefile or automated rocksdb build? The values can be easily changed in the script right now, but should I add some support for variable values or input to the script? I believe the script achieves its objective of unsafe crashes and reopening to expect sanity in the database.
Test Plan: python tools/db_crashtest.py
Reviewers: dhruba, vamsi, MarkCallaghan
Reviewed By: vamsi
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9369
Summary:
If a class owns an object:
- If the object can be null => use a unique_ptr. no delete
- If the object can not be null => don't even need new, let alone delete
- for runtime sized array => use vector, no delete.
Test Plan: make check
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb, zshao, sheki, emayanke, MarkCallaghan
Differential Revision: https://reviews.facebook.net/D9783
Summary:
RocksDB does a binary search to look at the files which might contain the requested sequence number at the call GetUpdatesSince.
There was a bug in the binary search => when the file pointed by the middle index of bsearch was empty/corrupt it needst to resize the vector and update indexes.
This now fixes that.
Test Plan: existing unit tests pass.
Reviewers: heyongqiang, dhruba
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9777
Summary:
If the vector returned by GetUpdatesSince is empty, it is still returned to the
user. This causes it throw an std::range error.
The probable file list is checked and it returns an IOError status instead of OK now.
Test Plan: added a unit test.
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9771
Summary:
Use non mmapd files for Write-Ahead log.
Earlier use of MMaped files. made the log iterator read ahead and miss records.
Now the reader and writer will point to the same physical location.
There is no perf regression :
./db_bench --benchmarks=fillseq --db=/dev/shm/mmap_test --num=$(million 20) --use_existing_db=0 --threads=2
with This diff :
fillseq : 10.756 micros/op 185281 ops/sec; 20.5 MB/s
without this dif :
fillseq : 11.085 micros/op 179676 ops/sec; 19.9 MB/s
Test Plan: unit test included
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9741
Summary:
Earlier Statistics object was a raw pointer. This meant the user had to clear up
the Statistics object after creating the database. In most use cases the database is created in a function and the statistics pointer is out of scope. Hence the statistics object would never be deleted.
Now Using a shared_ptr to manage this.
Want this in before the next release.
Test Plan: make all check.
Reviewers: dhruba, emayanke
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9735
Summary: rocksdb uses a single global lock to protect in memory metadata. We should minimize the mutex protected code section to increase the effective parallelism of the program. See https://our.intern.facebook.com/intern/tasks/?t=2218928
Test Plan:
make check
db_bench
Reviewers: dhruba, heyongqiang
CC: zshao, leveldb
Differential Revision: https://reviews.facebook.net/D9705
Summary:
The unit test fails as our solution does not work with MMap'd files.
Disable the failing unit test. Put it back with the next diff which should fix the problem.
Test Plan: db_test
Reviewers: heyongqiang
CC: dhruba
Differential Revision: https://reviews.facebook.net/D9645
Summary:
* Add a method to check if the log reader is at EOF.
* If we know a record has been flushed force the log_reader to believe it is not at EOF, using a new method UnMarkEof().
This does not work with MMpaed files.
Test Plan: added a unit test.
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9567
Summary: This caused compilation problems on some gcc platforms during the third-partyrelease
Test Plan: make
Reviewers: sheki
Reviewed By: sheki
Differential Revision: https://reviews.facebook.net/D9627
Summary: simple sed command to replace NULL in tools directory. Was missed by the previous codemod.
Test Plan: it compiles
Reviewers: emayanke
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9621
Summary:
The events that trigger compaction:
* opening the database
* Get -> only if seek compaction is not disabled and other checks are true
* MakeRoomForWrite -> when memtable is full
* BackgroundCall ->
If the background thread is about to do a compaction run, it schedules
a new background task to trigger a possible compaction. This will cause
additional background threads to find and process other compactions that
can run concurrently.
Test Plan: ran db_bench with overwrite and readonly alternatively.
Reviewers: sheki, MarkCallaghan
Reviewed By: sheki
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9579
Summary:
This patch allows an application to specify whether to use bufferedio,
reads-via-mmaps and writes-via-mmaps per database. Earlier, there
was a global static variable that was used to configure this functionality.
The default setting remains the same (and is backward compatible):
1. use bufferedio
2. do not use mmaps for reads
3. use mmap for writes
4. use readaheads for reads needed for compaction
I also added a parameter to db_bench to be able to explicitly specify
whether to do readaheads for compactions or not.
Test Plan: make check
Reviewers: sheki, heyongqiang, MarkCallaghan
Reviewed By: sheki
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9429
Summary: Getting rid of boost in our github codebase which caused problems on third-party
Test Plan: make ldb; python tools/ldb_test.py
Reviewers: sheki, dhruba
Reviewed By: sheki
Differential Revision: https://reviews.facebook.net/D9543
Summary: Was causing error(warning) in third-party saying unused result
Test Plan: make
Reviewers: sheki, dhruba
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D9447
Summary: Some comparisons left in log_test.cc and db_test.cc complained by make
Test Plan: make
Reviewers: dhruba, sheki
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D9537
Summary: Makefile had options to ignore sign-comparisons and unused-parameters, which should be there. Also fixed the specific errors in the code-base
Test Plan: make
Reviewers: chip, dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9531
Summary:
Add --benchmarks=levelstats option to report per-level stats (#files, #bytes)
Change readwhilewriting test to report response time for writes but exclude
them from the stats merged by all threads.
Prevent "NaN" in stats output by preventing division by 0.
Remove "o" file I committed by mistake.
Task ID: #
Blame Rev:
Test Plan:
make check
Revert Plan:
Database Impact:
Memcache Impact:
Other Notes:
EImportant:
- begin *PUBLIC* platform impact section -
Bugzilla: #
- end platform impact -
Reviewers: dhruba
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D9513
Summary:
Rocksdb can create 0 sized log files when it is opened and closed without any operations.
The GetUpdatesSince fails currently if there is a log file of size zero.
This diff fixes this. If there is a log file is 0, it is removed form the probable_file_list
Test Plan: unit test
Reviewers: dhruba, heyongqiang
Reviewed By: heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9507
Summary:
Instead of checking for number of files in L0. Check for number of files in the requested level.
Bug introduced in D4929 (diff trying to do too many things).
Test Plan: db_test.
Reviewers: dhruba, MarkCallaghan
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D9483
Summary: negation of the condition checked currently had to be checkd actually
Test Plan: make ldb; python ldb_test.py
Reviewers: sheki, dhruba
Reviewed By: sheki
Differential Revision: https://reviews.facebook.net/D9459
Summary: boost functions cause complications while deploying to third-party
Test Plan: make
Reviewers: sheki, dhruba
Reviewed By: sheki
Differential Revision: https://reviews.facebook.net/D9441
Summary:
Add --benchmarks=updaterandom for read-modify-write workloads. This is different
from --benchmarks=readrandomwriterandom in a few ways. First, an "operation" is the
combined time to do the read & write rather than treating them as two ops. Second,
the same key is used for the read & write.
Change RandomGenerator to support rows larger than 1M. That was using "assert"
to fail and assert is compiled-away when -DNDEBUG is used.
Add more options to db_bench
--duration - sets the number of seconds for tests to run. When not set the
operation count continues to be the limit. This is used by random operation
tests.
--use_snapshot - when set GetSnapshot() is called prior to each random read.
This is to measure the overhead from using snapshots.
--get_approx - when set GetApproximateSizes() is called prior to each random
read. This is to measure the overhead for a query optimizer.
Task ID: #
Blame Rev:
Test Plan:
run db_bench
Revert Plan:
Database Impact:
Memcache Impact:
Other Notes:
EImportant:
- begin *PUBLIC* platform impact section -
Bugzilla: #
- end platform impact -
Reviewers: dhruba
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D9267
Summary: Updated TOOL_CHAIN_LIB_BASE to use the third-party version for jemalloc-3.3.1 which contains a bug fix in quarantine.cc. This was detected while debugging valgrind issues with the rocksdb table_test
Test Plan: make table_test;valgrind --leak-check=full ./table_test
Reviewers: dhruba, sheki, vamsi
Reviewed By: sheki
Differential Revision: https://reviews.facebook.net/D9387
Summary:
Ftruncate does not throw an error on disk-full. This causes Sig-bus in
the case where the database tries to issue a Put call on a full-disk.
Use posix_fallocate for allocation instead of truncate.
Add a check to use MMaped files only on ext4, xfs and tempfs, as
posix_fallocate is very slow on ext3 and older.
Test Plan: make all check
Reviewers: dhruba, chip
Reviewed By: dhruba
CC: adsharma, leveldb
Differential Revision: https://reviews.facebook.net/D9291
Summary: ldb_test.py did a lot of assertFalse checks and displayed all the failed messages on the std output making it confusing to tell a successful from a failed run. Also many empty lines used to be needlessly printed. Also added some progression-"feel-good" lines in the tests
Test Plan: python ldb_test.py
Reviewers: dhruba, sheki, dilipj, chip
Reviewed By: dilipj
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9297
Summary: Fix for memory leaks in rocksdb tests. Also modified the variable NUM_FAILED_TESTS to print the actual number of failed tests.
Test Plan: make <test>; valgrind --leak-check=full ./<test>
Reviewers: sheki, dhruba
Reviewed By: sheki
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9333
Summary:
SizeBeingCompacted was called without any lock protection. This causes
crashes, especially when running db_bench with value_size=128K.
The fix is to compute SizeUnderCompaction while holding the mutex and
passing in these values into the call to Finalize.
(gdb) where
#4 leveldb::VersionSet::SizeBeingCompacted (this=this@entry=0x7f0b490931c0, level=level@entry=4) at db/version_set.cc:1827
#5 0x000000000043a3c8 in leveldb::VersionSet::Finalize (this=this@entry=0x7f0b490931c0, v=v@entry=0x7f0b3b86b480) at db/version_set.cc:1420
#6 0x00000000004418d1 in leveldb::VersionSet::LogAndApply (this=0x7f0b490931c0, edit=0x7f0b3dc8c200, mu=0x7f0b490835b0, new_descriptor_log=<optimized out>) at db/version_set.cc:1016
#7 0x00000000004222b2 in leveldb::DBImpl::InstallCompactionResults (this=this@entry=0x7f0b49083400, compact=compact@entry=0x7f0b2b8330f0) at db/db_impl.cc:1473
#8 0x0000000000426027 in leveldb::DBImpl::DoCompactionWork (this=this@entry=0x7f0b49083400, compact=compact@entry=0x7f0b2b8330f0) at db/db_impl.cc:1757
#9 0x0000000000426690 in leveldb::DBImpl::BackgroundCompaction (this=this@entry=0x7f0b49083400, madeProgress=madeProgress@entry=0x7f0b41bf2d1e, deletion_state=...) at db/db_impl.cc:1268
#10 0x0000000000428f42 in leveldb::DBImpl::BackgroundCall (this=0x7f0b49083400) at db/db_impl.cc:1170
#11 0x000000000045348e in BGThread (this=0x7f0b49023100) at util/env_posix.cc:941
#12 leveldb::(anonymous namespace)::PosixEnv::BGThreadWrapper (arg=0x7f0b49023100) at util/env_posix.cc:874
#13 0x00007f0b4a7cf10d in start_thread (arg=0x7f0b41bf3700) at pthread_create.c:301
#14 0x00007f0b49b4b11d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:115
Test Plan:
make check
I am running db_bench with a value size of 128K to see if the segfault is fixed.
Reviewers: MarkCallaghan, sheki, emayanke
Reviewed By: sheki
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9279
Summary:
This is a regression caused by
772f75b3fb
If you do "strings libleveldb.a | grep leveldb_build_git_datetime" it will
show you the time when the binary was built.
Test Plan: make check
Reviewers: emayanke
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9273
Summary:
Also added some comments and fixed some bugs in
stats reporting. Now the stats seem to match what is expected.
Test Plan:
[nponnekanti@dev902 /data/users/nponnekanti/rocksdb] ./db_stress --test_batches_snapshots=1 --ops_per_thread=1000 --threads=1 --max_key=320
LevelDB version : 1.5
Number of threads : 1
Ops per thread : 1000
Read percentage : 10
Delete percentage : 30
Max key : 320
Ratio #ops/#keys : 3
Num times DB reopens: 10
Batches/snapshots : 1
Num keys per lock : 4
Compression : snappy
------------------------------------------------
No lock creation because test_batches_snapshots set
2013/03/04-15:58:56 Starting database operations
2013/03/04-15:58:56 Reopening database for the 1th time
2013/03/04-15:58:56 Reopening database for the 2th time
2013/03/04-15:58:56 Reopening database for the 3th time
2013/03/04-15:58:56 Reopening database for the 4th time
Created bg thread 0x7f4542bff700
2013/03/04-15:58:56 Reopening database for the 5th time
2013/03/04-15:58:56 Reopening database for the 6th time
2013/03/04-15:58:56 Reopening database for the 7th time
2013/03/04-15:58:57 Reopening database for the 8th time
2013/03/04-15:58:57 Reopening database for the 9th time
2013/03/04-15:58:57 Reopening database for the 10th time
2013/03/04-15:58:57 Reopening database for the 11th time
2013/03/04-15:58:57 Limited verification already done during gets
Stress Test : 1811.551 micros/op 552 ops/sec
: Wrote 0.10 MB (0.05 MB/sec) (598% of 1011 ops)
: Wrote 6050 times
: Deleted 3050 times
: 500/900 gets found the key
: Got errors 0 times
[nponnekanti@dev902 /data/users/nponnekanti/rocksdb] ./db_stress --ops_per_thread=1000 --threads=1 --max_key=320
LevelDB version : 1.5
Number of threads : 1
Ops per thread : 1000
Read percentage : 10
Delete percentage : 30
Max key : 320
Ratio #ops/#keys : 3
Num times DB reopens: 10
Batches/snapshots : 0
Num keys per lock : 4
Compression : snappy
------------------------------------------------
Creating 80 locks
2013/03/04-15:58:17 Starting database operations
2013/03/04-15:58:17 Reopening database for the 1th time
2013/03/04-15:58:17 Reopening database for the 2th time
2013/03/04-15:58:17 Reopening database for the 3th time
2013/03/04-15:58:17 Reopening database for the 4th time
Created bg thread 0x7fc0f5bff700
2013/03/04-15:58:17 Reopening database for the 5th time
2013/03/04-15:58:17 Reopening database for the 6th time
2013/03/04-15:58:18 Reopening database for the 7th time
2013/03/04-15:58:18 Reopening database for the 8th time
2013/03/04-15:58:18 Reopening database for the 9th time
2013/03/04-15:58:18 Reopening database for the 10th time
2013/03/04-15:58:18 Reopening database for the 11th time
2013/03/04-15:58:18 Starting verification
Stress Test : 1836.258 micros/op 544 ops/sec
: Wrote 0.01 MB (0.01 MB/sec) (59% of 1011 ops)
: Wrote 605 times
: Deleted 305 times
: 50/90 gets found the key
: Got errors 0 times
2013/03/04-15:58:18 Verification successful
Revert Plan: OK
Task ID: #
Reviewers: emayanke, dhruba
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9081
Summary: This option is needed for compilation and the open-sourced rocksdb version wiull need to get it from Makefile
Test Plan: make clean;make
Reviewers: MarkCallaghan, dhruba, sheki, chip
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9243
Summary:
the valgrind version being used is in facebook specific path and should be moved to the fbcode.gcc471.sh file instead of the makefile.
The execution takes the environment's default valgrind version if the fbcode.gcc471.sh's valgrind_version is not available.
Test Plan: make valgrind_check
Reviewers: dhruba, sheki, akushner
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D9213
Summary:
1. Create only 2 levels so that manual compactions are fast.
2. Set target file size to a large value
Test Plan: make clean check
Reviewers: kailiu, zshao
Reviewed By: zshao
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9231
Summary: In light of the new option introduced by commit 806e264350 where the database has an option to compact before flushing to disk, we want the stress test to test both sides of the option. Have made it to 'deterministically' and configurably change that option for reopens.
Test Plan: make db_stress; ./db_stress with some differnet options
Reviewers: dhruba, vamsi
Reviewed By: dhruba
CC: leveldb, sheki
Differential Revision: https://reviews.facebook.net/D9165
Summary:
If there is an error while writing an edit to the manifest file, the manifest
file is closed and reopened to check if the edit made it in. However, if the
re-opening of the manifest is unsuccessful and options.paranoid_checks is set
t true, then the db refuses to accept new puts, effectively putting the db
in readonly mode.
In a future diff, I would like to make the default value of paranoid_check
to true.
Test Plan: make check
Reviewers: sheki
Reviewed By: sheki
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9201
Summary:
valgrind 3.7.0 used currently has a bug that needs LD_PRELOAD being set as a workaround. This caused problems when run on jenkins. 3.8.1 has fixed this issue and we should use it from third party
Also, have done away with log files. The whole output will be there on the terminal and the failed tests will be listed at the end. This is done because jenkins only lets us download the different files and not view them in the browser which is undesirable.
Test Plan: make valgrind_check
Reviewers: akushner, dhruba, vamsi, sheki, heyongqiang
Reviewed By: sheki
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9171
Summary:
Store the last flushed, seq no. in db_impl. Check against it in
transaction Log iterator. Do not attempt to read ahead if we do not know
if the data is flushed completely.
Does not work if flush is disabled. Any ideas on fixing that?
* Minor change, iter->Next is called the first time automatically for
* the first time.
Test Plan:
existing test pass.
More ideas on testing this?
Planning to run some stress test.
Reviewers: dhruba, heyongqiang
CC: leveldb
Differential Revision: https://reviews.facebook.net/D9087
Abhishek Kona