Summary:
1. prepare()
2. crash
3. recover
4. commit()
5. crash
6. data is lost
This is due to the transaction data still only residing in the WAL but because the logs were flushed on the first recovery the data is ignored on the second recovery. We must scan all logs found on recovery and only ignore redundant data at the time of replay. It is not possible to know which logs still contain relevant data at time of recovery. We cannot simply ignore a log because all of the non-2pc data it contains has already been written to L0.
The changes made to MemTableInserter are to ensure that prepared sections are still recovered even if all of the non-2pc data in that log has already been flushed to L0.
Test Plan: Provided test.
Reviewers: sdong
Subscribers: andrewkr, hermanlee4, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D57729
Summary:
Consider the following WAL with 4 batch entries prefixed with their sequence at time of memtable insert.
[1: BEGIN_PREPARE, PUT, PUT, PUT, PUT, END_PREPARE(a)]
[1: BEGIN_PREPARE, PUT, PUT, PUT, PUT, END_PREPARE(b)]
[4: COMMIT(a)]
[7: COMMIT(b)]
The first two batches do not consume any sequence numbers so are both prefixed with seq=1.
For 2pc commit, memtable insertion takes place before COMMIT batch is written to WAL.
We can see that sequence number consumption takes place between WAL entries giving us the seemingly sparse sequence prefix for WAL entries.
This is a valid WAL.
Because with 2PC markers one WriteBatch points to another batch containing its inserts a writebatch can consume more or less sequence numbers than the number of sequence consuming entries that it contains.
We can see that, given the entries in the WAL, 6 sequence ids were consumed. Yet on recovery the maximum sequence consumed would be 7 + 3 (the number of sequence numbers consumed by COMMIT(b))
So, now upon recovery we must track the actual consumption of sequence numbers.
In the provided scenario there will be no sequence gaps, but it is possible to produce a sequence gap. This should not be a problem though. correct?
Test Plan: provided test.
Reviewers: sdong
Subscribers: andrewkr, leveldb, dhruba, hermanlee4
Differential Revision: https://reviews.facebook.net/D57645
Summary: Adds three new WriteBatch data types: Prepare(xid), Commit(xid), Rollback(xid). Prepare(xid) should precede the (single) operation to which is applies. There can obviously be multiple Prepare(xid) markers. There should only be one Rollback(xid) or Commit(xid) marker yet not both. None of this logic is currently enforced and will most likely be implemented further up such as in the memtableinserter. All three markers are similar to PutLogData in that they are writebatch meta-data, ie stored but not counted. All three markers differ from PutLogData in that they will actually be written to disk. As for WriteBatchWithIndex, Prepare, Commit, Rollback are all implemented just as PutLogData and none are tested just as PutLogData.
Test Plan: single unit test in write_batch_test.
Reviewers: hermanlee4, sdong, anthony
Subscribers: leveldb, dhruba, vasilep, andrewkr
Differential Revision: https://reviews.facebook.net/D57867
Summary: Adds three new WriteBatch data types: Prepare(xid), Commit(xid), Rollback(xid). Prepare(xid) should precede the (single) operation to which is applies. There can obviously be multiple Prepare(xid) markers. There should only be one Rollback(xid) or Commit(xid) marker yet not both. None of this logic is currently enforced and will most likely be implemented further up such as in the memtableinserter. All three markers are similar to PutLogData in that they are writebatch meta-data, ie stored but not counted. All three markers differ from PutLogData in that they will actually be written to disk. As for WriteBatchWithIndex, Prepare, Commit, Rollback are all implemented just as PutLogData and none are tested just as PutLogData.
Test Plan: single unit test in write_batch_test.
Reviewers: hermanlee4, sdong, anthony
Subscribers: andrewkr, vasilep, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54093
Summary: Fix BackupableDBTest.NoDoubleCopy and BackupableDBTest.DifferentEnvs by mocking the db files in db_env instead of backup_env_
Test Plan: make check -j64
Reviewers: sdong, andrewkr
Reviewed By: andrewkr
Subscribers: andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D57273
Summary:
When db_env_ != backup_env_, InsertPathnameToSizeBytes() would
use the wrong Env during backup creation. This happened because this function
used backup_env_ instead of db_env_ to get WAL/data file sizes.
This diff adds an argument to InsertPathnameToSizeBytes() indicating which Env
to use.
Test Plan: ran @anirbanb's BackupTestTool
Reviewers: sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D57159
* Musl libc does not provide adaptive mutex. Added feature test for PTHREAD_MUTEX_ADAPTIVE_NP.
* Musl libc does not provide backtrace(3). Added a feature check for backtrace(3).
* Fixed compiler error.
* Musl libc does not implement backtrace(3). Added platform check for libexecinfo.
* Alpine does not appear to support gcc -pg option. By default (gcc has PIE option enabled) it fails with:
gcc: error: -pie and -pg|p|profile are incompatible when linking
When -fno-PIE and -nopie are used it fails with:
/usr/lib/gcc/x86_64-alpine-linux-musl/5.3.0/../../../../x86_64-alpine-linux-musl/bin/ld: cannot find gcrt1.o: No such file or directory
Added gcc -pg platform test and output PROFILING_FLAGS accordingly. Replaced pg var in Makefile with PROFILING_FLAGS.
* fix segfault when TEST_IOCTL_FRIENDLY_TMPDIR is undefined and default candidates are not suitable
* use ASSERT_DOUBLE_EQ instead of ASSERT_EQ
* When compiled with ROCKSDB_MALLOC_USABLE_SIZE UniversalCompactionFourPaths and UniversalCompactionSecondPathRatio tests fail due to premature memtable flushes on systems with 16-byte alignment. Arena runs out of block space before GenerateNewFile() completes.
Increased options.write_buffer_size.
Summary:
This interface is redundant and has been deprecated for a while.
It's also unused internally. Let's delete it.
I moved the comments to the corresponding functions in BackupEngine/
BackupEngineReadOnly. This caused the diff tool to not work cleanly.
Test Plan:
unit tests
$ ./backupable_db_test
Reviewers: yhchiang, sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D56331
Summary:
Several of backupable_db_test fails if running standalone, because of directory missing. Fix it by:
(1) garbage collector skips shared directory if it doesn't exit
(2) BackupableDBTest.Issue921Test to create the parent directory of the backup directory fist.
Test Plan: Run the tests individually and make sure they pass
Subscribers: leveldb, andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D56829
Summary:
- Need to use unsigned long long for 64-bit literals on windows
- Need size_t for backup meta-file length since clang doesn't let us assign size_t to int
Test Plan: backupable_db_test and options_test
Reviewers: IslamAbdelRahman, yhchiang, sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D56391
Summary:
My last diff introduced a warning when compiling under release mode
https://reviews.facebook.net/D55539
fix the warning
Test Plan:
DEBUG_LEVEL=0 make db_bench
make check
Reviewers: sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba
Differential Revision: https://reviews.facebook.net/D56295
Summary:
- Put key offset and key size in WriteBatchIndexEntry
- Use vector for comparators in WriteBatchEntryComparator
I use a slightly modified version of @yoshinorim code to benchmark
https://gist.github.com/IslamAbdelRahman/b120f4fba8d6ff7d58d2
For Put I create a transaction that put a 1000000 keys and measure the time spent without commit.
For GetForUpdate I read the keys that I added in the Put transaction.
Original time:
```
rm -rf /dev/shm/rocksdb-example/
./txn_bench put 1000000
1000000 OK Ops | took 3.679 seconds
./txn_bench get_for_update 1000000
1000000 OK Ops | took 3.940 seconds
```
New Time
```
rm -rf /dev/shm/rocksdb-example/
./txn_bench put 1000000
1000000 OK Ops | took 2.727 seconds
./txn_bench get_for_update 1000000
1000000 OK Ops | took 3.880 seconds
```
It looks like there is no significant improvement in GetForUpdate() but we can see ~30% improvement in Put()
Test Plan: unittests
Reviewers: yhchiang, anthony, sdong
Reviewed By: sdong
Subscribers: andrewkr, dhruba, yoshinorim
Differential Revision: https://reviews.facebook.net/D55539
Summary:
Rocksdb backup engine maintains metadata about backups in separate files. But,
there was no way to add extra application specific data to it. Adding support
for that.
In some use cases, applications decide to restore a backup based on some
metadata. This will help those cases to cheaply decide whether to restore or
not.
Test Plan:
Added a unit test. Existing ones are passing
Sample meta file for BinaryMetadata test-
```
1459454043
0
metadata 6162630A64656600676869
2
private/1/MANIFEST-000001 crc32 1184723444
private/1/CURRENT crc32 3505765120
```
Reviewers: sdong, ldemailly, andrewkr
Reviewed By: andrewkr
Subscribers: andrewkr, dhruba, ldemailly
Differential Revision: https://reviews.facebook.net/D56007
Summary:
This fixes a similar issue as D54711: "CURRENT" file can mutate between
GetLiveFiles() and copy to the tmp directory, in which case it would reference
the wrong manifest filename. To fix this, I forge the "CURRENT" file such that
it simply contains the filename for the manifest returned by GetLiveFiles().
- Changed CreateCheckpoint() to forge current file
- Added CreateFile() utility function
- Added test case that rolls manifest during checkpoint creation
Test Plan:
$ ./checkpoint_test
Reviewers: sdong, IslamAbdelRahman
Reviewed By: IslamAbdelRahman
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D55065
Summary: Refactored db_bench transaction stress tests so that they can be called from unit tests as well.
Test Plan: run new unit test as well as db_bench
Reviewers: yhchiang, IslamAbdelRahman, sdong
Reviewed By: IslamAbdelRahman
Subscribers: andrewkr, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D55203
Summary:
- Keep track of obsolete manifests in VersionSet
- Updated FindObsoleteFiles() to put obsolete manifests in the JobContext for later use by PurgeObsoleteFiles()
- Added test case that verifies a stale manifest is deleted by a non-full purge
Test Plan:
$ ./backupable_db_test --gtest_filter=BackupableDBTest.ChangeManifestDuringBackupCreation
Reviewers: IslamAbdelRahman, yoshinorim, sdong
Reviewed By: sdong
Subscribers: andrewkr, leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D55269
Summary: Previously, reusing a transaction (by passing it as an argument to BeginTransaction) would not clear the transaction's snapshot. This is not a clear, well-definited behavior.
Test Plan: improved test
Reviewers: sdong, IslamAbdelRahman, horuff, jkedgar
Reviewed By: jkedgar
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D55053
Summary:
Now that we get sizes efficiently, we no longer need the workaround to
embed file size in filename.
Test Plan:
$ ./backupable_db_test
Reviewers: sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D55035
Summary:
For VerifyBackup(), backup files can be spread across "shared/",
"shared_checksum/", and "private/" subdirectories, so we have to
bulk get all three.
For CreateNewBackup(), we make two separate bulk calls: one for the
data files and one for WAL files.
There is also a new helper function, ExtendPathnameToSizeBytes(),
that translates the file attributes vector to a map. I decided to leave
GetChildrenFileAttributes()'s (from D53781) return type as vector to
keep it consistent with GetChildren().
Depends on D53781.
Test Plan:
verified relevant unit tests
$ ./backupable_db_test
Reviewers: IslamAbdelRahman, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D53919
Summary: Add function to reinitialize a transaction object so that it can be reused. This is an optimization so users can potentially avoid reallocating transaction objects.
Test Plan: added tests
Reviewers: yhchiang, kradhakrishnan, IslamAbdelRahman, sdong
Reviewed By: sdong
Subscribers: jkedgar, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D53835
Summary:
Fixed two related race conditions in backup creation.
(1) CreateNewBackup() uses DB::DisableFileDeletions() to prevent table files
from being deleted while it is copying; however, the MANIFEST file could still
rotate during this time. The fix is to stop deleting the old manifest in the
rotation logic. It will be deleted safely later when PurgeObsoleteFiles() runs
(can only happen when file deletions are enabled).
(2) CreateNewBackup() did not account for the CURRENT file being mutable.
This is significant because the files returned by GetLiveFiles() contain a
particular manifest filename, but the manifest to which CURRENT refers can
change at any time. This causes problems when CURRENT changes between the call
to GetLiveFiles() and when it's copied to the backup directory. To workaround this, I
manually forge a CURRENT file referring to the manifest filename returned in
GetLiveFiles().
(2) also applies to the checkpointing code, so let me know if this approach is
good and I'll make the same change there.
Test Plan:
new test for roll manifest during backup creation.
running the test before this change:
$ ./backupable_db_test --gtest_filter=BackupableDBTest.ChangeManifestDuringBackupCreation
...
IO error: /tmp/rocksdbtest-9383/backupable_db/MANIFEST-000001: No such file or directory
running the test after this change:
$ ./backupable_db_test --gtest_filter=BackupableDBTest.ChangeManifestDuringBackupCreation
...
[ RUN ] BackupableDBTest.ChangeManifestDuringBackupCreation
[ OK ] BackupableDBTest.ChangeManifestDuringBackupCreation (2836 ms)
Reviewers: IslamAbdelRahman, anthony, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54711
Summary:
As titled. This fixes the tsan error caused by logger_ being used in
backup_engine_'s destructor. It does not fix the transient unit test failure,
which is caused by MANIFEST file changing while backup is happening.
Test Plan:
verified the tsan error no longer happens on either success or
failure.
$ COMPILE_WITH_TSAN=1 make -j32 backupable_db_test
$ while ./backupable_db_test --gtest_filter=BackupableDBTest.CorruptionsTest ; do : ; done
Reviewers: sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54669
Summary:
Relax the check condition of prefix_extractor in CheckOptionsCompatibility
by allowing changing value from non-nullptr to nullptr or nullptr to
non-nullptr.
Test Plan:
options_test
options_util_test
Reviewers: sdong, anthony, IslamAbdelRahman, kradhakrishnan, gunnarku
Reviewed By: gunnarku
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54477
Summary: Broke transaction locking in 4.4 in D52197. Will cherry-pick this change into 4.4 (which hasn't yet been fully released). Repro'd using db_bench.
Test Plan: unit tests and db_Bench
Reviewers: sdong, yhchiang, kradhakrishnan, ngbronson
Reviewed By: ngbronson
Subscribers: ngbronson, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54021
Summary: There is an issue in DBImpl::WriteImpl where if an empty writebatch comes in and sync=true then the logs will be marked as being synced yet the sync never actually happens because there is no data in the writebatch. This causes the next incoming batch to hang while waiting for the logs to complete syncing. This fix syncs logs even if the writebatch is empty.
Test Plan: DoubleEmptyBatch unit test in transaction_test.
Reviewers: yoshinorim, hermanlee4, sdong, ngbronson, anthony
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D54057
Summary:
One test in transaction_test.cc forgets to call SyncPoint::DisableProcessing().
As a result, a program might to access the SyncPoint singleton after it
already goes out of scope.
This patch fix this error by calling SyncPoint::DisableProcessing().
Test Plan: transaction_test
Reviewers: sdong, IslamAbdelRahman, kradhakrishnan, anthony
Reviewed By: anthony
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54033
Summary:
memory_test.cc has some tests that are not unstable but
hard to reproduce, and the cause is the test itself not
the code. Temporarily disable the tests until
we have a good fix.
Test Plan: memory_test
Reviewers: sdong, anthony, IslamAbdelRahman, rven, kradhakrishnan
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D54009
Summary: MyRocks wants to be able to un-lock a key that was just locked by GetForUpdate(). To do this safely, I am now keeping track of the number of reads(for update) and writes for each key in a transaction. UndoGetForUpdate() will only unlock a key if it hasn't been written and the read count reaches 0.
Test Plan: more unit tests
Reviewers: igor, rven, yhchiang, spetrunia, sdong
Reviewed By: spetrunia, sdong
Subscribers: spetrunia, dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D47043
Summary:
copy from task 8196669:
1) Optimistic transactions do not support batching writes from different threads.
2) Pessimistic transactions do not support batching writes if an expiration time is set.
In these 2 cases, we currently do not do any write batching in DBImpl::WriteImpl() because there is a WriteCallback that could decide at the last minute to abort the write. But we could support batching write operations with callbacks if we make sure to process the callbacks correctly.
To do this, we would first need to modify write_thread.cc to stop preventing writes with callbacks from being batched together. Then we would need to change DBImpl::WriteImpl() to call all WriteCallback's in a batch, only write the batches that succeed, and correctly set the state of each batch's WriteThread::Writer.
Test Plan: Added test WriteWithCallbackTest to write_callback_test.cc which creates multiple client threads and verifies that writes are batched and executed properly.
Reviewers: hermanlee4, anthony, ngbronson
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D52863
Summary:
Doing inline checking of transaction expiration instead of
using a callback.
Test Plan: To be added
Reviewers: anthony
Reviewed By: anthony
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D53673
Makefile adjust paths for solaris build
Makefile enable _GLIBCXX_USE_C99 so that std::to_string is available
db_compaction_test.cc Initialise a variable to avoid a compilation error
db_impl.cc Include <alloca.h>
db_test.cc Include <alloca.h>
Environment.java recognise solaris envrionment
options_bulder.cc Make log unambiguous
geodb_impl.cc Make log and floor unambiguous
Summary: fix memory leak in test code
Test Plan: ran test
Reviewers: rven, igor, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D52617
Summary:
See a bug report here: https://github.com/facebook/rocksdb/issues/921
The fix is to not check the shared/ directory if share_table_files is false. We could also check FileExists() before GetChildren(), but that will add extra latency when Env is Hdfs :(
Test Plan: added a unit test
Reviewers: rven, sdong, IslamAbdelRahman, yhchiang, anthony
Reviewed By: anthony
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D52593
Summary: Warning in release build.
Test Plan: Make release and make all
Subscribers: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D52305
Summary: Stopped using std::timed_mutex as it has known issues in older versiong of gcc. Ran into these problems when testing MongoRocks.
Test Plan: unit tests. Manual mongo testing on gcc 4.8.
Reviewers: igor, yhchiang, rven, IslamAbdelRahman, kradhakrishnan, sdong
Reviewed By: sdong
Subscribers: dhruba, leveldb
Differential Revision: https://reviews.facebook.net/D52197
