Summary:
@dhruba, I'm not sure where we need to sync the directory. I implemented the function in Env() and added the dir sync just after we close the newly created file in the builder.
Should I also add FsyncDir() to new files that get created by a compaction?
Test Plan: Confirmed that FsyncDir is returning Status::OK()
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb, dhruba
Differential Revision: https://reviews.facebook.net/D14751
Summary: By removing some includes form options.h and reply on forward declaration, we can more easily reason the dependencies.
Test Plan: make all check
Reviewers: kailiu, haobo, igor, dhruba
Reviewed By: kailiu
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15411
Summary: On a shutdown, freeing memory takes a long time. If we're shutting down, we don't really care about memory leaks. I added a call to Cache that will avoid freeing all objects in cache.
Test Plan:
I created a script to test the speedup and demonstrate how to use the call: https://phabricator.fb.com/P3864368
Clean shutdown took 7.2 seconds, while fast and dirty one took 6.3 seconds. Unfortunately, the speedup is not that big, but should be bigger with bigger block_cache. I have set up the capacity to 80GB, but the script filled up only ~7GB.
Reviewers: dhruba, haobo, MarkCallaghan, xjin
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15069
Summary:
A lot of our code implicitly assumes number_levels to be static. ReduceNumberOfLevels() breaks that assumption. For example, after calling ReduceNumberOfLevels(), DBImpl::NumberLevels() will be different from VersionSet::NumberLevels(). This is dangerous. Thankfully, it's not in public headers and is only used from LDB cmd tool. LDB tool is only using it statically, i.e. it never calls it with running DB instance. With this diff, we make it explicitly static. This way, we can assume number_levels to be immutable and not break assumption that lot of our code is relying upon. LDB tool can still use the method.
Also, I removed the method from a separate file since it breaks filename completition. version_se<TAB> now completes to "version_set." instead of "version_set" (without the dot). I don't see a big reason that the function should be in a different file.
Test Plan: reduce_levels_test
Reviewers: dhruba, haobo, kailiu, sdong
Reviewed By: kailiu
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15303
Summary:
For some reason, D15099 caused a big performance regression: https://fburl.com/16059000
After digging a bit, I figured out that the reason was that std::atomic_uint_fast64_t was allocated in an array. When I switched from an array to vector, the QPS returned to the previous level. I'm not sure why this is happening, but this diff seems to fix the performance regression.
Test Plan: I ran the regression script, observed the performance going back to normal
Reviewers: tnovak, kailiu, haobo
Reviewed By: kailiu
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15375
Summary:
In my MacOS, the member variables are populated with random numbers after initialization.
This diff fixes it by fill these arrays with 0.
Test Plan: make && ./table_test
Reviewers: igor
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15315
Summary: I'm separating code-cleanup part of https://reviews.facebook.net/D14517. This will make D14517 easier to understand and this diff easier to review.
Test Plan: make check
Reviewers: haobo, kailiu, sdong, dhruba, tnovak
Reviewed By: tnovak
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15099
Summary:
shared_ptr is slower than unique_ptr (which literally comes with no performance cost compare with raw pointers).
In memtable and memtable rep, we use shared_ptr when we'd actually should use unique_ptr.
According to igor's previous work, we are likely to make quite some performance gain from this diff.
Test Plan: make check
Reviewers: dhruba, igor, sdong, haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15213
Summary:
With column families VersionSet will not have a constant number of levels (each CF can have different options), so we'll need to eliminate call to VersionSet::NumberLevels()
This diff decreases number of callsites, but we're not there yet. It associates number of levels with Version (each version is associated with single CF) instead of VersionSet.
I have also slightly changed how VersionSet keeps track of manifest size.
This diff also modifies constructor of Compaction such that it takes input_version and automatically Ref()s it. Before this was done outside of constructor.
In next diffs I will continue to decrease number of callsites of VersionSet::NumberLevels() and also references to current_
Test Plan: make check
Reviewers: haobo, dhruba, kailiu, sdong
Reviewed By: sdong
Differential Revision: https://reviews.facebook.net/D15171
Summary:
When doing CompactRange(), we should first flush the memtable and then calculate max_level_with_files. Also, we want to compact all the levels that have files, including level `max_level_with_files`.
This patch fixed the unit test.
Test Plan: Added a failing unit test and a fix, so it's not failing anymore.
Reviewers: dhruba, haobo, sdong
Reviewed By: haobo
CC: leveldb, xjin
Differential Revision: https://reviews.facebook.net/D14421
Summary:
Added an option (max_successive_merges) that can be used to specify the
maximum number of successive merge operations on a key in the memtable.
This can be used to improve performance of the "get" operation. If many
successive merge operations are performed on a key, the performance of "get"
operations on the key deteriorates, as the value has to be computed for each
"get" operation by applying all the successive merge operations.
FB Task ID: #3428853
Test Plan:
make all check
db_bench --benchmarks=readrandommergerandom
counter_stress_test
Reviewers: haobo, vamsi, dhruba, sdong
Reviewed By: haobo
CC: zshao
Differential Revision: https://reviews.facebook.net/D14991
Summary: Currently, even if statistics is not enabled, StopWatch only for the stats still gets the time of the day, which is wasteful. This patch adds a new option to StopWatch to disable this get in this case.
Test Plan: make all check
Reviewers: dhruba, haobo, igor
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14703
Summary: Use two vectors for different types of memory allocation.
Test Plan: run all unit tests.
Reviewers: haobo, sdong
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D15027
This seems out of place as it's the only time RocksDB prints to stdout in the
normal course of operations. Thread IDs can still be retrieved from the LOG
file: cut -d ' ' -f2 LOG | sort | uniq | egrep -x '[0-9a-f]+'
Summary:
In some places we have NotFound status created with empty message, but it doesn't avoid a malloc. With this patch, the malloc is avoided for that case.
The motivation of it is that I found in db_bench readrandom test when all keys are not existing, about 4% of the total running time is spent on malloc of Status, plus a similar amount of CPU spent on free of them, which is not necessary.
Test Plan: make all check
Reviewers: dhruba, haobo, igor
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14691
Summary:
A vector that leverages pre-allocated stack-based array to achieve better
performance for array with small amount of items.
Test Plan:
Added tests for both correctness and performance
Here is the performance benchmark between vector and autovector
Please note that in the test "Creation and Insertion Test", the test case were designed with the motivation described below:
* no element inserted: internal array of std::vector may not really get
initialize.
* one element inserted: internal array of std::vector must have
initialized.
* kSize elements inserted. This shows the most time we'll spend if we
keep everything in stack.
* 2 * kSize elements inserted. The internal vector of
autovector must have been initialized.
Note: kSize is the capacity of autovector
=====================================================
Creation and Insertion Test
=====================================================
created 100000 vectors:
each was inserted with 0 elements
total time elapsed: 128000 (ns)
created 100000 autovectors:
each was inserted with 0 elements
total time elapsed: 3641000 (ns)
created 100000 VectorWithReserveSizes:
each was inserted with 0 elements
total time elapsed: 9896000 (ns)
-----------------------------------
created 100000 vectors:
each was inserted with 1 elements
total time elapsed: 11089000 (ns)
created 100000 autovectors:
each was inserted with 1 elements
total time elapsed: 5008000 (ns)
created 100000 VectorWithReserveSizes:
each was inserted with 1 elements
total time elapsed: 24271000 (ns)
-----------------------------------
created 100000 vectors:
each was inserted with 4 elements
total time elapsed: 39369000 (ns)
created 100000 autovectors:
each was inserted with 4 elements
total time elapsed: 10121000 (ns)
created 100000 VectorWithReserveSizes:
each was inserted with 4 elements
total time elapsed: 28473000 (ns)
-----------------------------------
created 100000 vectors:
each was inserted with 8 elements
total time elapsed: 75013000 (ns)
created 100000 autovectors:
each was inserted with 8 elements
total time elapsed: 18237000 (ns)
created 100000 VectorWithReserveSizes:
each was inserted with 8 elements
total time elapsed: 42464000 (ns)
-----------------------------------
created 100000 vectors:
each was inserted with 16 elements
total time elapsed: 102319000 (ns)
created 100000 autovectors:
each was inserted with 16 elements
total time elapsed: 76724000 (ns)
created 100000 VectorWithReserveSizes:
each was inserted with 16 elements
total time elapsed: 68285000 (ns)
-----------------------------------
=====================================================
Sequence Access Test
=====================================================
performed 100000 sequence access against vector
size: 4
total time elapsed: 198000 (ns)
performed 100000 sequence access against autovector
size: 4
total time elapsed: 306000 (ns)
-----------------------------------
performed 100000 sequence access against vector
size: 8
total time elapsed: 565000 (ns)
performed 100000 sequence access against autovector
size: 8
total time elapsed: 512000 (ns)
-----------------------------------
performed 100000 sequence access against vector
size: 16
total time elapsed: 1076000 (ns)
performed 100000 sequence access against autovector
size: 16
total time elapsed: 1070000 (ns)
-----------------------------------
Reviewers: dhruba, haobo, sdong, chip
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14655
Summary:
I have ran a get benchmark where all the data is in the cache and observed that most of the time is spent on waiting for lock in LRUCache.
This is an effort to optimize LRUCache.
Test Plan:
The data was loaded with fillseq. Then, I ran a benchmark:
/db_bench --db=/tmp/rocksdb_stat_bench --num=1000000 --benchmarks=readrandom --statistics=1 --use_existing_db=1 --threads=16 --disable_seek_compaction=1 --cache_size=20000000000 --cache_numshardbits=8 --table_cache_numshardbits=8
I ran the benchmark three times. Here are the results:
AFTER THE PATCH: 798072, 803998, 811807
BEFORE THE PATCH: 782008, 815593, 763017
Reviewers: dhruba, haobo, kailiu
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14571
Summary: So fflush() takes a lock which is heavyweight. I added flush_pending_, but more importantly, I removed LogFlush() from foreground threads.
Test Plan: ./db_test
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14535
Summary:
In this diff I present you BackupableDB v1. You can easily use it to backup your DB and it will do incremental snapshots for you.
Let's first describe how you would use BackupableDB. It's inheriting StackableDB interface so you can easily construct it with your DB object -- it will add a method RollTheSnapshot() to the DB object. When you call RollTheSnapshot(), current snapshot of the DB will be stored in the backup dir. To restore, you can just call RestoreDBFromBackup() on a BackupableDB (which is a static method) and it will restore all files from the backup dir. In the next version, it will even support automatic backuping every X minutes.
There are multiple things you can configure:
1. backup_env and db_env can be different, which is awesome because then you can easily backup to HDFS or wherever you feel like.
2. sync - if true, it *guarantees* backup consistency on machine reboot
3. number of snapshots to keep - this will keep last N snapshots around if you want, for some reason, be able to restore from an earlier snapshot. All the backuping is done in incremental fashion - if we already have 00010.sst, we will not copy it again. *IMPORTANT* -- This is based on assumption that 00010.sst never changes - two files named 00010.sst from the same DB will always be exactly the same. Is this true? I always copy manifest, current and log files.
4. You can decide if you want to flush the memtables before you backup, or you're fine with backing up the log files -- either way, you get a complete and consistent view of the database at a time of backup.
5. More things you can find in BackupableDBOptions
Here is the directory structure I use:
backup_dir/CURRENT_SNAPSHOT - just 4 bytes holding the latest snapshot
0, 1, 2, ... - files containing serialized version of each snapshot - containing a list of files
files/*.sst - sst files shared between snapshots - if one snapshot references 00010.sst and another one needs to backup it from the DB, it will just reference the same file
files/ 0/, 1/, 2/, ... - snapshot directories containing private snapshot files - current, manifest and log files
All the files are ref counted and deleted immediatelly when they get out of scope.
Some other stuff in this diff:
1. Added GetEnv() method to the DB. Discussed with @haobo and we agreed that it seems right thing to do.
2. Fixed StackableDB interface. The way it was set up before, I was not able to implement BackupableDB.
Test Plan:
I have a unittest, but please don't look at this yet. I just hacked it up to help me with debugging. I will write a lot of good tests and update the diff.
Also, `make asan_check`
Reviewers: dhruba, haobo, emayanke
Reviewed By: dhruba
CC: leveldb, haobo
Differential Revision: https://reviews.facebook.net/D14295
Summary:
This will help me a lot! When we hit an assertion in unittest, we get the whole stack trace now.
Also, changed stack trace a bit, we now include actual demangled C++ class::function symbols!
Test Plan: Added ASSERT_TRUE(false) to a test, observed a stack trace
Reviewers: haobo, dhruba, kailiu
Reviewed By: kailiu
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14499
Summary:
This adds 2 options for compression to db_bench:
* universal_compression_size_percent
* compression_level - to set zlib compression level
It also logs compression_size_percent at startup in LOG
Task ID: #
Blame Rev:
Test Plan:
make check, run db_bench
Revert Plan:
Database Impact:
Memcache Impact:
Other Notes:
EImportant:
- begin *PUBLIC* platform impact section -
Bugzilla: #
- end platform impact -
Reviewers: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14439
Summary:
Let's get rid of TransformRep and it's children. We have confirmed that HashSkipListRep works better with multifeed, so there is no benefit to keeping this around.
This diff is mostly just deleting references to obsoleted functions. I also have a diff for fbcode that we'll need to push when we switch to new release.
I had to expose HashSkipListRepFactory in the client header files because db_impl.cc needs access to GetTransform() function for SanitizeOptions.
Test Plan: make check
Reviewers: dhruba, haobo, kailiu, sdong
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14397
Summary: The preprocessor does not follow normal rules of && evaluation, tries to evaluate __GLIBC_PREREQ(2, 12) even though the defined() check fails. This breaks the build if __GLIBC_PREREQ is absent.
Test Plan: Try adding #undef __GLIBC_PREREQ above the offending line, build no longer breaks
Reviewed By: igor
Blame Rev: 4c81383628
Summary: Makes it easier to monitor performance with top
Test Plan: ./manual_compaction_test with `top -H` running. Previously was two `manual_compacti`, now one shows `rocksdb:bg0`.
Reviewers: igor, dhruba
Reviewed By: igor
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14367
Summary: liveness of the statistics object is already ensured by the shared pointer in DB options. There's no reason to pass again shared pointer among internal functions. Raw pointer is sufficient and efficient.
Test Plan: make check
Reviewers: dhruba, MarkCallaghan, igor
Reviewed By: dhruba
CC: leveldb, reconnect.grayhat
Differential Revision: https://reviews.facebook.net/D14289
Summary:
Provide a framework to profile a query in detail to figure out latency bottleneck. Currently, in Get(), Put() and iterators, 2-3 simple timing is used. We can easily add more profile counters to the framework later.
Test Plan: Enable this profiling in seveal existing tests.
Reviewers: haobo, dhruba, kailiu, emayanke, vamsi, igor
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14001
Summary:
Previously we introduce a `flush_block_policy_factory` in Options, however, that options is strongly releated to Table based tables.
It will make more sense to move it to block based table's own factory class.
Test Plan: make check to pass existing tests
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14211
Summary:
The primary motivation of the changes is to make it easier to figure out the inside of the tables.
* rename "table stats" to "table properties" since now we have more than "integers" to store in the property block.
* Add filter block size to the basic table properties.
* Whenever a table is built, we'll log the table properties (the sample output is in Test Plan).
* Make an api to expose deleted keys.
Test Plan:
Passed all existing test. and the sample output of table stats:
==================================================================
Basic Properties
------------------------------------------------------------------
# data blocks: 1
# entries: 1
raw key size: 9
raw average key size: 9
raw value size: 9
raw average value size: 0
data block size: 25
index block size: 27
filter block size: 18
(estimated) table size: 70
filter policy: rocksdb.BuiltinBloomFilter
==================================================================
User collected properties: InternalKeyPropertiesCollector
------------------------------------------------------------------
kDeletedKeys: 1
==================================================================
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14187
Summary:
1. Moved the compiler back to 4.8.1 and uses Centos 5.2 binaries if OS is Centos 5.2.
2. Fixes this issue: https://github.com/facebook/rocksdb/issues/7
3. We use lot of c++11 features, so we can't pretend we can compile without them. Makes it a first class dependency.
4. Fix blob_store_test, which failes on Ubuntu with "too many files opened" error
5. Removed dependency on port/port_chromium.h, which does not even exist on our system
Test Plan: make clean; make check
Reviewers: dhruba, kailiu
Reviewed By: kailiu
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14145
Summary: This diff invoves some more complicated issues in the posix environment.
Test Plan: works under mac os. will need to verify dev box.
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14061
Summary:
mac and our dev server has totally differnt definition of uint64_t, therefore fixing the warning in mac has actually made code in linux uncompileable.
Test Plan:
make clean && make -j32
Summary: The work to make sure mac os compiles rocksdb is not completed yet. But at least we can start cleaning some warnings captured only by g++ from mac os..
Test Plan: ran make in mac os
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D14049
Summary:
@haobo's suggestions from https://reviews.facebook.net/D13827
Renaming some variables, deprecating purge_log_after_flush, changing for loop into auto for loop.
I have not implemented deleting objects outside of mutex yet because it would require a big code change - we would delete object in db_impl, which currently does not know anything about object because it's defined in version_edit.h (FileMetaData). We should do it at some point, though.
Test Plan: Ran deletefile_test
Reviewers: haobo
Reviewed By: haobo
CC: leveldb, haobo
Differential Revision: https://reviews.facebook.net/D14025
Summary: In our project, when writing to the database, we want to form the value as the concatenation of a small header and a larger payload. It's a shame to have to copy the payload just so we can give RocksDB API a linear view of the value. Since RocksDB makes a copy internally, it's easy to support gather writes.
Test Plan: write_batch_test, new test case
Reviewers: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D13947
Summary:
Here's one solution we discussed on speeding up FindObsoleteFiles. Keep a set of all files in DBImpl and update the set every time we create a file. I probably missed few other spots where we create a file.
It might speed things up a bit, but makes code uglier. I don't really like it.
Much better approach would be to abstract all file handling to a separate class. Think of it as layer between DBImpl and Env. Having a separate class deal with file namings and deletion would benefit both code cleanliness (especially with huge DBImpl) and speed things up. It will take a huge effort to do this, though.
Let's discuss offline today.
Test Plan: Ran ./db_stress, verified that files are getting deleted
Reviewers: dhruba, haobo, kailiu, emayanke
Reviewed By: dhruba
Differential Revision: https://reviews.facebook.net/D13827
