Summary:
There is a use-case where we want to insert data into rocksdb as
fast as possible. Vector rep is used for this purpose.
The background flush thread needs to flush the vectorrep to
storage. It acquires the dblock then sorts the vector, releases
the dblock and then writes the sorted vector to storage. This is
suboptimal because the lock is held during the sort, which
prevents new writes for occuring.
This patch moves the sorting of the vector rep to outside the
db mutex. Performance is now as fastas the underlying storage
system. If you are doing buffered writes to rocksdb files, then
you can observe throughput upwards of 200 MB/sec writes.
This is an early draft and not yet ready to be reviewed.
Test Plan:
make check
Task ID: #
Blame Rev:
Reviewers: haobo
Reviewed By: haobo
CC: leveldb, haobo
Differential Revision: https://reviews.facebook.net/D12987
Summary: as title
Test Plan: make db_test; ./db_test
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D13005
Summary: as title
Test Plan: make db_test; ./db_test
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12981
Summary: Currently, when total number of files reaches level0_file_num_compaction_trigger, universal compaction will schedule a compaction job, but the job will not honor the compaction until the total number of files is level0_file_num_compaction_trigger+1. Fixed the condition for consistent behavior (start compaction on reaching level0_file_num_compaction_trigger).
Test Plan: make check; db_stress
Reviewers: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12945
Summary:
Added a new field called max_size_amplification_ratio in the
CompactionOptionsUniversal structure. This determines the maximum
percentage overhead of space amplification.
The size amplification is defined to be the ratio between the size of
the oldest file to the sum of the sizes of all other files. If the
size amplification exceeds the specified value, then min_merge_width
and max_merge_width are ignored and a full compaction of all files is done.
A value of 10 means that the size a database that stores 100 bytes
of user data could occupy 110 bytes of physical storage.
Test Plan: Unit test DBTest.UniversalCompactionSpaceAmplification added.
Reviewers: haobo, emayanke, xjin
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12825
Summary: As title. The DB log file life cycle is tied up with the memtable it backs. Once the memtable is flushed to sst and committed, we should be able to delete the log file, without holding the mutex. This is part of the bigger change to avoid FindObsoleteFiles at runtime. It deals with log files. sst files will be dealt with later.
Test Plan: make check; db_bench
Reviewers: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D11709
Summary: The pupose of this diff is to expose per user-call level precise timing of block read, so that we can answer questions like: a Get() costs me 100ms, is that somehow related to loading blocks from file system, or sth else? We will answer that with EXACTLY how many blocks have been read, how much time was spent on transfering the bytes from os, how much time was spent on checksum verification and how much time was spent on block decompression, just for that one Get. A nano second stopwatch was introduced to track time with higher precision. The cost/precision of the stopwatch is also measured in unit-test. On my dev box, retrieving one time instance costs about 30ns, on average. The deviation of timing results is good enough to track 100ns-1us level events. And the overhead could be safely ignored for 100us level events (10000 instances/s), for example, a viewstate thrift call.
Test Plan: perf_context_test, also testing with viewstate shadow traffic.
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb, xjin
Differential Revision: https://reviews.facebook.net/D12351
Summary:
There is an config option called Options.min_write_buffer_number_to_merge
that specifies the minimum number of write buffers to merge in memory
before flushing to a file in L0. But in the the case when the db is
being closed, we should not be using this config, instead we should
flush whatever write buffers were available at that time.
Test Plan: Unit test attached.
Reviewers: haobo, emayanke
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12717
Summary:
An iterator invokes reseek if the number of sequential skips over the
same userkey exceeds a configured number. This makes iter->Next()
faster (bacause of fewer key compares) if a large number of
adjacent internal keys in a table (sst or memtable) have the
same userkey.
Test Plan: Unit test DBTest.IterReseek.
Reviewers: emayanke, haobo, xjin
Reviewed By: xjin
CC: leveldb, xjin
Differential Revision: https://reviews.facebook.net/D11865
Summary: So that replication can just download from wherever LogFile.Pathname is pointing them.
Test Plan: make all check;./db_repl_stress
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12609
Summary:
Add new command "change_compaction_style" to ldb tool. For
universal->level, it shows "nothing to do". For level->universal, it
compacts all files into a single one and moves the file to level 0.
Also add check for number of files at level 1+ when opening db with
universal compaction style.
Test Plan:
'make all check'. New unit test for internal convertion function. Also manully test various
cmd like:
./ldb change_compaction_style --old_compaction_style=0
--new_compaction_style=1 --db=/tmp/leveldbtest-3088/db_test
Reviewers: haobo, dhruba
Reviewed By: haobo
CC: vamsi, emayanke
Differential Revision: https://reviews.facebook.net/D12603
Summary:
The way counters/statistics are implemented in rocksdb demands that enum Tickers and TickerNameMap follow the same order, otherwise statistics exposed from fbcode/rocks get out-of-sync. 2 counters for prefix had violated this order and when I built counters for fbcode/mcrocksdb, statistics for sequence number were appearing out-of-sync.
The other change is to record sequence-number using setTickerCount only and not recordTick. This is because of difference in statistics as understood by rocks/utils which uses ServiceData::statistics function and rocksdb statistics. In rocksdb there is just 1 counter for a countername. But in ServiceData there are 4 independent buckets for every countername-Count, Sum, Average and Rate. SetTickerCount and RecordTick update the same variable in rocksdb but different buckets in ServiceData. Therefore, I had to choose one consistent function from RecordTick or SetTickerCount for sequence number in rocksdb. I chose SetTickerCount because the statistics object in options passed during rocksdb-open is user-dependent and SetTickerCount makes sense there.
There will be a corresponding diff to mcorcksdb in fbcode shortly.
Test Plan: make all check; check ticker value using fprintfs
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12669
Summary: db->DeleteFile calls ParseFileName to check name that was returned for sst file. Now, sst filename is returned using TableFileName which uses MakeFileName. This puts a / at the front of the name and ParseFileName doesn't like that. Changed ParseFileName to tolerate /s at the beginning. The test delet_file_test used to pass earlier because this behaviour of MakeFileName had been changed a while back to not return a / during which delete_file_test was checked in. But MakeFileName had to be reverted to add / at the front because GetLiveFiles used at many places outside rocksdb used the previous behaviour of MakeFileName.
Test Plan: make;./delete_filetest;make all check
Reviewers: dhruba, haobo, vamsi
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12663
Summary: // won't hurt but a missing / hurts sometimes
Test Plan: make all check; ./db_repl_stress
Reviewers: vamsi
Reviewed By: vamsi
CC: dhruba
Differential Revision: https://reviews.facebook.net/D12621
Summary:
The DeleteFile API was removing files inside the db-lock. This
is now changed to remove files outside the db-lock.
The GetLiveFilesMetadata() returns the smallest and largest
seqnuence number of each file as well.
Test Plan: deletefile_test
Reviewers: emayanke, haobo
Reviewed By: haobo
CC: leveldb
Maniphest Tasks: T63
Differential Revision: https://reviews.facebook.net/D12567
Summary: Let TransformRepFactory own the passed in transform. Also make it better encapsulated.
Test Plan: make valgrind_check;
Reviewers: dhruba, emayanke
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12591
Summary:
If ReadOptions.non_blocking_io is set to true, then KeyMayExists
and Iterators will return data that is cached in RAM.
If the Iterator needs to do IO from storage to serve the data,
then the Iterator.status() will return Status::IsRetry().
Test Plan:
Enhanced unit test DBTest.KeyMayExist to detect if there were are IOs
issues from storage. Added DBTest.NonBlockingIteration to verify
nonblocking Iterations.
Reviewers: emayanke, haobo
Reviewed By: haobo
CC: leveldb
Maniphest Tasks: T63
Differential Revision: https://reviews.facebook.net/D12531
Summary:
As title. This is possible as tickers are atomic now.
db_bench on high qps in-memory muti-thread random get workload, showed ~5% throughput improvement.
Test Plan: make check; db_bench; db_stress
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12555
Summary: In KeyMayExist.db_test we do a Flush which causes sst file to be written and added as open file in TableCache, but block cache for the file is not populated. So value_found should have been false where it was true and KeyMayExist.db_test should not have passed earlier. But it passed because BlockReader in table/table.cc takes 2 default arguments at the end called for_compaction and no_io. Although I passed no_io=true from InternalGet to BlockReader, but it understood for_compaction=true and defaulted no_io to false. This is a bug and although will be removed by Dhruba's new patch to incorporate no_io in readoptions, I'm submitting this patch to fix this bug independently of that patch.
Test Plan: make all check
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12537
Summary: initialiszer list is fasteri/preferable because it can straightaway call the constructor for this object, otherwise it will be created first and then again initialized. Although gain may not be much in this case because files_ is just a pointer and not a complex object, this is recommended practice.
Test Plan: make all check
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12519
Summary: Fix code so that the filter_block layer only assumes keys are internal when prefix_extractor is set.
Test Plan: ./filter_block_test
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12501
Summary: Replace include/leveldb with include/rocksdb.
Test Plan:
make clean; make check
make clean; make release
Differential Revision: https://reviews.facebook.net/D12489
Summary:
This patch adds three new MemTableRep's: UnsortedRep, PrefixHashRep, and VectorRep.
UnsortedRep stores keys in an std::unordered_map of std::sets. When an iterator is requested, it dumps the keys into an std::set and iterates over that.
VectorRep stores keys in an std::vector. When an iterator is requested, it creates a copy of the vector and sorts it using std::sort. The iterator accesses that new vector.
PrefixHashRep stores keys in an unordered_map mapping prefixes to ordered sets.
I also added one API change. I added a function MemTableRep::MarkImmutable. This function is called when the rep is added to the immutable list. It doesn't do anything yet, but it seems like that could be useful. In particular, for the vectorrep, it means we could elide the extra copy and just sort in place. The only reason I haven't done that yet is because the use of the ArenaAllocator complicates things (I can elaborate on this if needed).
Test Plan:
make -j32 check
./db_stress --memtablerep=vector
./db_stress --memtablerep=unsorted
./db_stress --memtablerep=prefixhash --prefix_size=10
Reviewers: dhruba, haobo, emayanke
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12117
Summary: If use_prefix_filters is set and read_range>1, then the random seeks will set a the prefix filter to be the prefix of the key which was randomly selected as the target. Still need to add statistics (perhaps in a separate diff).
Test Plan: ./db_bench --benchmarks=fillseq,prefixscanrandom --num=10000000 --statistics=1 --use_prefix_blooms=1 --use_prefix_api=1 --bloom_bits=10
Reviewers: dhruba
Reviewed By: dhruba
CC: leveldb, haobo
Differential Revision: https://reviews.facebook.net/D12273
Summary: An api to query the level, key ranges, size etc for each SST file and an api to delete a specific file from the db and all associated state in the bookkeeping datastructures.
Notes: Editing the manifest version does not release the obsolete files right away. However deleting the file directly will mess up the iterator. We may need a more aggressive/timely file deletion api.
I have used std::unique_ptr - will switch to boost:: since this is external. thoughts?
Unit test is fragile right now as it expects the compaction at certain levels.
Test Plan: unittest
Reviewers: dhruba, vamsi, emayanke
CC: zshao, leveldb, haobo
Task ID: #
Blame Rev:
Summary:
Sometimes you don't need to iterate through the whole WriteBatch. This diff makes the Handler member functions return a bool that indicates whether to abort or not. If they return true, the iteration stops.
One thing I just thought of is that this will break backwards-compability. Maybe it would be better to add a virtual member function WriteBatch::Handler::ShouldAbort() that returns false by default. Comments requested.
I still have to add a new unit test for the abort code, but let's finalize the API first.
Test Plan: make -j32 check
Reviewers: dhruba, haobo, vamsi, emayanke
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12339
Summary: Was going through the iterator related code, did some cleanup along the way. Basically replaced array with vector and adopted range based loop where applicable.
Test Plan: make check; make valgrind_check
Reviewers: dhruba, emayanke
Reviewed By: emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12435
Test Plan:
- make all check;
- make release;
- make stringappend_test; ./stringappend_test
Reviewers: haobo, emayanke
Reviewed By: haobo
CC: leveldb, kailiu
Differential Revision: https://reviews.facebook.net/D12381
Summary: Also expanded class LogFile to have startSequene and FileSize and exposed it publicly
Test Plan: make all check
Reviewers: dhruba, haobo
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12087
Summary:
-Added null checks and revisions to DBIter::MergeValuesNewToOld()
-Added DBIter test to stringappend_test
-Major fix with Merge and TTL
More plans for fixes later.
Test Plan:
-make clean; make stringappend_test -j 32; ./stringappend_test
-make all check;
Reviewers: haobo, emayanke, vamsi, dhruba
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12315
Summary: statistic for sequence number is needed by wormhole. setTickerCount is demanded for this statistic. I can't simply recordTick(max_sequence) when db recovers because the statistic iobject is owned by client and may/may not be reset during reopen. Eg. statistic is reset in mcrocksdb whereas it is not in db_stress. Therefore it is best to go with setTickerCount
Test Plan: ./db_stress ... --statistics=1 and observed expected sequence number
Reviewers: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12327
Summary:
In release, "found variable assigned but not used anywhere". Changed it to work with
assert. Someone accept this :).
Test Plan: make release -j 32
Reviewers: haobo, dhruba, emayanke
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12309
Summary:
Updated db_bench and utilities/merge_operators.h to allow for dynamic benchmarking
of merge operators in db_bench. Added a new test (--benchmarks=mergerandom), which performs
a bunch of random Merge() operations over random keys. Also added a "--merge_operator=" flag
so that the tester can easily benchmark different merge operators. Currently supports
the PutOperator and UInt64Add operator. Support for stringappend or list append may come later.
Test Plan:
1. make db_bench
2. Test the PutOperator (simulating Put) as follows:
./db_bench --benchmarks=fillrandom,readrandom,updaterandom,readrandom,mergerandom,readrandom --merge_operator=put
--threads=2
3. Test the UInt64AddOperator (simulating numeric addition) similarly:
./db_bench --value_size=8 --benchmarks=fillrandom,readrandom,updaterandom,readrandom,mergerandom,readrandom
--merge_operator=uint64add --threads=2
Reviewers: haobo, dhruba, zshao, MarkCallaghan
Reviewed By: haobo
CC: leveldb
Differential Revision: https://reviews.facebook.net/D11535
Summary:
Minor fix to current codes, including: coding style, output format,
comments. No major logic change. There are only 2 real changes, please see my inline comments.
Test Plan: make all check
Reviewers: haobo, dhruba, emayanke
Differential Revision: https://reviews.facebook.net/D12297
Summary:
This patch adds the ability for the user to add sequences of arbitrary data (blobs) to write batches. These blobs are saved to the log along with everything else in the write batch. You can add multiple blobs per WriteBatch and the ordering of blobs, puts, merges, and deletes are preserved.
Blobs are not saves to SST files. RocksDB ignores blobs in every way except for writing them to the log.
Before committing this patch, I need to add some test code. But I'm submitting it now so people can comment on the API.
Test Plan: make -j32 check
Reviewers: dhruba, haobo, vamsi
Reviewed By: dhruba
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12195
Summary:
As title. No locking/atomic is needed due to thread local. There is also no need to modify the existing client interface, in order to expose related counters.
perf_context_test shows a simple example of retrieving the number of user key comparison done for each put and get call. More counters could be added later.
Sample output
./perf_context_test 1000000
==== Test PerfContextTest.KeyComparisonCount
Inserting 1000000 key/value pairs
...
total user key comparison get: 43446523
total user key comparison put: 8017877
max user key comparison get: 88939
avg user key comparison get:43
Basically, the current skiplist does well on average, but could perform poorly in extreme cases.
Test Plan: run perf_context_test <total number of entries to put/get>
Reviewers: dhruba
Differential Revision: https://reviews.facebook.net/D12225
Summary:
With Merge returning bool, it can keep failing silently(eg. While faling to fetch timestamp in TTL). We need to detect this through a rocksdb counter which can get bumped whenever Merge returns false. This will also be super-useful for the mcrocksdb-counter service where Merge may fail.
Added a counter NUMBER_MERGE_FAILURES and appropriately updated db/merge_helper.cc
I felt that it would be better to directly add counter-bumping in Merge as a default function of MergeOperator class but user should not be aware of this, so this approach seems better to me.
Test Plan: make all check
Reviewers: dnicholas, haobo, dhruba, vamsi
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12129
Summary: Similar to v2 (db and table code understands prefixes), but use ReadOptions as in v3. Also, make the CreateFilter code faster and cleaner.
Test Plan: make db_test; export LEVELDB_TESTS=PrefixScan; ./db_test
Reviewers: dhruba
Reviewed By: dhruba
CC: haobo, emayanke
Differential Revision: https://reviews.facebook.net/D12027
Summary:
If we have same compaction filter for each compaction,
application cannot know about the different compaction processes.
Later on, we can put in more details in compaction filter for the
application to consume and use it according to its needs. For e.g. In
the universal compaction, we have a compaction process involving all the
files while others don't involve all the files. Applications may want to
collect some stats only when during full compaction.
Test Plan: run existing unit tests
Reviewers: haobo, dhruba
Reviewed By: dhruba
CC: xinyaohu, leveldb
Differential Revision: https://reviews.facebook.net/D12057
Summary:
The pre-existing code was purging a DeleteMarker if thay key did not
exist in deeper levels. But in the Universal Compaction Style, all
files are in Level0. For compaction runs that did not include the
earliest file, we were erroneously purging the DeleteMarkers.
The fix is to purge DeleteMarkers only if the compaction includes
the earlist file.
Test Plan: DBTest.Randomized triggers this code path.
Differential Revision: https://reviews.facebook.net/D12081
Summary:
Continue fixing existing unit tests for universal compaction. I have
tried to apply universal compaction to all unit tests those haven't
called ChangeOptions(). I left a few which are either apparently not
applicable to universal compaction (because they check files/keys/values
at level 1 or above levels), or apparently not related to compaction
(e.g., open a file, open a db).
I also add a new unit test for universal compaction.
Good news is I didn't see any bugs during this round.
Test Plan: Ran "make all check" yesterday. Has rebased and is rerunning
Reviewers: haobo, dhruba
Differential Revision: https://reviews.facebook.net/D12135
Summary: Currently, VersionEdit::DebugString always display internal keys in the original ascii format. This could cause manifest dump to be truncated if internal keys contain special charactors (like null). Also added an option --input_key_hex for ldb idump to indicate that the passed in user keys are in hex.
Test Plan: run ldb manifest_dump
Reviewers: dhruba, emayanke
CC: leveldb
Differential Revision: https://reviews.facebook.net/D12111
Summary:
This is the first step to fix unit tests and bugs for universal
compactiion. I added universal compaction option to ChangeOptions(), and
fixed all unit tests calling ChangeOptions(). Some of these tests
obviously assume more than 1 level and check file number/values in level
1 or above levels. I set kSkipUniversalCompaction for these tests.
The major bug I found is manual compaction with universal compaction never stops. I have put a fix for
it.
I have also set universal compaction as the default compaction and found
at least 20+ unit tests failing. I haven't looked into the details. The
next step is to check all unit tests without calling ChangeOptions().
Test Plan: make all check
Reviewers: dhruba, haobo
Differential Revision: https://reviews.facebook.net/D12051
Dhruba Borthakur