Commit Graph

891 Commits

Author SHA1 Message Date
haoyuhuang
3778470061 Block cache analyzer: Compute correlation of features and human readable trace file. (#5596)
Summary:
- Compute correlation between a few features and predictions, e.g., number of accesses since the last access vs number of accesses till the next access on a block.
- Output human readable trace file so python can consume it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5596

Test Plan: make clean && USE_CLANG=1 make check -j32

Differential Revision: D16373200

Pulled By: HaoyuHuang

fbshipit-source-id: c848d26bc2e9210461f317d7dbee42d55be5a0cc
2019-07-22 17:51:34 -07:00
Yanqin Jin
a78503bd6c Temporarily disable snapshot list refresh for atomic flush stress test (#5581)
Summary:
Atomic flush test started to fail after https://github.com/facebook/rocksdb/issues/5099. Then https://github.com/facebook/rocksdb/issues/5278 provided a fix after
which the same error occurred much less frequently. However it still occur
occasionally. Not sure what the root cause is. This PR disables the feature of
snapshot list refresh, and we should keep an eye on the failure in the future.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5581

Differential Revision: D16295985

Pulled By: riversand963

fbshipit-source-id: c9e62e65133c52c21b07097de359632ca62571e4
2019-07-22 14:38:16 -07:00
sdong
6bb3b4b567 ldb idump to support non-default column families. (#5594)
Summary:
ldb idump now only works for default column family. Extend it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5594

Test Plan: Compile and run the tool against a multiple CF DB.

Differential Revision: D16380684

fbshipit-source-id: bfb8af36fdad1806837c90aaaab492d71528aceb
2019-07-19 11:36:59 -07:00
haoyuhuang
8a008d4170 Block access tracing: Trace referenced key for Get on non-data blocks. (#5548)
Summary:
This PR traces the referenced key for Get for all types of blocks. This is useful when evaluating hybrid row-block caches.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5548

Test Plan: make clean && USE_CLANG=1 make check -j32

Differential Revision: D16157979

Pulled By: HaoyuHuang

fbshipit-source-id: f6327411c9deb74e35e22a35f66cdbae09ab9d87
2019-07-17 13:05:58 -07:00
Levi Tamasi
3bde41b5a3 Move the filter readers out of the block cache (#5504)
Summary:
Currently, when the block cache is used for the filter block, it is not
really the block itself that is stored in the cache but a FilterBlockReader
object. Since this object is not pure data (it has, for instance, pointers that
might dangle, including in one case a back pointer to the TableReader), it's not
really sharable. To avoid the issues around this, the current code erases the
cache entries when the TableReader is closed (which, BTW, is not sufficient
since a concurrent TableReader might have picked up the object in the meantime).
Instead of doing this, the patch moves the FilterBlockReader out of the cache
altogether, and decouples the filter reader object from the filter block.
In particular, instead of the TableReader owning, or caching/pinning the
FilterBlockReader (based on the customer's settings), with the change the
TableReader unconditionally owns the FilterBlockReader, which in turn
owns/caches/pins the filter block. This change also enables us to reuse the code
paths historically used for data blocks for filters as well.

Note:
Eviction statistics for filter blocks are temporarily broken. We plan to fix this in a
separate phase.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5504

Test Plan: make asan_check

Differential Revision: D16036974

Pulled By: ltamasi

fbshipit-source-id: 770f543c5fb4ed126fd1e04bfd3809cf4ff9c091
2019-07-16 13:14:58 -07:00
haoyuhuang
68d43b4d30 A python script to plot graphs for cvs files generated by block_cache_trace_analyzer
Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/5563

Test Plan: Manually run the script on files generated by block_cache_trace_analyzer.

Differential Revision: D16214400

Pulled By: HaoyuHuang

fbshipit-source-id: 94485eed995e9b2b63e197c5dfeb80129fa7897f
2019-07-12 18:56:20 -07:00
haoyuhuang
3e9c5a3523 Block cache analyzer: Add more stats (#5516)
Summary:
This PR provides more command line options for block cache analyzer to better understand block cache access pattern.
-analyze_bottom_k_access_count_blocks
-analyze_top_k_access_count_blocks
-reuse_lifetime_labels
-reuse_lifetime_buckets
-analyze_callers
-access_count_buckets
-analyze_blocks_reuse_k_reuse_window
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5516

Test Plan: make clean && COMPILE_WITH_ASAN=1 make check -j32

Differential Revision: D16037440

Pulled By: HaoyuHuang

fbshipit-source-id: b9a4ac0d4712053fab910732077a4d4b91400bc8
2019-07-12 16:55:34 -07:00
haoyuhuang
1a59b6e2a9 Cache simulator: Add a ghost cache for admission control and a hybrid row-block cache. (#5534)
Summary:
This PR adds a ghost cache for admission control. Specifically, it admits an entry on its second access.
It also adds a hybrid row-block cache that caches the referenced key-value pairs of a Get/MultiGet request instead of its blocks.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5534

Test Plan: make clean && COMPILE_WITH_ASAN=1 make check -j32

Differential Revision: D16101124

Pulled By: HaoyuHuang

fbshipit-source-id: b99edda6418a888e94eb40f71ece45d375e234b1
2019-07-11 12:43:29 -07:00
Yanqin Jin
f786b4a5b4 Improve result print on atomic flush stress test failure (#5549)
Summary:
When atomic flush stress test fails, we print internal keys within the range with mismatched key/values for all column families.

Test plan (on devserver)
Manually hack the code to randomly insert wrong data. Run the test.
```
$make clean && COMPILE_WITH_TSAN=1 make -j32 db_stress
$./db_stress -test_atomic_flush=true -ops_per_thread=10000
```
Check that proper error messages are printed, as follows:
```
2019/07/08-17:40:14  Starting verification
Verification failed
Latest Sequence Number: 190903
[default] 000000000000050B => 56290000525350515E5F5C5D5A5B5859
[3] 0000000000000533 => EE100000EAEBE8E9E6E7E4E5E2E3E0E1FEFFFCFDFAFBF8F9
Internal keys in CF 'default', [000000000000050B, 0000000000000533] (max 8)
  key 000000000000050B seq 139920 type 1
  key 0000000000000533 seq 0 type 1
Internal keys in CF '3', [000000000000050B, 0000000000000533] (max 8)
  key 0000000000000533 seq 0 type 1
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5549

Differential Revision: D16158709

Pulled By: riversand963

fbshipit-source-id: f07fa87763f87b3bd908da03c956709c6456bcab
2019-07-09 16:27:22 -07:00
sdong
aa0367aabb Allow ldb to open DB as secondary (#5537)
Summary:
Right now ldb can open running DB through read-only DB. However, it might leave info logs files to the read-only DB directory. Add an option to open the DB as secondary to avoid it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5537

Test Plan:
Run
./ldb scan  --max_keys=10 --db=/tmp/rocksdbtest-2491/dbbench --secondary_path=/tmp --no_value --hex
and
./ldb get 0x00000000000000103030303030303030 --hex --db=/tmp/rocksdbtest-2491/dbbench --secondary_path=/tmp
against a normal db_bench run and observe the output changes. Also observe that no new info logs files are created under /tmp/rocksdbtest-2491/dbbench.
Run without --secondary_path and observe that new info logs created under /tmp/rocksdbtest-2491/dbbench.

Differential Revision: D16113886

fbshipit-source-id: 4e09dec47c2528f6ca08a9e7a7894ba2d9daebbb
2019-07-09 12:51:28 -07:00
Tim Hatch
a6a9213a36 Fix interpreter lines for files with python2-only syntax.
Reviewed By: lisroach

Differential Revision: D15362271

fbshipit-source-id: 48fab12ab6e55a8537b19b4623d2545ca9950ec5
2019-07-09 10:51:37 -07:00
sdong
872a261ffc db_stress to print some internal keys after verification failure (#5543)
Summary:
Print out some more information when db_tress fails with verification failures to help debugging problems.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5543

Test Plan:
Manually ingest some failures and observe the outputs are like this:

Verification failed
[default] 0000000000199A5A => 7C3D000078797A7B74757677707172736C6D6E6F68696A6B
[6] 000000000019C8BD => 65380000616063626D6C6F6E69686B6A
internal keys in default CF [0000000000199A5A, 000000000019C8BD] (max 8)
  key 0000000000199A5A seq 179246 type 1
  key 000000000019C8BD seq 163970 type 1
Lastest Sequence Number: 292234

Differential Revision: D16153717

fbshipit-source-id: b33fa50a828c190cbf8249a37955432044f92daf
2019-07-08 13:36:37 -07:00
sdong
e4dcf5fd22 db_bench to add a new "benchmark" to print out all stats history (#5532)
Summary:
Sometimes it is helpful to fetch the whole history of stats after benchmark runs. Add such an option
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5532

Test Plan: Run the benchmark manually and observe the output is as expected.

Differential Revision: D16097764

fbshipit-source-id: 10b5b735a22a18be198b8f348be11f11f8806904
2019-07-03 20:03:28 -07:00
haoyuhuang
66464d1fde Remove multiple declarations o kMicrosInSecond.
Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/5526

Test Plan:
OPT=-g V=1 make J=1 unity_test -j32
make clean && make -j32

Differential Revision: D16079315

Pulled By: HaoyuHuang

fbshipit-source-id: 294ab439cf0db8dd5da44e30eabf0cbb2bb8c4f6
2019-07-01 15:15:12 -07:00
Eli Pozniansky
3e6c185381 Formatting fixes in db_bench_tool (#5525)
Summary:
Formatting fixes in db_bench_tool that were accidentally omitted
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5525

Test Plan: Unit tests

Differential Revision: D16078516

Pulled By: elipoz

fbshipit-source-id: bf8df0e3f08092a91794ebf285396d9b8a335bb9
2019-07-01 14:57:28 -07:00
Eli Pozniansky
f872009237 Fix from some C-style casting (#5524)
Summary:
Fix from some C-style casting in bloom.cc and ./tools/db_bench_tool.cc
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5524

Differential Revision: D16075626

Pulled By: elipoz

fbshipit-source-id: 352948885efb64a7ef865942c75c3c727a914207
2019-07-01 13:05:34 -07:00
haoyuhuang
9f0bd56889 Cache simulator: Refactor the cache simulator so that we can add alternative policies easily (#5517)
Summary:
This PR creates cache_simulator.h file. It contains a CacheSimulator that runs against a block cache trace record. We can add alternative cache simulators derived from CacheSimulator later. For example, this PR adds a PrioritizedCacheSimulator that inserts filter/index/uncompressed dictionary blocks with high priority.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5517

Test Plan: make clean && COMPILE_WITH_ASAN=1 make check -j32

Differential Revision: D16043689

Pulled By: HaoyuHuang

fbshipit-source-id: 65f28ed52b866ffb0e6eceffd7f9ca7c45bb680d
2019-07-01 12:46:32 -07:00
Yanqin Jin
c360675750 Add secondary instance to stress test (#5479)
Summary:
This PR allows users to run stress tests on secondary instance.

Test plan (on devserver)
```
./db_stress -ops_per_thread=100000 -enable_secondary=true -threads=32 -secondary_catch_up_one_in=10000 -clear_column_family_one_in=1000 -reopen=100
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5479

Differential Revision: D16074325

Pulled By: riversand963

fbshipit-source-id: c0ed959e7b6c7cda3efd0b3070ab379de3b29f1c
2019-07-01 11:49:50 -07:00
sdong
10bae8ceb3 Add more release versions to tools/check_format_compatible.sh (#5518)
Summary:
tools/check_format_compatible.sh is lagged behind. Catch up.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5518

Test Plan: Run the command

Differential Revision: D16063180

fbshipit-source-id: d063eb42df9653dec06a2cf0fb982b8a60ca3d2f
2019-06-28 17:41:58 -07:00
Aaron Gao
5c2f13fb14 add create_column_family and drop_column_family cmd to ldb tool (#5503)
Summary:
`create_column_family` cmd already exists but was somehow missed in the help message.
also add `drop_column_family` cmd which can drop a cf without opening db.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5503

Test Plan: Updated existing ldb_test.py to test deleting a column family.

Differential Revision: D16018414

Pulled By: lightmark

fbshipit-source-id: 1fc33680b742104fea86b10efc8499f79e722301
2019-06-27 11:11:48 -07:00
haoyuhuang
554a6456aa Block cache trace analysis: Write time series graphs in csv files (#5490)
Summary:
This PR adds a feature in block cache trace analysis tool to write statistics into csv files.
1. The analysis tool supports grouping the number of accesses per second by various labels, e.g., block, column family, block type, or a combination of them.
2. It also computes reuse distance and reuse interval.

Reuse distance: The cumulated size of unique blocks read between two consecutive accesses on the same block.
Reuse interval: The time between two consecutive accesses on the same block.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5490

Differential Revision: D15901322

Pulled By: HaoyuHuang

fbshipit-source-id: b5454fea408a32757a80be63de6fe1c8149ca70e
2019-06-24 20:42:12 -07:00
Yanqin Jin
1bfeffab2d Stop printing after verification fails (#5493)
Summary:
Stop verification and printing once verification fails.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5493

Differential Revision: D15928992

Pulled By: riversand963

fbshipit-source-id: 699feac034a217d57280aa3fb50f5aba06adf317
2019-06-20 22:16:58 -07:00
haoyuhuang
705b8eecb4 Add more callers for table reader. (#5454)
Summary:
This PR adds more callers for table readers. These information are only used for block cache analysis so that we can know which caller accesses a block.
1. It renames the BlockCacheLookupCaller to TableReaderCaller as passing the caller from upstream requires changes to table_reader.h and TableReaderCaller is a more appropriate name.
2. It adds more table reader callers in table/table_reader_caller.h, e.g., kCompactionRefill, kExternalSSTIngestion, and kBuildTable.

This PR is long as it requires modification of interfaces in table_reader.h, e.g., NewIterator.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5454

Test Plan: make clean && COMPILE_WITH_ASAN=1 make check -j32.

Differential Revision: D15819451

Pulled By: HaoyuHuang

fbshipit-source-id: b6caa704c8fb96ddd15b9a934b7e7ea87f88092d
2019-06-20 14:31:48 -07:00
haoyuhuang
2e8ad03ab3 Add more stats in the block cache trace analyzer (#5482)
Summary:
This PR adds more stats in the block cache trace analyzer.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5482

Differential Revision: D15883553

Pulled By: HaoyuHuang

fbshipit-source-id: 6d440e4f657af75690420102d532d0ee1ed4e9cf
2019-06-18 18:38:42 -07:00
Huisheng Liu
92f631da33 replace sprintf with its safe version snprintf (#5475)
Summary:
sprintf is unsafe and has buffer overrun risk. Replace it with the safer version snprintf where buffer size is supplied to avoid overrun.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5475

Differential Revision: D15879481

Pulled By: sagar0

fbshipit-source-id: 7ae1958ffc9727fa50261dfbb98ddd74e70a72d8
2019-06-18 16:42:26 -07:00
haoyuhuang
bcfc53b436 Block cache tracing: Fix minor bugs with downsampling and some benchmark results. (#5473)
Summary:
As the code changes for block cache tracing are almost complete, I did a benchmark to compare the performance when block cache tracing is enabled/disabled.

 With 1% downsampling ratio, the performance overhead of block cache tracing is negligible. When we trace all block accesses, the throughput drops by 6 folds with 16 threads issuing random reads and all reads are served in block cache.

Setup:
RocksDB:    version 6.2
Date:       Mon Jun 17 17:11:13 2019
CPU:        24 * Intel Core Processor (Skylake)
CPUCache:   16384 KB
Keys:       20 bytes each
Values:     100 bytes each (100 bytes after compression)
Entries:    10000000
Prefix:    20 bytes
Keys per prefix:    0
RawSize:    1144.4 MB (estimated)
FileSize:   1144.4 MB (estimated)
Write rate: 0 bytes/second
Read rate: 0 ops/second
Compression: NoCompression
Compression sampling rate: 0
Memtablerep: skip_list
Perf Level: 1

I ran the readrandom workload for 1 minute. Detailed throughput results:  (ops/second)
Sample rate 0: no block cache tracing.
Sample rate 1: trace all block accesses.
Sample rate 100: trace accesses 1% blocks.
1 thread |   |   |  -- | -- | -- | --
Sample rate | 0 | 1 | 100
1 MB block cache size | 13,094 | 13,166 | 13,341
10 GB block cache size | 202,243 | 188,677 | 229,182

16 threads |   |   |  -- | -- | -- | --
Sample rate | 0 | 1 | 100
1 MB block cache size | 208,761 | 178,700 | 201,872
10 GB block cache size | 2,645,996 | 426,295 | 2,587,605
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5473

Differential Revision: D15869479

Pulled By: HaoyuHuang

fbshipit-source-id: 7ae802abe84811281a6af8649f489887cd7c4618
2019-06-17 17:59:02 -07:00
haoyuhuang
2d1dd5bce7 Support computing miss ratio curves using sim_cache. (#5449)
Summary:
This PR adds a BlockCacheTraceSimulator that reports the miss ratios given different cache configurations. A cache configuration contains "cache_name,num_shard_bits,cache_capacities". For example, "lru, 1, 1K, 2K, 4M, 4G".

When we replay the trace, we also perform lookups and inserts on the simulated caches.
In the end, it reports the miss ratio for each tuple <cache_name, num_shard_bits, cache_capacity> in a output file.

This PR also adds a main source block_cache_trace_analyzer so that we can run the analyzer in command line.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5449

Test Plan:
Added tests for block_cache_trace_analyzer.
COMPILE_WITH_ASAN=1 make check -j32.

Differential Revision: D15797073

Pulled By: HaoyuHuang

fbshipit-source-id: aef0c5c2e7938f3e8b6a10d4a6a50e6928ecf408
2019-06-17 16:41:12 -07:00
Zhongyi Xie
671d15cbdd Persistent Stats: persist stats history to disk (#5046)
Summary:
This PR continues the work in https://github.com/facebook/rocksdb/pull/4748 and https://github.com/facebook/rocksdb/pull/4535 by adding a new DBOption `persist_stats_to_disk` which instructs RocksDB to persist stats history to RocksDB itself. When statistics is enabled, and  both options `stats_persist_period_sec` and `persist_stats_to_disk` are set, RocksDB will periodically write stats to a built-in column family in the following form: key -> (timestamp in microseconds)#(stats name), value -> stats value. The existing API `GetStatsHistory` will detect the current value of `persist_stats_to_disk` and either read from in-memory data structure or from the hidden column family on disk.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5046

Differential Revision: D15863138

Pulled By: miasantreble

fbshipit-source-id: bb82abdb3f2ca581aa42531734ac799f113e931b
2019-06-17 15:21:50 -07:00
haoyuhuang
d43b4cd570 Integrate block cache tracing into db_bench (#5459)
Summary:
This PR integrates the block cache tracing into db_bench. It adds three command line arguments.
-block_cache_trace_file (Block cache trace file path.) type: string default: ""
-block_cache_trace_max_trace_file_size_in_bytes (The maximum block cache
trace file size in bytes. Block cache accesses will not be logged if the
trace file size exceeds this threshold. Default is 64 GB.) type: int64
default: 68719476736
-block_cache_trace_sampling_frequency (Block cache trace sampling
frequency, termed s. It uses spatial downsampling and samples accesses to
one out of s blocks.) type: int32 default: 1
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5459

Differential Revision: D15832031

Pulled By: HaoyuHuang

fbshipit-source-id: 0ecf2f2686557251fe741a2769b21170777efa3d
2019-06-17 11:08:21 -07:00
haoyuhuang
7a8d7358bb Integrate block cache tracer in block based table reader. (#5441)
Summary:
This PR integrates the block cache tracer into block based table reader. The tracer will write the block cache accesses using the trace_writer. The tracer is null in this PR so that nothing will be logged.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5441

Differential Revision: D15772029

Pulled By: HaoyuHuang

fbshipit-source-id: a64adb92642cd23222e0ba8b10d86bf522b42f9b
2019-06-14 17:40:31 -07:00
haoyuhuang
bb4178066d Integrate block cache tracer into db_impl (#5433)
Summary:
This PR integrates the block cache tracer class into db_impl.cc.
db_impl.cc contains a member variable of AtomicBlockCacheTraceWriter class and passes its reference to the block_based_table_reader.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5433

Differential Revision: D15728016

Pulled By: HaoyuHuang

fbshipit-source-id: 23d5659e8c82d556833dcc1a5558aac8c1f7db71
2019-06-13 15:43:10 -07:00
Maysam Yabandeh
f9842869cf Disable pipeline writes in stress test (#5445)
Summary:
The tsan crash tests are failing with a data race compliant with pipelined write option. Temporarily disable it until its concurrency issue are fixed.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5445

Differential Revision: D15783824

Pulled By: maysamyabandeh

fbshipit-source-id: 413a0c3230b86f524fc7eeea2cf8e8375406e65b
2019-06-12 11:12:36 -07:00
haoyuhuang
9bbccda01e First commit for block cache trace analyzer (#5425)
Summary:
This PR contains the first commit for block cache trace analyzer. It reads a block cache trace file and prints statistics of the traces.

We will extend this class to provide more functionalities.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5425

Differential Revision: D15709580

Pulled By: HaoyuHuang

fbshipit-source-id: 2f43bd2311f460ab569880819d95eeae217c20bb
2019-06-11 12:22:44 -07:00
Zhongyi Xie
d68f9f4580 simplify include directive involving inttypes (#5402)
Summary:
When using `PRIu64` type of printf specifier, current code base does the following:
```
#ifndef __STDC_FORMAT_MACROS
#define __STDC_FORMAT_MACROS
#endif
#include <inttypes.h>
```
However, this can be simplified to
```
#include <cinttypes>
```
as long as flag `-std=c++11` is used.
This should solve issues like https://github.com/facebook/rocksdb/issues/5159
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5402

Differential Revision: D15701195

Pulled By: miasantreble

fbshipit-source-id: 6dac0a05f52aadb55e9728038599d3d2e4b59d03
2019-06-06 13:56:07 -07:00
Siying Dong
5851cb7fdb Move util/trace_replay.* to trace_replay/ (#5376)
Summary:
util/ means for lower level libraries. trace_replay is highly integrated to DB and sometimes call DB. Move it out to a separate directory.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5376

Differential Revision: D15550938

Pulled By: siying

fbshipit-source-id: f46dce5ceffdc05a73f26379c7bb1b79ebe6c207
2019-06-03 13:25:26 -07:00
Siying Dong
000b9ec217 Move some logging related files to logging/ (#5387)
Summary:
Many logging related source files are under util/. It will be more structured if they are together.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5387

Differential Revision: D15579036

Pulled By: siying

fbshipit-source-id: 3850134ed50b8c0bb40a0c8ae1f184fa4081303f
2019-05-31 17:23:59 -07:00
Vijay Nadimpalli
49c5a12dbe Organizing rocksdb/db directory
Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/5390

Differential Revision: D15579388

Pulled By: vjnadimpalli

fbshipit-source-id: 5bfc95e31554b8ff05b97b76d6534113f527f366
2019-05-31 11:57:01 -07:00
Yanqin Jin
83f7a8eed0 Fix compilation error in LITE mode (#5391)
Summary:
Add macro ROCKSDB_LITE to fix compilation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5391

Differential Revision: D15574522

Pulled By: riversand963

fbshipit-source-id: 95aea83c5d9b2bf98a3ba0ef9167b63c9be2988b
2019-05-31 08:32:22 -07:00
Yanqin Jin
b9f5900658 Fix WAL replay by skipping old write batches (#5170)
Summary:
1. Fix a bug in WAL replay in which write batches with old sequence numbers are mistakenly inserted into memtables.
2. Add support for benchmarking secondary instance to db_bench_tool.
With changes made in this PR, we can start benchmarking secondary instance
using two processes. It is also possible to vary the frequency at which the
secondary instance tries to catch up with the primary. The info log of the
secondary can be found in a directory whose path can be specified with
'-secondary_path'.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5170

Differential Revision: D15564608

Pulled By: riversand963

fbshipit-source-id: ce97688ed3d33f69d3a0b9266ebbbbf887aa0ec8
2019-05-30 19:33:33 -07:00
Siying Dong
8843129ece Move some memory related files from util/ to memory/ (#5382)
Summary:
Move arena, allocator, and memory tools under util to a separate memory/ directory.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5382

Differential Revision: D15564655

Pulled By: siying

fbshipit-source-id: 9cd6b5d0d3d52b39606e19221fa154596e5852a5
2019-05-30 17:44:09 -07:00
Vijay Nadimpalli
50e470791d Organizing rocksdb/table directory by format
Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/5373

Differential Revision: D15559425

Pulled By: vjnadimpalli

fbshipit-source-id: 5d6d6d615582bedd96a4b879bb25d429a6de8b55
2019-05-30 14:51:11 -07:00
anand76
bd44ec2006 Fix reopen voting logic in db_stress when using MultiGet (#5374)
Summary:
When the --reopen option is non-zero, the DB is reopened after every ops_per_thread/(reopen+1) ops, with the check being done after every op. With MultiGet, we might do multiple ops in one iteration, which broke the logic that checked when to synchronize among the threads and reopen the DB. This PR fixes that logic.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5374

Differential Revision: D15559780

Pulled By: anand1976

fbshipit-source-id: ee6563a68045df7f367eca3cbc2500d3e26359ef
2019-05-30 11:41:08 -07:00
Siying Dong
e9e0101ca4 Move test related files under util/ to test_util/ (#5377)
Summary:
There are too many types of files under util/. Some test related files don't belong to there or just are just loosely related. Mo
ve them to a new directory test_util/, so that util/ is cleaner.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5377

Differential Revision: D15551366

Pulled By: siying

fbshipit-source-id: 0f5c8653832354ef8caa31749c0143815d719e2c
2019-05-30 11:25:51 -07:00
Siying Dong
545d206040 Move some file related files outside util/ (#5375)
Summary:
util/ means for lower level libraries, so it's a good idea to move the files which requires knowledge to DB out. Create a file/ and move some files there.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5375

Differential Revision: D15550935

Pulled By: siying

fbshipit-source-id: 61a9715dcde5386eebfb43e93f847bba1ae0d3f2
2019-05-29 20:47:06 -07:00
Maysam Yabandeh
eab4f49a2c WritePrepared: skip_concurrency_control option (#5330)
Summary:
This enables the user to set TransactionDBOptions::skip_concurrency_control so the standard `DB::Write(const WriteOptions& opts, WriteBatch* updates)` would skip the concurrency control. This would give higher throughput to the users who know their use case doesn't need concurrency control.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5330

Differential Revision: D15525932

Pulled By: maysamyabandeh

fbshipit-source-id: 68421ac1ba34f549a4a8de9ce4c2dccf6fb4b06b
2019-05-28 16:29:45 -07:00
Silver Chan
2095ae8858 fixed db_stress.cc build error (#5307)
Summary:
when building this file using Xcode 10.2.1 in MacOSX10.14, the compiler report this error:
`
rocksdb/tools/db_stress.cc:3613:33: error: implicit instantiation of
      undefined template 'std::__1::array<std::__1::basic_string<char>, 10>'
    std::array<std::string, 10> keys = {"0", "1", "2", "3", "4", "5", "6", "7", "8", "9"};
/usr/include/c++/v1/__tuple:223:64: note:
      template is declared here
template <class _Tp, size_t _Size> struct _LIBCPP_TEMPLATE_VIS array;
                                                               ^
1 error generated.
`
if including array, this error will be fixed.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5307

Differential Revision: D15475217

Pulled By: sagar0

fbshipit-source-id: b04a7658c2ca2573157028863b3a80f5ab52b9de
2019-05-23 14:03:25 -07:00
Zhichao Cao
a13026fb2f Added trace replay fast forward function (#5273)
Summary:
In the current db_bench trace replay, the replay process strictly follows the timestamp to issue the queries. In some cases, user does not care about the time. Therefore, fast forward is needed for users to speed up the replay process.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5273

Differential Revision: D15389232

Pulled By: zhichao-cao

fbshipit-source-id: 735d629b9d2a167b05af3e4fa0ddf9d5d0be1806
2019-05-16 20:21:18 -07:00
anand76
6492430eaf Fix a bug in db_stress and an incorrect assertion in FilePickerMultiGet (#5301)
Summary:
This PR has two fixes for crash test failures -
1. Fix a bug in TestMultiGet() in db_stress that was passing list of key to MultiGet() in the wrong order, thus ensuring that actual values don't match expected values
2. Remove an incorrect assertion in FilePickerMultiGet::GetNextFileInLevelWithKeys() that checks that files in a level are in sorted order. This is not true with MultiGet(), especially if there are duplicate keys and we may have to go back one file for the next key. Furthermore, this assertion makes more sense when a new version is created, rather than at lookup time

Test -
asan_crash and ubsan_crash tests
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5301

Differential Revision: D15337383

Pulled By: anand1976

fbshipit-source-id: 35092cb15bbc1700e5e823cbe07bfa62f1e9e6c6
2019-05-14 11:58:04 -07:00
Maysam Yabandeh
f383641a1d Unordered Writes (#5218)
Summary:
Performing unordered writes in rocksdb when unordered_write option is set to true. When enabled the writes to memtable are done without joining any write thread. This offers much higher write throughput since the upcoming writes would not have to wait for the slowest memtable write to finish. The tradeoff is that the writes visible to a snapshot might change over time. If the application cannot tolerate that, it should implement its own mechanisms to work around that. Using TransactionDB with WRITE_PREPARED write policy is one way to achieve that. Doing so increases the max throughput by 2.2x without however compromising the snapshot guarantees.
The patch is prepared based on an original by siying
Existing unit tests are extended to include unordered_write option.

Benchmark Results:
```
TEST_TMPDIR=/dev/shm/ ./db_bench_unordered --benchmarks=fillrandom --threads=32 --num=10000000 -max_write_buffer_number=16 --max_background_jobs=64 --batch_size=8 --writes=3000000 -level0_file_num_compaction_trigger=99999 --level0_slowdown_writes_trigger=99999 --level0_stop_writes_trigger=99999 -enable_pipelined_write=false -disable_auto_compactions  --unordered_write=1
```
With WAL
- Vanilla RocksDB: 78.6 MB/s
- WRITER_PREPARED with unordered_write: 177.8 MB/s (2.2x)
- unordered_write: 368.9 MB/s (4.7x with relaxed snapshot guarantees)

Without WAL
- Vanilla RocksDB: 111.3 MB/s
- WRITER_PREPARED with unordered_write: 259.3 MB/s MB/s (2.3x)
- unordered_write: 645.6 MB/s (5.8x with relaxed snapshot guarantees)

- WRITER_PREPARED with unordered_write disable concurrency control: 185.3 MB/s MB/s (2.35x)

Limitations:
- The feature is not yet extended to `max_successive_merges` > 0. The feature is also incompatible with `enable_pipelined_write` = true as well as with `allow_concurrent_memtable_write` = false.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5218

Differential Revision: D15219029

Pulled By: maysamyabandeh

fbshipit-source-id: 38f2abc4af8780148c6128acdba2b3227bc81759
2019-05-13 17:47:21 -07:00
Yi Wu
92c60547fe db_bench: fix hang on IO error (#5300)
Summary:
db_bench will wait indefinitely if there's background error. Fix by pass `abs_time_us` to cond var.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5300

Differential Revision: D15319945

Pulled By: miasantreble

fbshipit-source-id: 0034fb7f6ec7c3303c4ccf26e54c20fbdac8ab44
2019-05-13 11:30:35 -07:00
anand76
181bb43f08 Fix bugs in FilePickerMultiGet (#5292)
Summary:
This PR fixes a couple of bugs in FilePickerMultiGet that were causing db_stress test failures. The failures were caused by -
1. Improper handling of a key that matches the user key portion of an L0 file's largest key. In this case, the curr_index_in_curr_level file index in L0 for that key was getting incremented, but batch_iter_ was not advanced. By design, all keys in a batch are supposed to be checked against an L0 file before advancing to the next L0 file. Not advancing to the next key in the batch was causing a double increment of curr_index_in_curr_level due to the same key being processed again
2. Improper handling of a key that matches the user key portion of the largest key in the last file of L1 and higher. This was resulting in a premature end to the processing of the batch for that level when the next key in the batch is a duplicate. Typically, the keys in MultiGet will not be duplicates, but its good to handle that case correctly

Test -
asan_crash
make check
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5292

Differential Revision: D15282530

Pulled By: anand1976

fbshipit-source-id: d1a6a86e0af273169c3632db22a44d79c66a581f
2019-05-09 13:18:00 -07:00
anand76
930bfa5750 Disable MultiGet from db_stress (#5284)
Summary:
Disable it for now until we can get stress tests to pass consistently.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5284

Differential Revision: D15230727

Pulled By: anand1976

fbshipit-source-id: 239baacdb3c4cd4fb7c4447f7582b9042501d752
2019-05-06 18:26:50 -07:00
Maysam Yabandeh
6a40ee5eb1 Refresh snapshot list during long compactions (2nd attempt) (#5278)
Summary:
Part of compaction cpu goes to processing snapshot list, the larger the list the bigger the overhead. Although the lifetime of most of the snapshots is much shorter than the lifetime of compactions, the compaction conservatively operates on the list of snapshots that it initially obtained. This patch allows the snapshot list to be updated via a callback if the compaction is taking long. This should let the compaction to continue more efficiently with much smaller snapshot list.
For simplicity, to avoid the feature is disabled in two cases: i) When more than one sub-compaction are sharing the same snapshot list, ii) when Range Delete is used in which the range delete aggregator has its own copy of snapshot list.
This fixes the reverted https://github.com/facebook/rocksdb/pull/5099 issue with range deletes.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5278

Differential Revision: D15203291

Pulled By: maysamyabandeh

fbshipit-source-id: fa645611e606aa222c7ce53176dc5bb6f259c258
2019-05-03 17:30:22 -07:00
Zhongyi Xie
5d27d65bef multiget: fix memory issues due to vector auto resizing (#5279)
Summary:
This PR fixes three memory issues found by ASAN
* in db_stress, the key vector for MultiGet is created using `emplace_back` which could potentially invalidates references to the underlying storage (vector<string>) due to auto resizing. Fix by calling reserve in advance.
* Similar issue in construction of GetContext autovector in version_set.cc
* In multiget_context.h use T[] specialization for unique_ptr that holds a char array
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5279

Differential Revision: D15202893

Pulled By: miasantreble

fbshipit-source-id: 14cc2cda0ed64d29f2a1e264a6bfdaa4294ee75d
2019-05-03 15:58:43 -07:00
Zhongyi Xie
3e994809a1 fix implicit conversion error reported by clang check (#5277)
Summary:
fix the following clang check errors
```
tools/db_stress.cc:3609:30: error: implicit conversion loses integer precision: 'std::vector::size_type' (aka 'unsigned long') to 'int' [-Werror,-Wshorten-64-to-32]
    int num_keys = rand_keys.size();
        ~~~~~~~~   ~~~~~~~~~~^~~~~~
tools/db_stress.cc:3888:30: error: implicit conversion loses integer precision: 'std::vector::size_type' (aka 'unsigned long') to 'int' [-Werror,-Wshorten-64-to-32]
    int num_keys = rand_keys.size();
        ~~~~~~~~   ~~~~~~~~~~^~~~~~
2 errors generated.
make: *** [tools/db_stress.o] Error 1
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5277

Differential Revision: D15196620

Pulled By: miasantreble

fbshipit-source-id: d56b1420d4a9f1df875fc52877a5fbb342bc7cae
2019-05-03 10:02:27 -07:00
anand76
434ccf2df4 Add option to use MultiGet in db_stress (#5264)
Summary:
The new option will pick a batch size randomly in the range 1-64. It will then space the keys in the batch by random intervals.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5264

Differential Revision: D15175522

Pulled By: anand1976

fbshipit-source-id: c16baa69d0f1ff4cf53c55c813ddd82c8aeb58fc
2019-05-01 23:06:56 -07:00
Andrew Kryczka
b02d0c238d Init compression dict handle before reading meta-blocks (#5267)
Summary:
At least one of the meta-block loading functions (`ReadRangeDelBlock`)
uses the same block reading function (`NewDataBlockIterator`) as data
block reads, which means it uses the dictionary handle. However, the
dictionary handle was uninitialized while reading meta-blocks, causing
readers to receive an error. This situation was only noticed when
`cache_index_and_filter_blocks=true`.

This PR initializes the handle to null while reading meta-blocks to
prevent the error. It also adds support to `db_stress` /
`db_crashtest.py` for `cache_index_and_filter_blocks`.

Fixes #5263.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5267

Differential Revision: D15149264

Pulled By: maysamyabandeh

fbshipit-source-id: 991d38a306c62db5976778bfb050fa3cd4a0671b
2019-04-30 09:50:49 -07:00
Yanqin Jin
210b49cac9 Disable pipelined write in atomic flush stress test (#5266)
Summary:
Since currently pipelined write allows one thread to perform memtable writes
while another thread is traversing the `flush_scheduler_`, it will cause an
assertion failure in `FlushScheduler::Clear`. To unblock crash recoery tests,
we temporarily disable pipelined write when atomic flush is enabled.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5266

Differential Revision: D15142285

Pulled By: riversand963

fbshipit-source-id: a0c20fe4ac543e08feaed602414f982054df7831
2019-04-30 08:12:42 -07:00
Sagar Vemuri
3548e4220d Improve explicit user readahead performance (#5246)
Summary:
Improve the iterators performance when the user explicitly sets the readahead size via `ReadOptions.readahead_size`.

1. Stop creating new table readers when the user explicitly sets readahead size.
2. Make use of an internal buffer based on `FilePrefetchBuffer` instead of using `ReadaheadRandomAccessFileReader`, to handle the user readahead requests (for both buffered and direct io cases).
3. Add `readahead_size` to db_bench.

**Benchmarks:**
https://gist.github.com/sagar0/53693edc320a18abeaeca94ca32f5737

For 1 MB readahead, Buffered IO performance improves by 28% and Direct IO performance improves by 50%.
For 512KB readahead, Buffered IO performance improves by 30% and Direct IO performance improves by 67%.

**Test Plan:**
Updated `DBIteratorTest.ReadAhead` test to make sure that:
- no new table readers are created for iterators on setting ReadOptions.readahead_size
- At least "readahead" number of bytes are actually getting read on each iterator read.

TODO later:
- Use similar logic for compactions as well.
- This ties in nicely with #4052 and paves the way for removing ReadaheadRandomAcessFile later.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5246

Differential Revision: D15107946

Pulled By: sagar0

fbshipit-source-id: 2c1149729ca7d779e4e8b7710ba6f4e8cbfd3bea
2019-04-26 21:24:10 -07:00
Adam Retter
990b2f4cb3 Fix compilation on db_bench_tool.cc on Windows (#5227)
Summary:
I needed this change to be able to build the v6.0.1 release on Windows.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5227

Differential Revision: D15033815

Pulled By: sagar0

fbshipit-source-id: 579f3b8e694c34c0d43527eb2fa37175e37f5911
2019-04-23 11:16:51 -07:00
Fosco Marotto
6c2bf9e916 Add copyright headers per FB open-source checkup tool. (#5199)
Summary:
internal task: T35568575
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5199

Differential Revision: D14962794

Pulled By: gfosco

fbshipit-source-id: 93838ede6d0235eaecff90d200faed9a8515bbbe
2019-04-18 10:55:01 -07:00
Zhongyi Xie
baa5302447 Avoid double-compacting data in bottom level in manual compactions (#5138)
Summary:
Depending on the config, manual compaction (leveled compaction style) does following compactions:
L0->L1
L1->L2
...
Ln-1 -> Ln
Ln -> Ln
The final Ln -> Ln compaction is partly unnecessary as it recompacts all the files that were just generated by the Ln-1 -> Ln. We should avoid recompacting such files. This rule should be applied to Lmax only.
Resolves issue https://github.com/facebook/rocksdb/issues/4995
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5138

Differential Revision: D14940106

Pulled By: miasantreble

fbshipit-source-id: 8d3cf5507a17e76f3333cfd4bac5256d005636e5
2019-04-16 23:32:20 -07:00
Yi Wu
b70967aac7 db_bench: support seek to non-exist prefix (#5163)
Summary:
Add `--seek_missing_prefix` flag to db_bench to allow benchmarking seeking to non-existing prefix. Usage example:
```
./db_bench --db=/dev/shm/db_bench --use_existing_db=false --benchmarks=fillrandom --num=100000000 --prefix_size=9 --keys_per_prefix=10
./db_bench --db=/dev/shm/db_bench --use_existing_db=true --benchmarks=seekrandom --disable_auto_compactions=true --num=100000000 --prefix_size=9 --keys_per_prefix=10 --reads=1000 --prefix_same_as_start=true --seek_missing_prefix=true
```
Also adding `--total_order_seek` and `--prefix_same_as_start` flags.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5163

Differential Revision: D14935724

Pulled By: riversand963

fbshipit-source-id: 7c41023f007febe373eb1589861f215432a9e18a
2019-04-15 10:54:58 -07:00
Yanqin Jin
3189398c00 Fix bugs detected by clang analyzer (#5185)
Summary:
as titled. False positive included, fixed anyway to make the check
pass.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5185

Differential Revision: D14909384

Pulled By: riversand963

fbshipit-source-id: dc5177e72b1929ccfd6175a60e2cd7bdb9bd80f3
2019-04-12 10:45:56 -07:00
anand76
fefd4b98c5 Introduce a new MultiGet batching implementation (#5011)
Summary:
This PR introduces a new MultiGet() API, with the underlying implementation grouping keys based on SST file and batching lookups in a file. The reason for the new API is twofold - the definition allows callers to allocate storage for status and values on stack instead of std::vector, as well as return values as PinnableSlices in order to avoid copying, and it keeps the original MultiGet() implementation intact while we experiment with batching.

Batching is useful when there is some spatial locality to the keys being queries, as well as larger batch sizes. The main benefits are due to -
1. Fewer function calls, especially to BlockBasedTableReader::MultiGet() and FullFilterBlockReader::KeysMayMatch()
2. Bloom filter cachelines can be prefetched, hiding the cache miss latency

The next step is to optimize the binary searches in the level_storage_info, index blocks and data blocks, since we could reduce the number of key comparisons if the keys are relatively close to each other. The batching optimizations also need to be extended to other formats, such as PlainTable and filter formats. This also needs to be added to db_stress.

Benchmark results from db_bench for various batch size/locality of reference combinations are given below. Locality was simulated by offsetting the keys in a batch by a stride length. Each SST file is about 8.6MB uncompressed and key/value size is 16/100 uncompressed. To focus on the cpu benefit of batching, the runs were single threaded and bound to the same cpu to eliminate interference from other system events. The results show a 10-25% improvement in micros/op from smaller to larger batch sizes (4 - 32).

Batch   Sizes

1        | 2        | 4         | 8      | 16  | 32

Random pattern (Stride length 0)
4.158 | 4.109 | 4.026 | 4.05 | 4.1 | 4.074        - Get
4.438 | 4.302 | 4.165 | 4.122 | 4.096 | 4.075 - MultiGet (no batching)
4.461 | 4.256 | 4.277 | 4.11 | 4.182 | 4.14        - MultiGet (w/ batching)

Good locality (Stride length 16)
4.048 | 3.659 | 3.248 | 2.99 | 2.84 | 2.753
4.429 | 3.728 | 3.406 | 3.053 | 2.911 | 2.781
4.452 | 3.45 | 2.833 | 2.451 | 2.233 | 2.135

Good locality (Stride length 256)
4.066 | 3.786 | 3.581 | 3.447 | 3.415 | 3.232
4.406 | 4.005 | 3.644 | 3.49 | 3.381 | 3.268
4.393 | 3.649 | 3.186 | 2.882 | 2.676 | 2.62

Medium locality (Stride length 4096)
4.012 | 3.922 | 3.768 | 3.61 | 3.582 | 3.555
4.364 | 4.057 | 3.791 | 3.65 | 3.57 | 3.465
4.479 | 3.758 | 3.316 | 3.077 | 2.959 | 2.891

dbbench command used (on a DB with 4 levels, 12 million keys)-
TEST_TMPDIR=/dev/shm numactl -C 10  ./db_bench.tmp -use_existing_db=true -benchmarks="readseq,multireadrandom" -write_buffer_size=4194304 -target_file_size_base=4194304 -max_bytes_for_level_base=16777216 -num=12000000 -reads=12000000 -duration=90 -threads=1 -compression_type=none -cache_size=4194304000 -batch_size=32 -disable_auto_compactions=true -bloom_bits=10 -cache_index_and_filter_blocks=true -pin_l0_filter_and_index_blocks_in_cache=true -multiread_batched=true -multiread_stride=4
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5011

Differential Revision: D14348703

Pulled By: anand1976

fbshipit-source-id: 774406dab3776d979c809522a67bedac6c17f84b
2019-04-11 14:28:26 -07:00
datonli
f0edf9d575 #5145 , rename port/dirent.h to port/port_dirent.h to avoid compile err when use port dir as header dir output (#5152)
Summary:
mv port/dirent.h to port/port_dirent.h to avoid compile err when use port dir as header dir output
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5152

Differential Revision: D14779409

Pulled By: siying

fbshipit-source-id: d4162c47c979c6e8cc6a9e601802864ab3768ecb
2019-04-04 11:38:19 -07:00
Yanqin Jin
d77476ef55 Fix db_stress for custom env (#5122)
Summary:
Fix some hdfs-related code so that it can compile and run 'db_stress'
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5122

Differential Revision: D14675495

Pulled By: riversand963

fbshipit-source-id: cac280479efcf5451982558947eac1732e8bc45a
2019-03-28 19:20:27 -07:00
Siying Dong
2b4d5ceb47 Remove some "using std::..." from header files. (#5113)
Summary:
The code convention we are following, Google C++ Style, discourage
alias in header files, especially public headers:
https://google.github.io/styleguide/cppguide.html#Aliases
Remove some of them. Might removed some from .cc files as well to be consistent.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5113

Differential Revision: D14633030

Pulled By: siying

fbshipit-source-id: b990edc919d5de60295992284f980195e501d424
2019-03-27 10:28:21 -07:00
Yanqin Jin
9358178edc Support for single-primary, multi-secondary instances (#4899)
Summary:
This PR allows RocksDB to run in single-primary, multi-secondary process mode.
The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary.
Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`.

This PR has several components:
1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary.

2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue.

3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`.
3.1 Tail the primary's MANIFEST during recovery.
3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`.
3.3 Tailing WAL will be in a future PR.

4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899

Differential Revision: D14510945

Pulled By: riversand963

fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
2019-03-26 16:45:31 -07:00
Zhongyi Xie
52e6404e0f ldb command parsing: allow option values to contain equals signs (#5088)
Summary:
Right now ldb command doesn't allow cases where option values contain equals sign. For example,
```
ldb --db=/tmp/test scan --from='q=3' --max_keys=1
```
after parsing, ldb will have one option 'db', 'max_keys' and one flag 'from'.
This PR updates the parsing logic so that it now supports the above mentioned cases
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5088

Differential Revision: D14600869

Pulled By: miasantreble

fbshipit-source-id: c6ef518c74a98d7b6675ea5954ae08b1bda5554e
2019-03-25 13:23:11 -07:00
Shobhit Dayal
b45b1cde3e Feature for sampling and reporting compressibility (#4842)
Summary:
This is a feature to sample data-block compressibility and and report them as stats. 1 in N (tunable) blocks is sampled for compressibility using two algorithms:
1. lz4 or snappy for fast compression
2. zstd or zlib for slow but higher compression.

The stats are reported to the caller as raw-bytes and compressed-bytes. The block continues to be compressed for storage using the specified CompressionType.

The db_bench_tool how has a command line option for specifying the sampling rate. It's default value is 0 (no sampling). To test the overhead for a certain value, users can compare the performance of db_bench_tool, varying the sampling rate. It is unlikely to have a noticeable impact for high values like 20.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4842

Differential Revision: D13629011

Pulled By: shobhitdayal

fbshipit-source-id: 14ca668bcab6499b2a1734edf848eb62a4f4fafa
2019-03-18 12:15:34 -07:00
Andrew Kryczka
2263f86901 exercise WAL recycling in crash test (#5070)
Summary:
Since this feature affects the WAL behavior, it seems important our crash-recovery tests cover it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5070

Differential Revision: D14470085

Pulled By: miasantreble

fbshipit-source-id: 9b9682a718a926d57d055e0a5ec867efbd2eb9c1
2019-03-15 12:03:26 -07:00
Zhichao Cao
dcde292c3b Add the -try_process_corrupted_trace option to trace_analyzer (#5067)
Summary:
In the current trace_analyzer implementation, once the trace file has corrupted content, which can be caused by unexpected tracing operations or other reasons, trace_analyzer will print the error and stop analyzing.

By adding the -try_process_corrupted_trace option, user can try to process the corrupted trace file and get the analyzing results of the trace records from the beginning to the the first corrupted point in the trace file. Analyzing might fail even this option is enabled.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5067

Differential Revision: D14433037

Pulled By: zhichao-cao

fbshipit-source-id: d095233ba371726869af0def0cdee23b69896831
2019-03-14 20:03:01 -07:00
Andrew Kryczka
5a5c0492db ldb: set total_order_seek for scans (#5066)
Summary:
Without `total_order_seek=true`, using this command with `prefix_extractor` set skips over lots of keys.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5066

Differential Revision: D14425967

Pulled By: sagar0

fbshipit-source-id: f6f142733258d92604f920615be9266e1fe797f8
2019-03-12 13:10:39 -07:00
Zhichao Cao
05ebfebc17 Fixed the potential stack overflow of MixGraph in db_bench (#5051)
Summary:
In the MixGraph benchmark of db_bench, The max buffer size used for value of KV-pair might be extremely large (64MB), which might cause function stack overflow in some platforms, reduced to 1MB.

Added the finished ops printing in MixGraph benchmark.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5051

Differential Revision: D14379571

Pulled By: zhichao-cao

fbshipit-source-id: 24084fbe38f60f2902d9a40f6bc9a25e4e2c9bb9
2019-03-08 14:10:17 -08:00
Andrew Kryczka
18d2e4beb7 Run db_bench on database generated externally (#5017)
Summary:
Added an option, `-use_existing_keys`, which can be set to run
benchmarks against an arbitrary existing database. Now users can
benchmark against their actual database rather than synthetic data.

Before the run begins, it loads all the keys into memory, then uses that
set of keys rather than synthesizing new ones in `GenerateKeyFromInt`.
This is mainly intended for small-scale DBs where the memory consumption
is not a concern.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5017

Differential Revision: D14270303

Pulled By: riversand963

fbshipit-source-id: 6328df9dffb5e19170270dd00a69f4bbe424e5ed
2019-03-01 11:19:03 -08:00
Siying Dong
aef763b6d6 Make statistics's stats_level change thread-safe (#5030)
Summary:
Right now, users can change statistics.stats_level while DB is running, but TSAN may report
data race. We make stats_level_ to be atomic, and access them using accessors.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5030

Differential Revision: D14267519

Pulled By: siying

fbshipit-source-id: 37d7ebeff7a43a406230143422a16af899163f73
2019-03-01 10:42:09 -08:00
Maysam Yabandeh
0b80f6b380 WritePrepared: script to analyze stress test failures (#5033)
Summary:
This the hackish script we used to find the root cause of failures in transaction stress tests. It is not well-written and does not require rigorous reviewing but it is better than starting from scratch each time we observe an issue. The stress tests would just say that at which snapshots the sum of all the keys in a set is inconsistent with another set. To help debugging one need to know which key exactly returned inconsistent results. The script looks at the transactions between two conflicting snapshots, and performs thee changes manually to see for which key the read value was inconsistent.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5033

Differential Revision: D14280362

Pulled By: maysamyabandeh

fbshipit-source-id: d5826055c46711460ba81480d96cb5ea082814a5
2019-03-01 09:18:40 -08:00
Siying Dong
5e298f865b Add two more StatsLevel (#5027)
Summary:
Statistics cost too much CPU for some use cases. Add two stats levels
so that people can choose to skip two types of expensive stats, timers and
histograms.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5027

Differential Revision: D14252765

Pulled By: siying

fbshipit-source-id: 75ecec9eaa44c06118229df4f80c366115346592
2019-02-28 10:27:59 -08:00
Zhongyi Xie
c4f5d0aa15 add GetStatsHistory to retrieve stats snapshots (#4748)
Summary:
This PR adds public `GetStatsHistory` API to retrieve stats history in the form of an std map. The key of the map is the timestamp in microseconds when the stats snapshot is taken, the value is another std map from stats name to stats value (stored in std string). Two DBOptions are introduced: `stats_persist_period_sec` (default 10 minutes) controls the intervals between two snapshots are taken; `max_stats_history_count` (default 10) controls the max number of history snapshots to keep in memory. RocksDB will stop collecting stats snapshots if `stats_persist_period_sec` is set to 0.

(This PR is the in-memory part of https://github.com/facebook/rocksdb/pull/4535)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4748

Differential Revision: D13961471

Pulled By: miasantreble

fbshipit-source-id: ac836d401ecb84ea92216bf9966f969dedf4ad04
2019-02-20 15:52:54 -08:00
Zhongyi Xie
ed995c6a69 add whole key bloom filter support in memtables (#4985)
Summary:
MyRocks calls `GetForUpdate` on `INSERT`, for unique key check, and in almost all cases GetForUpdate returns empty result. For such cases, whole key bloom filter is helpful.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4985

Differential Revision: D14118257

Pulled By: miasantreble

fbshipit-source-id: d35cb7109c62fd5ad541a26968e3a3e16d3e85ea
2019-02-19 12:15:39 -08:00
Siying Dong
4db46aa2e6 Fix LITE Build (#4989)
Summary:
LITE mode has EventListener to be an empty class. However in db_bench,
it is used. When "override" is added to the functions, the build breaks. Fix it
by keeping the listener empty in LITE mode.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4989

Differential Revision: D14108132

Pulled By: siying

fbshipit-source-id: 80121aab35b1120e502b37b782301dd700692697
2019-02-15 16:13:11 -08:00
Aubin Sanyal
3231a2e581 Deprecate ttl option from CompactionOptionsFIFO (#4965)
Summary:
We introduced ttl option in CompactionOptionsFIFO when ttl-based file
deletion (compaction) was supported only as part of FIFO Compaction. But
with the extension of ttl semantics even to Level compaction,
CompactionOptionsFIFO.ttl can now be deprecated. Instead we will start
using ColumnFamilyOptions.ttl for FIFO compaction as well.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4965

Differential Revision: D14072960

Pulled By: sagar0

fbshipit-source-id: c98cc2ae695a28136295787cd88d36a220fc219e
2019-02-15 09:51:41 -08:00
Michael Liu
ca89ac2ba9 Apply modernize-use-override (2nd iteration)
Summary:
Use C++11’s override and remove virtual where applicable.
Change are automatically generated.

Reviewed By: Orvid

Differential Revision: D14090024

fbshipit-source-id: 1e9432e87d2657e1ff0028e15370a85d1739ba2a
2019-02-14 14:41:36 -08:00
Yanqin Jin
4fc442029a Avoid using kInAtomicGroup tag for single-cf op (#4981)
Summary:
if an operation just involves a single column family, then we do
not have to set the kInAtomicGroup tag when writing to MANIFEST. This change
can fix a compatibility test failure, i.e. 5.15 and earlier cannot recognize
kInAtomicGroup tag.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4981

Differential Revision: D14072687

Pulled By: riversand963

fbshipit-source-id: 46b0c61e399f16c6b7169de0b33430d0ed90d6d4
2019-02-13 18:33:42 -08:00
Andrew Kryczka
62f70f6d14 Reduce scope of compression dictionary to single SST (#4952)
Summary:
Our previous approach was to train one compression dictionary per compaction, using the first output SST to train a dictionary, and then applying it on subsequent SSTs in the same compaction. While this was great for minimizing CPU/memory/I/O overhead, it did not achieve good compression ratios in practice. In our most promising potential use case, moderate reductions in a dictionary's scope make a major difference on compression ratio.

So, this PR changes compression dictionary to be scoped per-SST. It accepts the tradeoff during table building to use more memory and CPU. Important changes include:

- The `BlockBasedTableBuilder` has a new state when dictionary compression is in-use: `kBuffered`. In that state it accumulates uncompressed data in-memory whenever `Add` is called.
- After accumulating target file size bytes or calling `BlockBasedTableBuilder::Finish`, a `BlockBasedTableBuilder` moves to the `kUnbuffered` state. The transition (`EnterUnbuffered()`) involves sampling the buffered data, training a dictionary, and compressing/writing out all buffered data. In the `kUnbuffered` state, a `BlockBasedTableBuilder` behaves the same as before -- blocks are compressed/written out as soon as they fill up.
- Samples are now whole uncompressed data blocks, except the final sample may be a partial data block so we don't breach the user's configured `max_dict_bytes` or `zstd_max_train_bytes`. The dictionary trainer is supposed to work better when we pass it real units of compression. Previously we were passing 64-byte KV samples which was not realistic.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4952

Differential Revision: D13967980

Pulled By: ajkr

fbshipit-source-id: 82bea6f7537e1529c7a1a4cdee84585f5949300f
2019-02-11 19:47:32 -08:00
Andrew Kryczka
1218704b61 Fix compression_zstd_max_train_bytes coverage in stress test (#4957)
Summary:
Previously `finalize_and_sanitize` function was always zeroing out `compression_zstd_max_train_bytes`. It was only supposed to do that when non-ZSTD compression was used. But since `--compression_type` was an unknown argument (i.e., one that `db_crashtest.py` does not recognize and blindly forwards to `db_stress`), `finalize_and_sanitize` could not tell whether ZSTD was used. This PR fixes it simply by making `--compression_type` a known argument with snappy as default (same as `db_stress`).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4957

Differential Revision: D13994302

Pulled By: ajkr

fbshipit-source-id: 1b0baea7331397822830970d3698642eb7a7df65
2019-02-11 14:56:39 -08:00
Siying Dong
cf3a671733 Remove cuckoo hash memtable (#4953)
Summary:
Cuckoo Hash is less useful than we initially expected. Remove it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4953

Differential Revision: D13979264

Pulled By: siying

fbshipit-source-id: 2a60afdaa989f045357398b43a1cc5d46f4492ed
2019-02-07 16:15:27 -08:00
Siying Dong
d9c9f3c809 db_bench: fix "micros/op" reporting (#4949)
Summary:
4985a9f73b (diff-e5276985b26a0551957144f4420a594bR511)
changes the meaning of latency reporting from running time per query, to elapse_time / #ops, without providing a reason why.
Considering that this is a counter-intuitive reporting, Reverting the change.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4949

Differential Revision: D13964684

Pulled By: siying

fbshipit-source-id: d6304d3d4b5a802daa292302623c7dbca9a680bc
2019-02-05 17:20:02 -08:00
Alexander Zinoviev
32a6dd9a41 Add a new CPU time counter to compaction report (#4889)
Summary:
Measure CPU time consumed for a compaction and report it in the stats report
Enable NowCPUNanos() to work for MacOS
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4889

Differential Revision: D13701276

Pulled By: zinoale

fbshipit-source-id: 5024e5bbccd4dd10fd90d947870237f436445055
2019-01-29 17:24:00 -08:00
zhichao-cao
e2547103fd Fix the build error caused by the dynamic array (#4918)
Summary:
In the MixGraph benchmark of db_bench #4788 , the char array is initialized with an argument from user's input, which can cause build error on some platforms. Also, the msg char array size can be potentially smaller than the printed data, which should be extended from 100 to 256.

Tested with make check.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4918

Differential Revision: D13844298

Pulled By: sagar0

fbshipit-source-id: 33c4809c5c4438f0a9f7b289d3f42e20c545bbab
2019-01-28 12:39:57 -08:00
Andrew Kryczka
e242fa4664 Add latest toolchain (gcc-8, etc.) build support for fbcode users (#4923)
Summary:
- When building with internal dependencies, specify this toolchain by setting `ROCKSDB_FBCODE_BUILD_WITH_PLATFORM007=1`
- It is not enabled by default. However, it is enabled for TSAN builds in CI since there is a known problem with TSAN in gcc-5: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71090
- I did not add support for Lua since (1) we agreed to deprecate it, and (2) we only have an internal build for v5.3 with this toolchain while that has breaking changes compared to our current version (v5.2).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4923

Differential Revision: D13827226

Pulled By: ajkr

fbshipit-source-id: 9aa3388ed3679777cfb15ef8cbcb83c07f62f947
2019-01-28 11:26:32 -08:00
Sagar Vemuri
0cead31d10 Fix Clang static analyzer warning in db_bench (#4910)
Summary:
Fixed clang static analyzer warning about division by 0.
```
ar: creating librocksdb_debug.a
tools/db_bench_tool.cc:4650:43: warning: Division by zero
      int pos = static_cast<int>(rand_num % range_);
                                 ~~~~~~~~~^~~~~~~~
1 warning generated.
make: *** [analyze] Error 1
```

This is from the new code I recently merged in ce8e88d2d7.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4910

Differential Revision: D13788037

Pulled By: sagar0

fbshipit-source-id: f48851dca85047c19fbb1a361e25ce643aa4c7ea
2019-01-23 13:33:02 -08:00
Zhongyi Xie
cbe0239270 add cast to avoid loss of precision error (#4906)
Summary:
this PR address the following error:
> tools/db_bench_tool.cc:4776:68: error: implicit conversion loses integer precision: 'int64_t' (aka 'long') to 'unsigned int' [-Werror,-Wshorten-64-to-32]
        s = db_with_cfh->db->Put(write_options_, key, gen.Generate(value_size));
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4906

Differential Revision: D13780185

Pulled By: miasantreble

fbshipit-source-id: 1c83a77d341099518c72f0f4a63e97ab9c4784b3
2019-01-22 22:44:17 -08:00
Zhichao Cao
ce8e88d2d7 Generate mixed workload with Get, Put, Seek in db_bench (#4788)
Summary:
Based on the specific workload models (key access distribution, value size distribution, and iterator scan length distribution, the QPS variation), the MixGraph benchmark generate the synthetic workload according to these distributions which can reflect the real-world workload characteristics.

After user enable the tracing function, they will get the trace file. By analyzing the trace file with the trace_analyzer tool, user can generate a set of statistic data files including. The *_accessed_key_stats.txt,  *-accessed_value_size_distribution.txt, *-iterator_length_distribution.txt, and *-qps_stats.txt are mainly used to fit the Matlab model fitting. After that, user can get the parameters of the workload distributions (the modeling details are described: [here](https://github.com/facebook/rocksdb/wiki/RocksDB-Trace%2C-Replay%2C-and-Analyzer))

The key access distribution follows the The two-term power model. The probability density function is: `f(x) = ax^{b}+c`. The corresponding parameters are key_dist_a, key_dist_b, and key_dist_c in db_bench

For the value size distribution and iterator scan length distribution, they both follow the Generalized Pareto Distribution. The probability density function is `f(x) = (1/sigma)(1+k*(x-theta)/sigma))^{-1-1/k)`. The parameters are: value_k, value_theta, value_sigma and iter_k, iter_theta, iter_sigma. For more information about the Generalized Pareto Distribution, users can find the [wiki](https://en.wikipedia.org/wiki/Generalized_Pareto_distribution) and [Matalb page](https://www.mathworks.com/help/stats/generalized-pareto-distribution.html)

As for the QPS, it follows the diurnal pattern. So Sine is a good model to fit it. `F(x) = sine_a*sin(sine_b*x + sine_c) + sine_d`. The trace_will tell you the average QPS in the print out resutls, which is sine_d. After user fit the "*-qps_stats.txt" to the Matlab model, user can get the sine_a, sine_b, and sine_c. By using the 4 parameters, user can control the QPS variation including the period, average, changes.

To use the bench mark, user can indicate the following parameters as examples:
```
-benchmarks="mixgraph" -key_dist_a=0.002312 -key_dist_b=0.3467 -value_k=0.9233 -value_sigma=226.4092 -iter_k=2.517 -iter_sigma=14.236 -mix_get_ratio=0.7 -mix_put_ratio=0.25 -mix_seek_ratio=0.05 -sine_mix_rate_interval_milliseconds=500 -sine_a=15000 -sine_b=1 -sine_d=20000
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4788

Differential Revision: D13573940

Pulled By: sagar0

fbshipit-source-id: e184c27e07b4f1bc0b436c2be36c5090c1fb0222
2019-01-22 10:44:26 -08:00
Andrew Kryczka
01013ae766 Digest ZSTD compression dictionary once when writing SST file (#4849)
Summary:
This is essentially a re-submission of #4251 with a few improvements:

- Split `CompressionDict` into two separate classes: `CompressionDict` and `UncompressionDict`
- Eliminated `Init` functions. Instead do all initialization work in constructors.
- Added test case for parallel DB open, which is the scenario where #4251 failed under TSAN.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4849

Differential Revision: D13606039

Pulled By: ajkr

fbshipit-source-id: 08c236059798c710db9cbf545fce0f371232d447
2019-01-18 19:12:57 -08:00
Siying Dong
4e37251b4d With ldb --try_load_options and wal_dir doesn't exist, ignore it (#4875)
Summary:
LDB is frequently used to exam data copied. wal_dir in option file is not modified and it usually points to the path it copied from.
The user experience will be better if when ldb sees wal_dir pointed by the option file doesn't exist, rather than fail, just ignore it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4875

Differential Revision: D13643173

Pulled By: siying

fbshipit-source-id: 2e64d4ea2ec49a6794b9a706b7fc1ba901128bb8
2019-01-11 16:48:32 -08:00
Yanqin Jin
ffc9f84649 Free memory after use
Summary: Pull Request resolved: https://github.com/facebook/rocksdb/pull/4857

Differential Revision: D13602688

Pulled By: riversand963

fbshipit-source-id: 993419a6afb982a7a701ff71daebebb4b4a6b265
2019-01-08 17:19:09 -08:00
Yanqin Jin
e686caffec Remove unnecessary assersion in AtomicFlushStressTest::TestCheckpoint (#4846)
Summary:
as titled.
We can remove the assersion because we do not perform verification in
AtomicFlushStressTest::TestCheckpoint for similar reasons to TestGet, TestPut,
etc.
Therefore, we override TestCheckpoint in AtomicFlushStressTest so that the
assertion `rand_column_families.size() == rand_keys.size()' is removed, and we
do not verify the DB in this function.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4846

Differential Revision: D13583377

Pulled By: riversand963

fbshipit-source-id: 03647f3da67e27a397413fd666e3bb43003bf596
2019-01-07 16:47:26 -08:00
Huachao Huang
74f7d7551e tools: use provided options instead of the default (#4839)
Summary:
The current implementation hardcode the default options in different
places, which makes it impossible to support other environments (like
encrypted environment).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4839

Differential Revision: D13573578

Pulled By: sagar0

fbshipit-source-id: 76b58b4b758902798d10ff2f52d9f39abff015e7
2019-01-03 11:23:49 -08:00