Commit Graph

43 Commits

Author SHA1 Message Date
Zhichao Cao
82a70e1470 Trace MultiGet Keys and CF_IDs to the trace file (#8421)
Summary:
Tracing the MultiGet information including timestamp, keys, and CF_IDs to the trace file for analyzing and replay.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8421

Test Plan: make check, add test to trace_analyzer_test

Reviewed By: anand1976

Differential Revision: D29221195

Pulled By: zhichao-cao

fbshipit-source-id: 30c677d6c39ab31ef4bbdf7e0d1fa1fd79f295ff
2021-06-18 15:04:05 -07:00
Akanksha Mahajan
689b13e639 Add request_id in IODebugContext. (#8045)
Summary:
Add request_id in IODebugContext which will be populated by
    underlying FileSystem for IOTracing purposes. Update IOTracer to trace
    request_id in the tracing records. Provided API
    IODebugContext::SetRequestId which will set the request_id and enable
    tracing for request_id. The API hides the implementation and underlying
    file system needs to call this API directly.

Update DB::StartIOTrace API and remove redundant Env* from the
    argument as its not used and DB already has Env that is passed down to
    IOTracer.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8045

Test Plan: Update unit test.

Differential Revision: D26899871

Pulled By: akankshamahajan15

fbshipit-source-id: 56adef52ee5af0fb3060b607c3af1ec01635fa2b
2021-04-01 13:14:51 -07:00
mrambacher
a34dafe5ff Fix clang analyze for trace replace GetPayload (#8121)
Summary:
For some branches, I see an error during analyze on this code.  I do not know why it is not persistent, but this should address the error:

Logic error | Result of operation is garbage or undefined | trace_replay.cc | Replay | 436 | 30 | View Report

DecodeCFAndKey(trace.payload, &get_payload.cf_id, &get_payload.get_key);
--
433 | } else {
434 | TracerHelper::DecodeGetPayload(&trace, &get_payload);
  | 25←Calling 'TracerHelper::DecodeGetPayload'→ | 25 | ← | Calling 'TracerHelper::DecodeGetPayload' | →
25 | ← | Calling 'TracerHelper::DecodeGetPayload' | →
  | 29←Returning from 'TracerHelper::DecodeGetPayload'→ | 29 | ← | Returning from 'TracerHelper::DecodeGetPayload' | →
29 | ← | Returning from 'TracerHelper::DecodeGetPayload' | →
435 | }
436 | if (get_payload.cf_id > 0 &&
  | 30←The left operand of '>' is a garbage value | 30 | ← | The left operand of '>' is a garbage value
30 | ← | The left operand of '>' is a garbage value
437 | cf_map_.find(get_payload.cf_id) == cf_map_.end()) {
438 | return Status::Corruption("Invalid Column Family ID.");
439 | }

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8121

Reviewed By: zhichao-cao

Differential Revision: D27366022

Pulled By: mrambacher

fbshipit-source-id: 309c05dbab08cd7ab7f15389e8456f09196f37f6
2021-03-26 17:48:31 -07:00
mrambacher
3dff28cf9b Use SystemClock* instead of std::shared_ptr<SystemClock> in lower level routines (#8033)
Summary:
For performance purposes, the lower level routines were changed to use a SystemClock* instead of a std::shared_ptr<SystemClock>.  The shared ptr has some performance degradation on certain hardware classes.

For most of the system, there is no risk of the pointer being deleted/invalid because the shared_ptr will be stored elsewhere.  For example, the ImmutableDBOptions stores the Env which has a std::shared_ptr<SystemClock> in it.  The SystemClock* within the ImmutableDBOptions is essentially a "short cut" to gain access to this constant resource.

There were a few classes (PeriodicWorkScheduler?) where the "short cut" property did not hold.  In those cases, the shared pointer was preserved.

Using db_bench readrandom perf_level=3 on my EC2 box, this change performed as well or better than 6.17:

6.17: readrandom   :      28.046 micros/op 854902 ops/sec;   61.3 MB/s (355999 of 355999 found)
6.18: readrandom   :      32.615 micros/op 735306 ops/sec;   52.7 MB/s (290999 of 290999 found)
PR: readrandom   :      27.500 micros/op 871909 ops/sec;   62.5 MB/s (367999 of 367999 found)

(Note that the times for 6.18 are prior to revert of the SystemClock).

Pull Request resolved: https://github.com/facebook/rocksdb/pull/8033

Reviewed By: pdillinger

Differential Revision: D27014563

Pulled By: mrambacher

fbshipit-source-id: ad0459eba03182e454391b5926bf5cdd45657b67
2021-03-15 04:34:11 -07:00
Zhichao Cao
b0fd1cc45a Introduce a new trace file format (v 0.2) for better extension (#7977)
Summary:
The trace file record and payload encode is fixed, which requires complex backward compatibility resolving. This PR introduce a new trace file format, which makes it easier to add new entries to the payload and does not have backward compatible issues. V 0.1 is still supported in this PR. Added the tracing for lower_bound and upper_bound for iterator.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7977

Test Plan: make check. tested with old trace file in replay and analyzing.

Reviewed By: anand1976

Differential Revision: D26529948

Pulled By: zhichao-cao

fbshipit-source-id: ebb75a127ce3c07c25a1ccc194c551f917896a76
2021-02-18 23:05:35 -08:00
mrambacher
12f1137355 Add a SystemClock class to capture the time functions of an Env (#7858)
Summary:
Introduces and uses a SystemClock class to RocksDB.  This class contains the time-related functions of an Env and these functions can be redirected from the Env to the SystemClock.

Many of the places that used an Env (Timer, PerfStepTimer, RepeatableThread, RateLimiter, WriteController) for time-related functions have been changed to use SystemClock instead.  There are likely more places that can be changed, but this is a start to show what can/should be done.  Over time it would be nice to migrate most (if not all) of the uses of the time functions from the Env to the SystemClock.

There are several Env classes that implement these functions.  Most of these have not been converted yet to SystemClock implementations; that will come in a subsequent PR.  It would be good to unify many of the Mock Timer implementations, so that they behave similarly and be tested similarly (some override Sleep, some use a MockSleep, etc).

Additionally, this change will allow new methods to be introduced to the SystemClock (like https://github.com/facebook/rocksdb/issues/7101 WaitFor) in a consistent manner across a smaller number of classes.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7858

Reviewed By: pdillinger

Differential Revision: D26006406

Pulled By: mrambacher

fbshipit-source-id: ed10a8abbdab7ff2e23d69d85bd25b3e7e899e90
2021-01-25 22:09:11 -08:00
Akanksha Mahajan
1d226018af In IOTracing, add filename with each operation in trace file. (#7885)
Summary:
1. In IOTracing, add filename with each IOTrace record. Filename is stored in file object (Tracing Wrappers).
         2. Change the logic of figuring out which additional information (file_size,
            length, offset etc) needs to be store with each operation
            which is different for different operations.
            When new information will be added in future (depends on operation),
            this change would make the future additions simple.

Logic: In IOTraceRecord, io_op_data is added and its
         bitwise positions represent which additional information need
         to added in the record from enum IOTraceOp. Values in IOTraceOp represent bitwise positions.
         So if length and offset needs to be stored (IOTraceOp::kIOLen
         is 1 and IOTraceOp::kIOOffset is 2), position 1 and 2 (from rightmost bit) will be set
         and io_op_data will contain 110.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7885

Test Plan: Updated io_tracer_test and verified the trace file manually.

Reviewed By: anand1976

Differential Revision: D25982353

Pulled By: akankshamahajan15

fbshipit-source-id: ebfc5539cc0e231d7794a6b42b73f5403e360b22
2021-01-25 14:37:35 -08:00
Akanksha Mahajan
fbac1b3f80 Add tests in ASSERT_STATUS_CHECKED (#7793)
Summary:
add io_tracer_parser_test and prefetch_test under
ASSERT_STATUS_CHECKED

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7793

Test Plan: ASSERT_STATUS_CHECKED=1 make check -j64

Reviewed By: jay-zhuang

Differential Revision: D25673464

Pulled By: akankshamahajan15

fbshipit-source-id: 50e0b6f17160ddda206a521a7b47ee33e699a2d4
2020-12-22 10:31:13 -08:00
Jay Zhuang
861b0d1a49 Fix Windows build in block_cache_tracer_test (#7795)
Summary:
The test was added to cmake in https://github.com/facebook/rocksdb/issues/7783

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7795

Reviewed By: akankshamahajan15

Differential Revision: D25671010

Pulled By: jay-zhuang

fbshipit-source-id: 2146ff9559cdd7266c4d78476672488c62654a6d
2020-12-21 15:22:16 -08:00
Peter Dillinger
08552b19d3 Genericize and clean up FastRange (#7436)
Summary:
A generic algorithm in progress depends on a templatized
version of fastrange, so this change generalizes it and renames
it to fit our style guidelines, FastRange32, FastRange64, and now
FastRangeGeneric.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7436

Test Plan: added a few more test cases

Reviewed By: jay-zhuang

Differential Revision: D23958153

Pulled By: pdillinger

fbshipit-source-id: 8c3b76101653417804997e5f076623a25586f3e8
2020-09-28 11:35:00 -07:00
Akanksha Mahajan
98ac6b646a Add IO Tracer Parser (#7333)
Summary:
Implement a parsing tool io_tracer_parser that takes IO trace file (binary file) with command line argument --io_trace_file and output file with --output_file and dumps the IO trace records in outputfile in human readable form.

Also added unit test cases that generates IO trace records and calls io_tracer_parse to parse those records.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7333

Test Plan:
make check -j64,
 Add unit test cases.

Reviewed By: anand1976

Differential Revision: D23772360

Pulled By: akankshamahajan15

fbshipit-source-id: 9c20519c189362e6663352d08863326f3e496271
2020-09-23 15:50:26 -07:00
mrambacher
b7e1c5213f Add some simulator cache and block tracer tests to ASSERT_STATUS_CHECKED (#7305)
Summary:
More tests now pass.  When in doubt, I added a TODO comment to check what should happen with an ignored error.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7305

Reviewed By: akankshamahajan15

Differential Revision: D23301262

Pulled By: ajkr

fbshipit-source-id: 5f120edc7393560aefc0633250277bbc7e8de9e6
2020-08-24 16:43:31 -07:00
Akanksha Mahajan
493f425e77 Add support to start and end IOTracing through DB APIs (#7203)
Summary:
1. Add support to start io tracing through DB::StartIOTrace(Env*, const TraceOptions&, std::unique_ptr<TraceWriter>&&) and end tracing through DB::EndIOTrace(). This doesn't trace DB::Open.

User side code:

//Open DB
DB::Open(options, dbname, &db);

/* Start tracing */
db->StartIOTrace(env, trace_opt, std::move(trace_writer));

/* Perform Operations */

/*End tracing*/
db->EndIOTrace();

2. Fix the build errors for Windows.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7203

Test Plan: make check -j64

Reviewed By: anand1976

Differential Revision: D22901947

Pulled By: akankshamahajan15

fbshipit-source-id: e59c0b785a802168e6f1aa028d99c224a35cb30c
2020-08-04 18:41:45 -07:00
Akanksha Mahajan
d93bd3ce25 Add FileSystem wrapper classes for IO tracing. (#7002)
Summary:
1. Add the wrapper classes FileSystemTracingWrapper, FSSequentialFileTracingWrapper, FSRandomAccessFileTracingWrapper, FSWritableFileTracingWrapper, FSRandomRWFileTracingWrapper that forward the calls to underlying storage system and then pass the file operation information to IOTracer. IOTracer dumps the record in binary format for tracing.
2. Add the wrapper classes FileSystemPtr, FSSequentialFilePtr, FSRandomAccessFilePtr, FSWritableFilePtr and FSRandomRWFilePtr that overload operator-> and return ptr to underlying storage system or Tracing wrapper class based on enabling/disabling of IO tracing. These classes are added to bypass Tracing Wrapper classes when we disable tracing.
3. Add enums in trace.h that distinguish which options need to be added for different file operations(Read, close, write etc) as part of tracing record.

Pull Request resolved: https://github.com/facebook/rocksdb/pull/7002

Test Plan: make check -j64

Reviewed By: anand1976

Differential Revision: D22127897

Pulled By: akankshamahajan15

fbshipit-source-id: 74cff58ce5661c9a3832dfaa52483f3b2d8565e0
2020-07-13 16:36:55 -07:00
sdong
50d6969816 Fix unity build broken by #7007 (#7024)
Summary:
https://github.com/facebook/rocksdb/pull/7007 broken the unity build. Fix it by moving the const inside the function
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7024

Test Plan: make unity and see it to build.

Reviewed By: zhichao-cao

Differential Revision: D22212028

fbshipit-source-id: 5daff7383b691808164d4745ab543238502d946b
2020-06-24 13:40:48 -07:00
Zhichao Cao
d739318ba7 Fix double define in IO_tracer (#7007)
Summary:
Fix the following error

"./trace_replay/io_tracer.h:20:20: error: redefinition of ‘const unsigned int rocksdb::{anonymous}::kCharSize’
 const unsigned int kCharSize = 1;
                    ^~~~~~~~~
In file included from unity.cc:177:
trace_replay/block_cache_tracer.cc:22:20: note: ‘const unsigned int rocksdb::{anonymous}::kCharSize’ previously defined here
 const unsigned int kCharSize = 1;"
Pull Request resolved: https://github.com/facebook/rocksdb/pull/7007

Reviewed By: akankshamahajan15

Differential Revision: D22142618

Pulled By: zhichao-cao

fbshipit-source-id: e6dcd51ccc21d1f58df52cdc7a1c88e54cf4f6e8
2020-06-22 10:20:13 -07:00
Akanksha Mahajan
552fd765b3 Add IOTracer reader, writer classes for reading/writing IO operations in a binary file (#6958)
Summary:
1. As part of IOTracing project, Add a class IOTracer,
IOTraceReader and IOTracerWriter that writes the file operations
information in a binary file. IOTrace Record contains record information
and right now it contains access_timestamp, file_operation, file_name,
io_status, len, offset and later other options will be added when file
system APIs will be call IOTracer.

2. Add few unit test cases that verify that reading and writing to a IO
Trace file is working properly and before start trace and after ending
trace nothing is added to the binary file.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6958

Test Plan:
1. make check -j64
                 2. New testcases for IOTracer.

Reviewed By: anand1976

Differential Revision: D21943375

Pulled By: akankshamahajan15

fbshipit-source-id: 3532204e2a3eab0104bf411ab142e3fdd4fbce54
2020-06-18 10:46:11 -07:00
Peter Dillinger
c7432cc3c0 Fix more defects reported by Coverity Scan (#6935)
Summary:
Mostly uninitialized values: some probably written before use, but some seem like bugs. Also, destructor needs to be virtual, and possible use-after-free in test
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6935

Test Plan: make check

Reviewed By: siying

Differential Revision: D21885484

Pulled By: pdillinger

fbshipit-source-id: e2e7cb0a0cf196f2b55edd16f0634e81f6cc8e08
2020-06-04 15:35:08 -07:00
Zhichao Cao
c8643edfc3 Fix multiple CF replay failure in db_bench replay (#6787)
Summary:
The multiple CF hash map is not passed to the multi-thread worker. When using multi-thread replay for multiple CFs, it will cause segment fault. Pass the cf_map to the argument.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6787

Test Plan: pass trace replay test.

Reviewed By: yhchiang

Differential Revision: D21339941

Pulled By: zhichao-cao

fbshipit-source-id: 434482b492287e6722c7cd5a706f057c5ec170ce
2020-05-01 00:03:38 -07:00
sdong
fdf882ded2 Replace namespace name "rocksdb" with ROCKSDB_NAMESPACE (#6433)
Summary:
When dynamically linking two binaries together, different builds of RocksDB from two sources might cause errors. To provide a tool for user to solve the problem, the RocksDB namespace is changed to a flag which can be overridden in build time.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6433

Test Plan: Build release, all and jtest. Try to build with ROCKSDB_NAMESPACE with another flag.

Differential Revision: D19977691

fbshipit-source-id: aa7f2d0972e1c31d75339ac48478f34f6cfcfb3e
2020-02-20 12:09:57 -08:00
Peter Dillinger
ca7ccbe2ea Misc hashing updates / upgrades (#5909)
Summary:
- Updated our included xxhash implementation to version 0.7.2 (== the latest dev version as of 2019-10-09).
- Using XXH_NAMESPACE (like other fb projects) to avoid potential name collisions.
- Added fastrange64, and unit tests for it and fastrange32. These are faster alternatives to hash % range.
- Use preview version of XXH3 instead of MurmurHash64A for NPHash64
-- Had to update cache_test to increase probability of passing for any given hash function.
- Use fastrange64 instead of % with uses of NPHash64
-- Had to fix WritePreparedTransactionTest.CommitOfDelayedPrepared to avoid deadlock apparently caused by new hash collision.
- Set default seed for NPHash64 because specifying a seed rarely makes sense for it.
- Removed unnecessary include xxhash.h in a popular .h file
- Rename preview version of XXH3 to XXH3p for clarity and to ease backward compatibility in case final version of XXH3 is integrated.

Relying on existing unit tests for NPHash64-related changes. Each new implementation of fastrange64 passed unit tests when manipulating my local build to select it. I haven't done any integration performance tests, but I consider the improved performance of the pieces being swapped in to be well established.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5909

Differential Revision: D18125196

Pulled By: pdillinger

fbshipit-source-id: f6bf83d49d20cbb2549926adf454fd035f0ecc0d
2019-10-24 17:16:46 -07:00
Zhichao Cao
0933360644 Fix the potential memory leak in trace_replay (#5955)
Summary:
In the previous PR https://github.com/facebook/rocksdb/issues/5934 , in the while loop, if/else if is used without ending with else to free the object referenced by ra, it might cause potential memory leak (warning during compiling). Fix it by changing the last "else if" to "else".
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5955

Test Plan: pass make asan check, pass the USE_CLANG=1 TEST_TMPDIR=/dev/shm/rocksdb OPT=-g make -j64 analyze.

Differential Revision: D18071612

Pulled By: zhichao-cao

fbshipit-source-id: 51c00023d0c97c2921507254329aed55d56e1786
2019-10-22 16:39:46 -07:00
Zhichao Cao
7245fb5f63 Fix the potential memory leak of ReplayMultiThread (#5949)
Summary:
The pointer ra needs to be freed the status s returns not OK. In the previous  PR https://github.com/facebook/rocksdb/issues/5934  , the ra is not freed which might cause potential memory leak. Fix this issue by moving the clarification of ra inside the while loop and freeing it as desired.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5949

Test Plan: pass make asan check.

Differential Revision: D18045726

Pulled By: zhichao-cao

fbshipit-source-id: d5445b7b832c8bb1dafe008bafea7bfe9eb0b1ce
2019-10-21 15:05:01 -07:00
Zhichao Cao
526e3b9763 Enable trace_replay with multi-threads (#5934)
Summary:
In the current trace replay, all the queries are serialized and called by single threads. It may not simulate the original application query situations closely. The multi-threads replay is implemented in this PR. Users can set the number of threads to replay the trace. The queries generated according to the trace records are scheduled in the thread pool job queue.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5934

Test Plan: test with make check and real trace replay.

Differential Revision: D17998098

Pulled By: zhichao-cao

fbshipit-source-id: 87eecf6f7c17a9dc9d7ab29dd2af74f6f60212c8
2019-10-18 14:13:50 -07:00
haoyuhuang
3da225716c Block cache analyzer: Support reading from human readable trace file. (#5679)
Summary:
This PR adds support in block cache trace analyzer to read from human readable trace file. This is needed when a user does not have access to the binary trace file.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5679

Test Plan: USE_CLANG=1 make check -j32

Differential Revision: D16697239

Pulled By: HaoyuHuang

fbshipit-source-id: f2e29d7995816c389b41458f234ec8e184a924db
2019-08-09 13:13:54 -07:00
haoyuhuang
6e78fe3c8d Pysim more algorithms (#5644)
Summary:
This PR adds four more eviction policies.
- OPT [1]
- Hyperbolic caching [2]
- ARC [3]
- GreedyDualSize [4]

[1] L. A. Belady. 1966. A Study of Replacement Algorithms for a Virtual-storage Computer. IBM Syst. J. 5, 2 (June 1966), 78-101. DOI=http://dx.doi.org/10.1147/sj.52.0078
[2] Aaron Blankstein, Siddhartha Sen, and Michael J. Freedman. 2017. Hyperbolic caching: flexible caching for web applications. In Proceedings of the 2017 USENIX Conference on Usenix Annual Technical Conference (USENIX ATC '17). USENIX Association, Berkeley, CA, USA, 499-511.
[3] Nimrod Megiddo and Dharmendra S. Modha. 2003. ARC: A Self-Tuning, Low Overhead Replacement Cache. In Proceedings of the 2nd USENIX Conference on File and Storage Technologies (FAST '03). USENIX Association, Berkeley, CA, USA, 115-130.
[4] N. Young. The k-server dual and loose competitiveness for paging. Algorithmica, June 1994, vol. 11,(no.6):525-41. Rewritten version of ''On-line caching as cache size varies'', in The 2nd Annual ACM-SIAM Symposium on Discrete Algorithms, 241-250, 1991.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5644

Differential Revision: D16548817

Pulled By: HaoyuHuang

fbshipit-source-id: 838f76db9179f07911abaab46c97e1c929cfcd63
2019-08-06 18:50:59 -07:00
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
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
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
haoyuhuang
6edc5d0719 Block cache tracing: Associate a unique id with Get and MultiGet (#5514)
Summary:
This PR associates a unique id with Get and MultiGet. This enables us to track how many blocks a Get/MultiGet request accesses. We can also measure the impact of row cache vs block cache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5514

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

Differential Revision: D16032681

Pulled By: HaoyuHuang

fbshipit-source-id: 775b05f4440badd58de6667e3ec9f4fc87a0af4c
2019-07-03 19:35:41 -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
haoyuhuang
68b46a2e36 Block cache tracer: StartTrace return busy if trace is already started. (#5519)
Summary:
This PR is needed for integration into MyRocks. A second call on StartTrace returns Busy so that MyRocks may return an error to the user.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5519

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

Differential Revision: D16055476

Pulled By: HaoyuHuang

fbshipit-source-id: a51772fb0965c873922757eb470a332b1e02a91d
2019-06-30 20:03:01 -07:00
haoyuhuang
a8975b6245 Block cache tracer: Do not populate block cache trace record when tracing is disabled. (#5510)
Summary:
This PR makes sure that trace record is not populated when tracing is disabled.

Before this PR:
DB path: [/data/mysql/rocks_regression_tests/OPTIONS-myrocks-40-33-10000000/2019-06-26-13-04-41/db]
readwhilewriting :       9.803 micros/op 1550408 ops/sec;  107.9 MB/s (5000000 of 5000000 found)
Microseconds per read:
Count: 80000000 Average: 9.8045  StdDev: 12.64
Min: 1  Median: 7.5246  Max: 25343
Percentiles: P50: 7.52 P75: 12.10 P99: 37.44 P99.9: 75.07 P99.99: 133.60

After this PR:
DB path: [/data/mysql/rocks_regression_tests/OPTIONS-myrocks-40-33-10000000/2019-06-26-14-08-21/db]
readwhilewriting :       8.723 micros/op 1662882 ops/sec;  115.8 MB/s (5000000 of 5000000 found)
Microseconds per read:
Count: 80000000 Average: 8.7236  StdDev: 12.19
Min: 1  Median: 6.7262  Max: 25229
Percentiles: P50: 6.73 P75: 10.50 P99: 31.54 P99.9: 74.81 P99.99: 132.82
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5510

Differential Revision: D16016428

Pulled By: HaoyuHuang

fbshipit-source-id: 3b3d11e6accf207d18ec2545b802aa01ee65901f
2019-06-27 08:34:08 -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
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
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
89695bfbaa Remove unused variable (#5457)
Summary:
This PR removes the unused variable that causes CLANG build to fail.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5457

Differential Revision: D15825027

Pulled By: HaoyuHuang

fbshipit-source-id: 72c847c39ca310560efcbc5938cffa6f31164068
2019-06-14 09:17:09 -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
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
haoyuhuang
5efa0d6b0d Create a BlockCacheLookupContext to enable fine-grained block cache tracing. (#5421)
Summary:
BlockCacheLookupContext only contains the caller for now.
We will trace block accesses at five places:
1. BlockBasedTable::GetFilter.
2. BlockBasedTable::GetUncompressedDict.
3. BlockBasedTable::MaybeReadAndLoadToCache. (To trace access on data, index, and range deletion block.)
4. BlockBasedTable::Get. (To trace the referenced key and whether the referenced key exists in a fetched data block.)
5. BlockBasedTable::MultiGet. (To trace the referenced key and whether the referenced key exists in a fetched data block.)

We create the context at:
1. BlockBasedTable::Get. (kUserGet)
2. BlockBasedTable::MultiGet. (kUserMGet)
3. BlockBasedTable::NewIterator. (either kUserIterator, kCompaction, or external SST ingestion calls this function.)
4. BlockBasedTable::Open. (kPrefetch)
5. Index/Filter::CacheDependencies. (kPrefetch)
6. BlockBasedTable::ApproximateOffsetOf. (kCompaction or kUserApproximateSize).

I loaded 1 million key-value pairs into the database and ran the readrandom benchmark with a single thread. I gave the block cache 10 GB to make sure all reads hit the block cache after warmup. The throughput is comparable.
Throughput of this PR: 231334 ops/s.
Throughput of the master branch: 238428 ops/s.

Experiment setup:
RocksDB:    version 6.2
Date:       Mon Jun 10 10:42:51 2019
CPU:        24 * Intel Core Processor (Skylake)
CPUCache:   16384 KB
Keys:       20 bytes each
Values:     100 bytes each (100 bytes after compression)
Entries:    1000000
Prefix:    20 bytes
Keys per prefix:    0
RawSize:    114.4 MB (estimated)
FileSize:   114.4 MB (estimated)
Write rate: 0 bytes/second
Read rate: 0 ops/second
Compression: NoCompression
Compression sampling rate: 0
Memtablerep: skip_list
Perf Level: 1

Load command: ./db_bench --benchmarks="fillseq" --key_size=20 --prefix_size=20 --keys_per_prefix=0 --value_size=100 --statistics --cache_index_and_filter_blocks --cache_size=10737418240 --disable_auto_compactions=1 --disable_wal=1 --compression_type=none --min_level_to_compress=-1 --compression_ratio=1 --num=1000000

Run command: ./db_bench --benchmarks="readrandom,stats" --use_existing_db --threads=1 --duration=120 --key_size=20 --prefix_size=20 --keys_per_prefix=0 --value_size=100 --statistics --cache_index_and_filter_blocks --cache_size=10737418240 --disable_auto_compactions=1 --disable_wal=1 --compression_type=none --min_level_to_compress=-1 --compression_ratio=1 --num=1000000 --duration=120

TODOs:
1. Create a caller for external SST file ingestion and differentiate the callers for iterator.
2. Integrate tracer to trace block cache accesses.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5421

Differential Revision: D15704258

Pulled By: HaoyuHuang

fbshipit-source-id: 4aa8a55f8cb1576ffb367bfa3186a91d8f06d93a
2019-06-10 15:33:27 -07:00
haoyuhuang
aa71718ac3 Add block cache tracer. (#5410)
Summary:
This PR adds a help class block cache tracer to read/write block cache accesses. It uses the trace reader/writer to perform this task.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5410

Differential Revision: D15612843

Pulled By: HaoyuHuang

fbshipit-source-id: f30fd1e1524355ca87db5d533a5c086728b141ea
2019-06-06 11:24:39 -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