Commit graph

20 commits

Author SHA1 Message Date
Dan Hermann
48594baf2b Add license reporting task
Fixes #9331
2018-05-02 15:35:42 +00:00
Armin
26fe96c038 MINOR: Remove unused CheckpointIOFactory
Fixes #9054
2018-02-05 07:13:10 +00:00
Armin
0a1a761442 CLEANUP: Remove redundant PageIOFactory
Fixes #9012
2018-01-23 20:22:25 +00:00
Armin
4091ddfdb4 Remove mock file io
Fixes #8999
2018-01-22 23:22:15 +00:00
Colin Surprenant
e75eca4483 maximize PQ read bach size by using same read timeout semantic as MQ #8702
add from another LongVector

use timedout read batch and added test for maximization

WIP maximized batch read size

fixes and cleanups

cleanups

use the await returned boolean to assess timeout

remove useless constant

use Callable<Void> and cosmetic fix
2017-12-05 13:40:35 -05:00
Armin
823feaf69a MINOR: Remove JMH main entrypoints
Fixes #8591
2017-11-06 13:46:24 +00:00
Armin
44e68b4627 MINOR: Clean up some dead code around PQ + PQ Benchmarks
Fixes #8232
2017-09-13 13:19:19 +00:00
Armin
f06e859717 #7998 Serialization benchmark and Afterburner removal
Fixes #8002
2017-08-16 16:08:09 +00:00
Armin
96ff365997 Fixes #7676 and generally improves the performance of org.logstash.StringInterpolation#evaluate by:
* Don't use regex to find the contents between `%{}`, it's way too inefficient for a simple case like this with all its allocations
* Flatten logic out of the many helper classes
* Use a reusable, thread-local `StringBuilder` to serialize into
* Don't use `.substring` for comparisons
* `static` `ObjectMapper`
* `org.joda.time.format.DateTimeFormat#forPattern` not being cached is not an issue, since it's cached in `org.joda.time.format.DateTimeFormat#cPatternCache`

* Added benchmark
   * ~15% faster for a few cases I ran, even for repeated patterns with interpolation + couldn't find any case even with date formatting, where this implementation is slower than the cached one
   *  no leaking on dynamic patterns

Fixes #7692
2017-07-25 18:29:10 +00:00
Armin
5b0c86c4a4 MINOR: Align JMH JVM settings with production default JVM settings
Fixes #7534
2017-06-25 17:20:20 +00:00
Pier-Hugues Pellerin
6a4766a01f Centralize the JRuby version
We have multiple places to declare the jruby version, this commit
centralize everything into the `version.yml` file. This is now used by
the ruby part and the gradle build file to correctly fetch the jruby
library.

Fixes #7434
2017-06-13 15:44:31 +00:00
Pier-Hugues Pellerin
6e6589e773 Use uniform version withing our gradle dependencies
Make sure we use the same provided and runtime JRuby with our gradle
projects

Fixes: #7421

Fixes #7425
2017-06-13 12:44:01 +00:00
Armin
941446a6dc PERFORMANCE: Queue RW benchmark for in memory queue
Fixes #7410
2017-06-10 15:05:13 +00:00
Armin
4926f5b6bb PERFORMANCE: Queue RW benchmark baseline
Fixes #7400
2017-06-09 17:22:39 +00:00
Armin
b6e051acbc #7317 Queue RW Benchmark
Fixes #7355
2017-06-08 13:38:41 +00:00
Armin
54fa8d06a0 MINOR: Add log config to benchmarks
Fixes #7359
2017-06-08 04:51:33 +00:00
Armin
e80bfe45e7 MINOR: Dry up gradle settings
Fixes #7298
2017-06-03 07:07:28 +00:00
Armin
c80eabbc9d #7111 cleaned up one case of redundant CRC32 digest and byte[] creation. Adjusted benchmark checkpoint interval to default value to be able to judge performance impact of the move.
Fixes #7166
2017-05-26 04:47:19 +00:00
Armin
1a2c0c8597 #7053 Refactor Queue Settings Class to Using a Builder Pattern
Fixes #7062
2017-05-18 15:57:43 +00:00
Armin
b6c855a0a6 Added JMH Benchmark Suite + Queue Benchmark
Fixes #7118
2017-05-18 06:31:17 +00:00