-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/28617/
-----------------------------------------------------------
(Updated Feb. 4, 2015, 11:38 p.m.)
Review request for Aurora, Kevin Sweeney and Bill Farner.
Changes
-------
Rebased and modified benchmarks to support victim-less preemption run.
Bugs: AURORA-909
https://issues.apache.org/jira/browse/AURORA-909
Repository: aurora
Description (updated)
-------
Modified the task offer/task matching logic to skip offer matching for tasks
previously vetoed statically.
Real life testing in vagrant (see pictures) shows close to 50% improvement in
task scheduling performance.
Testing with JMH shows over 97% better perf when testing with disabled
preemptor (1 scheduling loop):
```
Master
Benchmark
Mode Samples Score Error Units
o.a.a.b.SchedulingBenchmarks.ConstraintMismatchsSchedulingBenchmark.runBenchmark
avgt 100 8291046.074 ± 145251.995 ns/op
o.a.a.b.SchedulingBenchmarks.InsufficientResourcesSchedulingBenchmark.runBenchmark
avgt 100 7522269.050 ± 142446.265 ns/op
This RB
Benchmark
Mode Samples Score Error Units
o.a.a.b.SchedulingBenchmarks.ConstraintMismatchsSchedulingBenchmark.runBenchmark
avgt 100 204171.046 ± 3800.124 ns/op
o.a.a.b.SchedulingBenchmarks.InsufficientResourcesSchedulingBenchmark.runBenchmark
avgt 100 215854.129 ± 8959.851 ns/op
```
Testing with preemptor enabled and no tasks eligible for preemption gives
around 40% improvement (2 scheduling loops):
```
Master
Benchmark
Mode Samples Score Error Units
o.a.a.b.SchedulingBenchmarks.ConstraintMismatchsSchedulingBenchmark.runBenchmark
avgt 100 1767479.299 ± 26907.571 ns/op
o.a.a.b.SchedulingBenchmarks.InsufficientResourcesSchedulingBenchmark.runBenchmark
avgt 100 1538682.287 ± 119119.911 ns/op
This RB
Benchmark
Mode Samples Score Error Units
o.a.a.b.SchedulingBenchmarks.ConstraintMismatchsSchedulingBenchmark.runBenchmark
avgt 100 1105731.141 ± 10040.721 ns/op
o.a.a.b.SchedulingBenchmarks.InsufficientResourcesSchedulingBenchmark.runBenchmark
avgt 100 939230.662 ± 11091.505 ns/op
```
Testing with preemptor enabled and running the worst case possible scenario
(every slave is eligible and all tasks are victims) yields the least
improvement 2-3% (3 scheduling loops).
```
Master
Benchmark
Mode Samples Score Error Units
o.a.a.b.SchedulingBenchmarks.ConstraintMismatchsSchedulingBenchmark.runBenchmark
avgt 100 11043701.243 ± 40550.259 ns/op
o.a.a.b.SchedulingBenchmarks.InsufficientResourcesSchedulingBenchmark.runBenchmark
avgt 100 10478631.055 ± 178833.158 ns/op
o.a.a.b.SchedulingBenchmarks.PreemptorFallbackForLargeClusterBenchmark.runBenchmark
avgt 100 116258653.000 ± 403080.017 ns/op
This RB
Benchmark
Mode Samples Score Error Units
o.a.a.b.SchedulingBenchmarks.ConstraintMismatchsSchedulingBenchmark.runBenchmark
avgt 100 10886116.889 ± 193934.324 ns/op
o.a.a.b.SchedulingBenchmarks.InsufficientResourcesSchedulingBenchmark.runBenchmark
avgt 100 10182572.955 ± 35740.891 ns/op
o.a.a.b.SchedulingBenchmarks.PreemptorFallbackForLargeClusterBenchmark.runBenchmark
avgt 100 113656994.000 ± 424163.759 ns/op
```
Diffs (updated)
-----
src/jmh/java/org/apache/aurora/benchmark/SchedulingBenchmarks.java
8c11ef8bd6609f3e4d97ca154d922898f8362446
src/jmh/java/org/apache/aurora/benchmark/Tasks.java
1a35f9ee9e8e76def0f9bf5454cf8cbdf6a89c25
src/main/java/org/apache/aurora/scheduler/TaskVars.java
f017cdd26ca40138a7e141f21613ed567314c399
src/main/java/org/apache/aurora/scheduler/async/OfferQueue.java
f66383830140e5eaba436f35ebb5192eee65947a
src/main/java/org/apache/aurora/scheduler/async/TaskScheduler.java
ce47ff152e303fd2116bc3b9e91c0c1a8f76f258
src/main/java/org/apache/aurora/scheduler/filter/SchedulingFilter.java
6a43bcd1719e8aa32fd3fcb7387d0318c3c0b804
src/main/java/org/apache/aurora/scheduler/filter/SchedulingFilterImpl.java
f06fdaeb92e154d0982bdabed5df93e7bcba9048
src/main/java/org/apache/aurora/scheduler/state/TaskAssigner.java
e1c29747c9854cf75bf63f6f085cf40ca68989af
src/test/java/org/apache/aurora/scheduler/TaskVarsTest.java
4e7efb3c1214c3d193afd61f162713490eb8effb
src/test/java/org/apache/aurora/scheduler/async/OfferQueueImplTest.java
4cf602ad32b972c18eb5a81e9b2f59c67859bdb2
src/test/java/org/apache/aurora/scheduler/async/TaskSchedulerImplTest.java
5647349854a5e04de749c4d809684a0066d4da06
src/test/java/org/apache/aurora/scheduler/async/TaskSchedulerTest.java
6cc13231560996b144101eba36577f49017aba06
src/test/java/org/apache/aurora/scheduler/filter/SchedulingFilterImplTest.java
52ee7c1e3742d9315c7e7aaa77677121e1e9288d
src/test/java/org/apache/aurora/scheduler/state/TaskAssignerImplTest.java
411a55a8d85f60bb2703468f2d69b64b2736eee4
Diff: https://reviews.apache.org/r/28617/diff/
Testing
-------
./gradlew -Pq build
Tested in vagrant
File Attachments
----------------
NoStaticVetoFiltering.png
https://reviews.apache.org/media/uploaded/files/2014/12/03/7945c60b-4135-4016-a9bf-8d4815a4a573__NoStaticVetoFiltering.png
StaticVetoFiltering.png
https://reviews.apache.org/media/uploaded/files/2014/12/03/2f73b94a-5ba9-43b6-922e-e9e4ec18d0bb__StaticVetoFiltering.png
Thanks,
Maxim Khutornenko
