You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Piotr Nowojski (Jira)" <ji...@apache.org> on 2020/01/06 08:56:00 UTC
[jira] [Closed] (FLINK-15103) Performance regression on 3.12.2019
in various benchmarks
[ https://issues.apache.org/jira/browse/FLINK-15103?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Piotr Nowojski closed FLINK-15103.
----------------------------------
Resolution: Won't Do
I've spent sometime bisecting this issue on the remote machine. For one thing, it seems like there were two regressions in yours [~xintongsong] commits, first one a bit smaller, that is really difficult to pick up from the benchmarking noise. I think the second one bigger was caused by either *4b8ed64* or *6681e11*, but that's also not sure, because of the benchmarking noise.
I don't like that I still do not understand what has exactly happened, but I'm more and more convinced that this must be some benign change in the amount of allocated buffers/memory or something like that. I have a feeling that spending more time on this is an overkill, so I'm closing this ticket.
I have already reconfigured the micro benchmarks [to always use 1000 network buffers|https://github.com/dataArtisans/flink-benchmarks/pull/49], which as expected improved results by quite a lot.
> Performance regression on 3.12.2019 in various benchmarks
> ---------------------------------------------------------
>
> Key: FLINK-15103
> URL: https://issues.apache.org/jira/browse/FLINK-15103
> Project: Flink
> Issue Type: Bug
> Components: Benchmarks
> Reporter: Piotr Nowojski
> Assignee: Piotr Nowojski
> Priority: Blocker
> Labels: pull-request-available
> Fix For: 1.10.0
>
>
> Various benchmarks show a performance regression that happened on December 3rd:
> [arrayKeyBy (probably the most easily visible)|http://codespeed.dak8s.net:8000/timeline/#/?exe=1&ben=arrayKeyBy&env=2&revs=200&equid=off&quarts=on&extr=on]
> [tupleKeyBy|http://codespeed.dak8s.net:8000/timeline/#/?exe=1&ben=tupleKeyBy&env=2&revs=200&equid=off&quarts=on&extr=on]
> [twoInputMapSink|http://codespeed.dak8s.net:8000/timeline/#/?exe=1&ben=twoInputMapSink&env=2&revs=200&equid=off&quarts=on&extr=on]
> [globalWindow (small one)|http://codespeed.dak8s.net:8000/timeline/#/?exe=1&ben=globalWindow&env=2&revs=200&equid=off&quarts=on&extr=on]
> and possible others.
> Probably somewhere between those commits: -8403fd4- 2d67ee0..60b3f2f
--
This message was sent by Atlassian Jira
(v8.3.4#803005)