[
https://issues.apache.org/jira/browse/BEAM-9436?focusedWorklogId=408856&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-408856
]
ASF GitHub Bot logged work on BEAM-9436:
----------------------------------------
Author: ASF GitHub Bot
Created on: 24/Mar/20 15:39
Start Date: 24/Mar/20 15:39
Worklog Time Spent: 10m
Work Description: echauchot commented on issue #11055: [BEAM-9436]
Improve GBK in spark structured streaming runner
URL: https://github.com/apache/beam/pull/11055#issuecomment-603314471
> > I have already given them. see above
> > > I could not make jenkins launch the script from the PR (!) So I ran it
locally here are the compared results: on a -Xms12g -Xmx20g JVM, 200000
elements of K=100K and V=100K
> > > before the change on GBK: run time 403.033s
> > > after the change on GBK: run time 283.426s
>
> @echauchot I saw this but it was not clear said which kind of tests it was
running. I'd expect to run Nexmark (as our Beam default standard) for such
performance testing. Also, it would be great, as @iemejia mentioned, to run it
in distributed environment (not only locally).
Yes I'm waiting for the run on a cluster that we discused with @jbonofre
@RyanSkraba.
Also I ran nexmark and times are similar (in PR nexmark does not print
graphs) that is why I ran pure load tests (nexmark does more than just GBK)
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 408856)
Time Spent: 11h 10m (was: 11h)
> Try to avoid elements list materialization in GBK
> -------------------------------------------------
>
> Key: BEAM-9436
> URL: https://issues.apache.org/jira/browse/BEAM-9436
> Project: Beam
> Issue Type: Improvement
> Components: runner-spark
> Reporter: Etienne Chauchot
> Assignee: Etienne Chauchot
> Priority: Major
> Labels: structured-streaming
> Time Spent: 11h 10m
> Remaining Estimate: 0h
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
