You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Jiayi Liao (Jira)" <ji...@apache.org> on 2020/02/11 15:15:00 UTC
[jira] [Comment Edited] (FLINK-16001) Avoid using Java Streams in
construction of ExecutionGraph
[ https://issues.apache.org/jira/browse/FLINK-16001?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17034538#comment-17034538 ]
Jiayi Liao edited comment on FLINK-16001 at 2/11/20 3:14 PM:
-------------------------------------------------------------
[~gjy] Thanks for your attention. Currently we're trying to optimize the job with topology of A(p=8000) -> B(p=8000), ALL_TO_ALL pattern, which takes JM nearly a minute to finish function {{attachJobGraph}}. Part of the cost is from duplicate calculating of {{PipelinedRegionComputeUtil.computePipelinedRegions}} in {{RegionPartitionReleaseStrategy}} and {{AdaptedRestartPipelinedRegionStrategyNG}}, which I can open another issue if you think the improvement is needed.
was (Author: wind_ljy):
[~gjy] Thanks for your attention. Currently we're trying to optimize the job with topology of A(p=8000) -> B(p=8000), ALL_TO_ALL pattern, which takes us nearly a minute to finish function {{attachJobGraph}}. Part of the cost is from duplicate calculating of {{PipelinedRegionComputeUtil.computePipelinedRegions}} in {{RegionPartitionReleaseStrategy}} and {{AdaptedRestartPipelinedRegionStrategyNG}}, which I can open another issue if you think the improvement is needed.
> Avoid using Java Streams in construction of ExecutionGraph
> ----------------------------------------------------------
>
> Key: FLINK-16001
> URL: https://issues.apache.org/jira/browse/FLINK-16001
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Coordination
> Affects Versions: 1.10.0
> Reporter: Jiayi Liao
> Priority: Major
>
> I think we should avoid {{Java Streams}} in construction of {{ExecutionGraph}} like function {{toPipelinedRegionsSet}} in {{PipelinedRegionComputeUtil}} because the job submission is definitely performance sensitive, especially when {{distinctRegions}} has a large cardinality.
> Also includes some other places in package {{org.apache.flink.runtime.executiongraph}}
> cc [~trohrmann] [~gjy] [~zhuzh]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)