You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Weijie Guo (Jira)" <ji...@apache.org> on 2024/04/24 03:53:00 UTC

[jira] [Commented] (FLINK-35166) Improve the performance of Hybrid Shuffle when enable memory decoupling

    [ https://issues.apache.org/jira/browse/FLINK-35166?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17840277#comment-17840277 ] 

Weijie Guo commented on FLINK-35166:
------------------------------------

master(1.20) via 712d9800e6c02e66b51747ca65331ad9e5295ffb.

> Improve the performance of Hybrid Shuffle when enable memory decoupling
> -----------------------------------------------------------------------
>
>                 Key: FLINK-35166
>                 URL: https://issues.apache.org/jira/browse/FLINK-35166
>             Project: Flink
>          Issue Type: Improvement
>          Components: Runtime / Network
>            Reporter: Jiang Xin
>            Assignee: Jiang Xin
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 1.20.0
>
>
> Currently, the tiered result partition creates the SortBufferAccumulator with the number of expected buffers as min(numSubpartitions+1, 512), thus the SortBufferAccumulator may obtain very few buffers when the parallelism is small. We can easily make the number of expected buffers 512 by default to have a better performance when the buffers are sufficient.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)