You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Zhu Zhu (Jira)" <ji...@apache.org> on 2022/07/01 03:46:00 UTC

[jira] [Closed] (FLINK-28136) Implement ExecutionTimeBasedSlowTaskDetector

     [ https://issues.apache.org/jira/browse/FLINK-28136?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Zhu Zhu closed FLINK-28136.
---------------------------
    Resolution: Done

Done via ef07590403ad7448a523ef5da78da0306cccdae7

> Implement ExecutionTimeBasedSlowTaskDetector
> --------------------------------------------
>
>                 Key: FLINK-28136
>                 URL: https://issues.apache.org/jira/browse/FLINK-28136
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Runtime / Coordination
>            Reporter: Zhu Zhu
>            Assignee: wangwj
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 1.16.0
>
>
> In the first version of speculative execution, an ExecutionTimeBasedSlowTaskDetector will be used to detect slow tasks. For ExecutionTimeBasedSlowTaskDetector, if a task's execution time is much longer than that of most tasks of the same JobVertex, the task will be identified as slow. More specifically, it will compute an execution time baseline for each JobVertex. Tasks which execute longer than or equals to the baseline will be identified as slow tasks.



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