flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "vinoyang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-8946) TaskManager stop sending metrics after JobManager failover
Date Wed, 21 Mar 2018 11:12:00 GMT

    [ https://issues.apache.org/jira/browse/FLINK-8946?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16407756#comment-16407756

vinoyang commented on FLINK-8946:

[~till.rohrmann] I just tracked this issue from source code, For Flip-6, specific code in
*TaskExecutor*, the *TaskManagerMetricGroup's* instance be injected with the constructor parameter,
and the metricGroup's instance is created at *TaskManagerRunner.startTaskManager*. In *TaskExecutor*,
the instance is not be closed, so it seems like it's not a problem in Flip-6. 

> TaskManager stop sending metrics after JobManager failover
> ----------------------------------------------------------
>                 Key: FLINK-8946
>                 URL: https://issues.apache.org/jira/browse/FLINK-8946
>             Project: Flink
>          Issue Type: Bug
>          Components: Metrics, TaskManager
>    Affects Versions: 1.4.2
>            Reporter: Truong Duc Kien
>            Assignee: vinoyang
>            Priority: Critical
>             Fix For: 1.5.0
> Running in Yarn-standalone mode, when the Job Manager performs a failover, all TaskManager
that are inherited from the previous JobManager will not send metrics to the new JobManager
and other registered metric reporters.
> A cursory glance reveal that these line of code might be the cause
> [https://github.com/apache/flink/blob/a3478fdfa0f792104123fefbd9bdf01f5029de51/flink-runtime/src/main/scala/org/apache/flink/runtime/taskmanager/TaskManager.scala#L1082-L1086]
> Perhap the TaskManager close its metrics group when disassociating JobManager, but not
creating a new one on fail-over association ?

This message was sent by Atlassian JIRA

View raw message