hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yongjun Zhang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-11363) Need more diagnosis info when seeing Slow waitForAckedSeqno
Date Wed, 01 Feb 2017 22:30:52 GMT

    [ https://issues.apache.org/jira/browse/HDFS-11363?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15849039#comment-15849039

Yongjun Zhang commented on HDFS-11363:

Thanks [~xiaochen] for the patch.

The patch is largely what I want, but some comments/suggestions:
1. print out the array in the current patch way won't work. suggest to use {{Arrays.toString(array)}}
2. Replace {{File:}} with {{File being written:}}
3. Replace {{datanodes:}} with {{Write pipeline datanodes:}}

Does that make sense to you?


> Need more diagnosis info when seeing Slow waitForAckedSeqno
> -----------------------------------------------------------
>                 Key: HDFS-11363
>                 URL: https://issues.apache.org/jira/browse/HDFS-11363
>             Project: Hadoop HDFS
>          Issue Type: Bug
>            Reporter: Yongjun Zhang
>            Assignee: Xiao Chen
>              Labels: supportability
>         Attachments: HDFS-11363.01.patch
> When client writes a file, it may get the following message when ACK doesn't get back
in a timely manner:
> WARN hdfs.DFSClient: Slow waitForAckedSeqno took 39264ms (threshold=30000ms)
> It would be nice to tell what file it's writing, and what DataNodes are in the pipeline,
together with this message, to facilitate investigating related performance issue.

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org

View raw message