[
https://issues.apache.org/jira/browse/HDFS-15280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17111459#comment-17111459
]
Wei-Chiu Chuang commented on HDFS-15280:
----------------------------------------
This is basically creating a random backoff mechanism for DN to retry if NN
signals it is busy.
[~hexiaoqiao] or [~kihwal], thoughts?
> Datanode delay random time to report block if BlockManager is busy
> ------------------------------------------------------------------
>
> Key: HDFS-15280
> URL: https://issues.apache.org/jira/browse/HDFS-15280
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: namenode
> Reporter: Yang Yun
> Assignee: Yang Yun
> Priority: Minor
> Attachments: HDFS-15280.001.diff, HDFS-15280.002.patch
>
>
> When many Datanodes are reporting at the same time, the cluster may respond
> slowly. Limit the concurrent reporting number. If BlockManager is busy, it
> rejects new request and the Datanode delay a few random time to report.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
