[
https://issues.apache.org/jira/browse/HDFS-13852?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16684759#comment-16684759
]
Akira Ajisaka commented on HDFS-13852:
--------------------------------------
The test failure is related to HADOOP-15916 and it should be backported to
HDFS-13891 branch.
> RBF: The DN_REPORT_TIME_OUT and DN_REPORT_CACHE_EXPIRE should be configured
> in RBFConfigKeys.
> ---------------------------------------------------------------------------------------------
>
> Key: HDFS-13852
> URL: https://issues.apache.org/jira/browse/HDFS-13852
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: federation, hdfs
> Affects Versions: 3.1.0, 2.9.1, 3.0.1
> Reporter: yanghuafeng
> Assignee: yanghuafeng
> Priority: Major
> Attachments: HDFS-13852-HDFS-13891.0.patch, HDFS-13852.001.patch,
> HDFS-13852.002.patch, HDFS-13852.003.patch, HDFS-13852.004.patch
>
>
> In the NamenodeBeanMetrics the router will invokes 'getDataNodeReport'
> periodically. And we can set the dfs.federation.router.dn-report.time-out and
> dfs.federation.router.dn-report.cache-expire to avoid time out. But when we
> start the router, the FederationMetrics will also invoke the method to get
> node usage. If time out error happened, we cannot adjust the parameter
> time_out. And the time_out in the FederationMetrics and NamenodeBeanMetrics
> should be the same.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
