[
https://issues.apache.org/jira/browse/HDFS-10588?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15407277#comment-15407277
]
Wei-Chiu Chuang commented on HDFS-10588:
----------------------------------------
Looks good to me. Thanks to [~cheersyang] for finding the bug and the updated
patch.
The unit test failures are unrelated to this patch.
After removing the error logging, {{DiskBalancer#checkDiskBalancerEnabled}}
still throws an exception. if {{DiskBalancer#checkDiskBalancerEnabled}} is
invoked due to JMX, the exception is recorded as a debug message, so this won't
spam DataNode log. If {{DiskBalancer#checkDiskBalancerEnabled}} is invoked
because the client sends a diskbalancer command, the DataNode catches the
exception and log it at INFO level, and client also gets the exception.
I think the verbosity of the log message will be adequate. +1 but I'd like to
let it sit for a day to let any one else to comment.
> False alarm in datanode log - ERROR - Disk Balancer is not enabled
> ------------------------------------------------------------------
>
> Key: HDFS-10588
> URL: https://issues.apache.org/jira/browse/HDFS-10588
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: datanode, hdfs
> Reporter: Weiwei Yang
> Assignee: Weiwei Yang
> Attachments: HDFS-10588.001.patch, HDFS-10588.002.patch,
> HDFS-10588.003.patch, HDFS-10588.004.patch, HDFS-10588.005.patch
>
>
> Noticed error message in namenode log
> {code}2016-06-28 19:49:12,221 ERROR datanode.DiskBalancer
> (DiskBalancer.java:checkDiskBalancerEnabled(297)) - Disk Balancer is not
> enabled.
> {code}
> even with default configuration dfs.disk.balancer.enabled=false.This is
> triggered when accessing datanode web UI, because
> {{DataNode#getDiskBalancerStatus}} calls the check.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
