[
https://issues.apache.org/jira/browse/HDFS-10246?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15230045#comment-15230045
]
zhangyubiao commented on HDFS-10246:
------------------------------------
[~mingma],thank you. I disable retry cache and set the
dfs.namenode.checkpoint.txns to 10000000,The standby become normal.
But I find the performance spikes still exits in every three min .
And we have dfs.ha.tail-edits.period value 60s and
the dfs.ha.log-roll.period value 120s. Should we reduce the value ?
I Update the Rpc CallQueueLength image with standby
> Standby NameNode dfshealth.jsp Response very slow
> -----------------------------------------------------
>
> Key: HDFS-10246
> URL: https://issues.apache.org/jira/browse/HDFS-10246
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: namenode
> Affects Versions: 2.2.0
> Environment: CentOS6.3 Hadoop-2.2.0
> Reporter: zhangyubiao
> Labels: bug
> Attachments: stacks.txt
>
>
> HDFS Standby NameNode dfshealth.jsp Response very slow
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
