[
https://issues.apache.org/jira/browse/HADOOP-12364?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14945995#comment-14945995
]
Siqi Li commented on HADOOP-12364:
----------------------------------
I think it significantly shrinks the window of the race, originally the window
would be 5 seconds by default. If we are not satisfied with the reduced window,
we might need to introduce synchronization of accessing pid file.
Also, can you point me the location of unit test framework for shell code.
> Deleting pid file after stop is causing the daemons to keep restarting
> -----------------------------------------------------------------------
>
> Key: HADOOP-12364
> URL: https://issues.apache.org/jira/browse/HADOOP-12364
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 2.6.0
> Reporter: Siqi Li
> Assignee: Siqi Li
> Priority: Minor
> Attachments: HADOOP-12364.v1.patch, HADOOP-12364.v2.patch
>
>
> pid files are deleting in 5 seconds after we stop the daemons. If a start
> command were executed within the 5 seconds, the pid file will be overwrite
> with a new pid. However, this pid file is going to be deleted by the former
> stop command. This is causing the monitoring service to lose track of the
> daemons, hence keep rebooting them
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
