[
https://issues.apache.org/jira/browse/HADOOP-4290?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12635673#action_12635673
]
Vinod K V commented on HADOOP-4290:
-----------------------------------
Changing the issue dependencies, I've found out the actual reason to be
HADOOP-4288. Because of HADOOP-4288, a tasktracker that gets a
NullPointerException in its heartbeat resends the heartbeat repetitively for
ever. And, on each heartbeat, JT finds a completed task reported by the TT to
be already marked as completed in its data-structures, so marks it as killed
and thus the state toggle happens repetitively. (JobInProgress.java +619).
It wasn't mentioned earlier, but, note that this issue surfaces only after
applying HADOOP-4053, which in turn triggers HADOOP-4288.
> In jobdetails.jsp page the values of Pending and Complete tasks for map keeps
> on interchanging with each other.
> ---------------------------------------------------------------------------------------------------------------
>
> Key: HADOOP-4290
> URL: https://issues.apache.org/jira/browse/HADOOP-4290
> Project: Hadoop Core
> Issue Type: Bug
> Components: contrib/capacity-sched
> Affects Versions: 0.19.0
> Reporter: Karam Singh
>
> In jobdetails.jsp page the values of Pending and Complete tasks for map keeps
> on interchanging with each other.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
