[
https://issues.apache.org/jira/browse/HADOOP-1610?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12514255
]
Doug Cutting commented on HADOOP-1610:
--------------------------------------
The names of these metrics ("tasks_unresponsive" and "failedToPingTracker") are
neither consistent nor do they indicate that these are task failure counts.
Perhaps they ought to be called something like "tasks_failed_timeout" and
"tasks_failed_ping"?
> Add metrics for failed tasks
> ----------------------------
>
> Key: HADOOP-1610
> URL: https://issues.apache.org/jira/browse/HADOOP-1610
> Project: Hadoop
> Issue Type: Improvement
> Components: mapred
> Affects Versions: 0.14.0
> Reporter: Devaraj Das
> Assignee: Devaraj Das
> Fix For: 0.15.0
>
> Attachments: 1610.1.patch, 1610.patch
>
>
> Add metrics for failed tasks - specifically tasks that kill themselves due to
> communication failure with the tasktracker (exit status 65), and those that
> are killed due to lack of progress reporting.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
