[
https://issues.apache.org/jira/browse/MAPREDUCE-2928?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13105125#comment-13105125
]
Eli Collins commented on MAPREDUCE-2928:
----------------------------------------
Owen, Ravi - thank you for the feedback.
Will re-introduce getGoodLocalDirsString.
Wrt #1 are we worried about this case? It doesn't seem like something that
would happen in practice as disk failures aren't typically recoverable, and per
MAPREDUCE-3011 I'm thinking once a local path is identified as bad we should
remove it from the config. In the DN we maintain a count of total failures, how
about we do that here and use it instead of the # of valid dirs? This value
always increases and is therefore not susceptible to the bug you described.
Wrt #2 - good point, will fix this.
> MR-2413 improvements
> --------------------
>
> Key: MAPREDUCE-2928
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-2928
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: tasktracker
> Reporter: Eli Collins
> Assignee: Eli Collins
> Fix For: 0.20.205.0
>
> Attachments: mapreduce-2928-1.patch
>
>
> Tracks improvements to MR-2413. See [this
> comment|https://issues.apache.org/jira/browse/MAPREDUCE-2413?focusedCommentId=13095073&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13095073].
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
