You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-issues@hadoop.apache.org by "Harsh J (JIRA)" <ji...@apache.org> on 2012/07/10 05:49:35 UTC
[jira] [Commented] (MAPREDUCE-482) Each shuffle should proceed as a
DFA
[ https://issues.apache.org/jira/browse/MAPREDUCE-482?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13410034#comment-13410034 ]
Harsh J commented on MAPREDUCE-482:
-----------------------------------
I wonder if this makes sense to do in the ShuffleHandler today in MR2, or if it is already done. It would fit in with the event/transition model used in other places and does sound like a good idea.
> Each shuffle should proceed as a DFA
> ------------------------------------
>
> Key: MAPREDUCE-482
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-482
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Reporter: Devaraj Das
> Priority: Minor
>
> The shuffle code (for fetching map outputs) should proceed as a DFA. The states (as Owen had mentioned on HADOOP-1183) are (INITIAL, LOCATED, FETCHING, DONE, FAILED). This is to ensure correctness of map output fetching.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira