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 "Omkar Vinit Joshi (JIRA)" <ji...@apache.org> on 2013/07/31 23:07:49 UTC
[jira] [Commented] (MAPREDUCE-5436) Better exception handling for
Invalid AMRMToken exception for MapReduce
[ https://issues.apache.org/jira/browse/MAPREDUCE-5436?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13725721#comment-13725721 ]
Omkar Vinit Joshi commented on MAPREDUCE-5436:
----------------------------------------------
I liked the idea..we should definitely not loop...there is no point as far as I see in trying to communicate back.. probably we can have something like retry policy default to < 3? and trying after waiting for a while may be a second or so?.. I am not sure but probably we should check if there is some problem with RPC and token is corrupted during transmission then we get same exception or different? if same then we should definitely try again to avoid network issues..
> Better exception handling for Invalid AMRMToken exception for MapReduce
> -----------------------------------------------------------------------
>
> Key: MAPREDUCE-5436
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-5436
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Reporter: Jian He
> Assignee: Jian He
>
> Nowadays, even though MR gets an InvalidToken Exception from RM, RMCommunicator just ignores it and continues looping and retry. we should explicitly handle such exception.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira