[ https://issues.apache.org/jira/browse/YARN-1113?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13752417#comment-13752417 ]
Jason Lowe commented on YARN-1113: ---------------------------------- This is related to, and possibly just a duplicate of, YARN-257. > Job failing when one of the NM local dir got filled > --------------------------------------------------- > > Key: YARN-1113 > URL: https://issues.apache.org/jira/browse/YARN-1113 > Project: Hadoop YARN > Issue Type: Bug > Affects Versions: 2.0.1-alpha > Reporter: Nishan Shetty > > 1.In NodeManager only one disk is configured for NM local dir > 2.Make that disk full > 3.Run job > Problems > -Tasks assigned to that disk filled NM is waiting for container expiry > time(10min) > -After expiry time that containers will be killed and new task attempt spawned > -All the other tasks attempts are getting assigned to same node only and > failing the tasks 4 times intern job fails -- 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