[ https://issues.apache.org/jira/browse/IGNITE-2354?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15232338#comment-15232338 ]
Ivan Veselovsky commented on IGNITE-2354: ----------------------------------------- Pull request: https://github.com/apache/ignite/pull/622 . New test: org.apache.ignite.internal.processors.hadoop.HadoopMapReduceErrorResilienceTest . > Ensure that Hadoop client is still operational after failed map-reduce jobs. > ---------------------------------------------------------------------------- > > Key: IGNITE-2354 > URL: https://issues.apache.org/jira/browse/IGNITE-2354 > Project: Ignite > Issue Type: Bug > Components: hadoop > Affects Versions: ignite-1.4 > Reporter: Vladimir Ozerov > Assignee: Ivan Veselovsky > Priority: Critical > Labels: important > Fix For: 1.6 > > > We need to ensure that MR failures on the server doesn't block further job > submissions. > First of all we should cover such situations with tests. > 1) Test exceptions on all job stages (setup, map, reduce, combine, cleanup) > 2) Test several exception types: > - Exceptions expected from Hadoop interface > - Exceptions unexpected form Hadoop interface > - Errors. -- This message was sent by Atlassian JIRA (v6.3.4#6332)