[
https://issues.apache.org/jira/browse/HADOOP-5152?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12711467#action_12711467
]
haiyang commented on HADOOP-5152:
---------------------------------
i also met this situation.but i can't explain what happened to the running task.
> better messaging when tasktacker cannot access userlogs
> -------------------------------------------------------
>
> Key: HADOOP-5152
> URL: https://issues.apache.org/jira/browse/HADOOP-5152
> Project: Hadoop Core
> Issue Type: Improvement
> Components: io
> Affects Versions: 0.18.2
> Environment: 15 node cluster
> CentOS 5
> 2.6.18-8.el5 #1 SMP
> java version "1.6.0_10-ea"
> Reporter: David O'Dell
> Priority: Minor
>
> When the tasktracker cannot access the userlogs directory all the tasks fail
> but the only message in the logs is:
> 2009-01-28 16:00:37,024 WARN org.apache.hadoop.mapred.TaskRunner:
> attempt_200901280756_0001_m_000534_1 Child Error
> java.io.IOException: Task process exit with nonzero status of 1.
> at org.apache.hadoop.mapred.TaskRunner.runChild(TaskRunner.java:462)
> at org.apache.hadoop.mapred.TaskRunner.run(TaskRunner.java:403)
> It should write to the log what the actual problem is.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
