Disk problems encountered by a running task are not correctly reported to the
user
----------------------------------------------------------------------------------
Key: HADOOP-1323
URL: https://issues.apache.org/jira/browse/HADOOP-1323
Project: Hadoop
Issue Type: Improvement
Components: mapred
Reporter: Devaraj Das
Currently, if a running task encounters a disk problem, it report FSError to
the TaskTracker. The TaskTracker then reinitializes itself and the state of the
running tasks are lost. The JobTracker stops getting status reports about those
tasks from that TaskTracker, and fails the tasks with a "Lost TaskTracker"
diagnostic message. This error reporting should be improved to something like
"The node running the task encountered disk problems".
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
