[
https://issues.apache.org/jira/browse/YARN-1438?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13829997#comment-13829997
]
Steve Loughran commented on YARN-1438:
--------------------------------------
the issue here is that while the NM logs contain the cause, the container
diagnostics sent to the AM don't include the exit code, and they don't call
Exception.toString():
{code}
LOG.warn("Exception from container-launch with container ID: "
+ containerId + " and exit code: " + exitCode , e);
logOutput(shExec.getOutput());
String diagnostics = "Exception from container-launch: \n"
+ StringUtils.stringifyException(e) + "\n" + shExec.getOutput();
container.handle(new ContainerDiagnosticsUpdateEvent(containerId,
diagnostics));
{code}
You need to get at the NM logs to find out the problem
> When a container fails, the text of the exception isn't included in the
> diagnostics
> -----------------------------------------------------------------------------------
>
> Key: YARN-1438
> URL: https://issues.apache.org/jira/browse/YARN-1438
> Project: Hadoop YARN
> Issue Type: Bug
> Components: nodemanager
> Reporter: Steve Loughran
>
> The diagnostics text generated when a container execution thrown an exception
> doesn't include the exception message -only the stack trace. This makes
> debugging harder than necessary.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
