[
https://issues.apache.org/jira/browse/SAMZA-414?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14137546#comment-14137546
]
Yan Fang commented on SAMZA-414:
--------------------------------
Yes, I am using the root /tmp. ./tmp will be wiped... ><
{quote}
The problem that I have with putting the dump in /tmp is that /tmp is a shared
resource, and is usually not monitored all that well. The dumps can easily get
to be 500+ megs. This could very easily fill /tmp space, which will break a lot
of things.
{quote}
Yes, this is true. Three points:
* Putting the dump to ./tmp or $SAMZA_LOG_DIR causes problems. Seems so far the
/tmp is the only option we have.
* Some tools, such as jvisualvm shipped with Java, will dump the heapDump to
the /tmp anyway when the users debug. I was thinking it was fine that we dump
to the /tmp as well.
* But the problem with Samza is that, it is possible that two or more
containers will run in the same machine, as a result, that machine will have
more than one dumps. That will definitely fill the /tmp space.
Maybe we should give a caveat to the users and make this optional ?
> Enable HeapDumpOnOutOfMemoryError by default
> --------------------------------------------
>
> Key: SAMZA-414
> URL: https://issues.apache.org/jira/browse/SAMZA-414
> Project: Samza
> Issue Type: Bug
> Components: container
> Affects Versions: 0.8.0
> Reporter: Chris Riccomini
> Assignee: Chris Riccomini
> Fix For: 0.8.0
>
> Attachments: SAMZA-414-0.patch, SAMZA-414-1.patch, log-hprof.png
>
>
> It would be nice if Samza's run-class.sh defaulted to use
> -XX:+HeapDumpOnOutOfMemoryError. According to
> [this|http://stackoverflow.com/questions/542979/using-heapdumponoutofmemoryerror-parameter-for-heap-dump-for-jboss]
> post, it puts the heap dump in CWD by default, which should be fine.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
