[
https://issues.apache.org/jira/browse/CASSANDRA-6541?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13861662#comment-13861662
]
Jason Brown edited comment on CASSANDRA-6541 at 1/3/14 4:53 PM:
----------------------------------------------------------------
Yeah, a brute force check of some clusters confirms some are on 1.6.0_41, _37,
and _45 - and, of course, they are mixed within the cluster, depending on when
the instance was launched (if it was with an older AMI with older JVM).
I haven't narrowed the problem down to which instances had troubles with which
JVM, but some clusters with < u45 did not have any problems, so I suspect it's
at least > u41 (our most highest version), but probably [~benedict] is correct
with u45.
was (Author: jasobrown):
Yeah, a brute force check of some clusters confirms some are on 1.6.0_41, _37,
and _45 - and, of course, they are mixed within the cluster, depending on when
the instance was launched (if it was with an older AMI with older JVM).
> Add JVM_OPTS="$JVM_OPTS -XX:+CMSClassUnloadingEnabled" to cassandra-env.sh
> --------------------------------------------------------------------------
>
> Key: CASSANDRA-6541
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6541
> Project: Cassandra
> Issue Type: Bug
> Components: Config
> Reporter: jonathan lacefield
> Priority: Minor
>
> Newer versions of Oracle's Hotspot JVM , post 6u45, are experiencing issues
> with GC and JMX where heap slowly fills up overtime until OOM or a full GC
> event occurs, specifically when CMS is leveraged. Running repair exacerbates
> this issue. The configuration added to the Summary line helps alleviate this
> behavior and should be included in the C* config files by default.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
