accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Keith Turner (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-1379) PermGen leak
Date Tue, 10 Sep 2013 22:45:52 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-1379?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13763643#comment-13763643
] 

Keith Turner commented on ACCUMULO-1379:
----------------------------------------

Reference counting would be the simplest code change.   When all instances are closed, then
the background threads can be stopped.   However this should not preclude instances from being
created in the future.

If ref counting is done, calling the instance close() method ZKI.finalize() could be tricky.
One hypothetical problem is the following situation.

 # create ZookeeperInstance (ZKI)
 # create Connector from ZKI and drops ref to ZKI
 # create Scanner from Connector and drop ref to Connector

Currently this type of code works w/ no problem.  If Scanner does not reference ZKI, then
ZKI will be garbage collected and possibly close resources the scanner needs.  I think the
scanner does ref the instance, so it would be ok in this case, was using it as an example.





                
>  PermGen leak 
> --------------
>
>                 Key: ACCUMULO-1379
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-1379
>             Project: Accumulo
>          Issue Type: Bug
>          Components: client
>    Affects Versions: 1.4.3, 1.5.0
>         Environment: Linux/JBoss
>            Reporter: Mike Giordano
>            Assignee: John Vines
>             Fix For: 1.6.0
>
>         Attachments: ACCUMULO-1379.patch, ACCUMULO-1379_v2.patch
>
>
> Under version 1.3.7 we are using the following code to initialize a cloudbase connection
during initialization of our web app:
>                         ZooKeeperInstance instance = new ZooKeeperInstance(instanceName,
zooKeepers);
>                         connector = instance.getConnector(userId, password.getBytes());
> The problem is that under the hood, this call creates several threads that are not cleaned
up when the app is undeployed in JBoss. This is occurring without performing any scans or
interacting with cloudbase in any other way. After relatively few redeploys of the app, the
PermGen Space is OOM.
> I can't find any reference in the cloudbase API akin to a close() method for the Connector
object. This is a classloader leak effecting any webapp that is accessing cloudbase directly.
The result of this leak is not simply orphaned threads, but thousands of classes not gc'd
because the classloader itself can't be gc'd. This is what is filling up PermGen.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message