Error getting current time from controller java.lang.OutOfMemoryError: PermGen space

[CommunityUser]

Hi all,

We have an application (stand alone under Solaris) and when we start the JVM together with the monitoring it crashes after some time. We enabled the trace logging in the agent and we found that it is out of memory exception due to PermGen space

[AD Thread Pool-Global1] 11 maj 2015 15:53:45,575 ERROR ControllerTimeSkewHandler - Error getting current time from controller
java.lang.OutOfMemoryError: PermGen space

The thing is that this happens only if the agent is attached to the process and in addition we already have increased the PermGen to 1024m which is quite a lot.

I hope you can provide us with some idea why this happens and also the correct memory settings for the agent.

Thank you in advance.

Regards,

Ned

[CommunityUser]

I tried to investigate a bit more and I belive it is related to the increased amount of PermGen space for the JVM when the agent is attched. The default value for our application is 80 MB and it seem the JVM creshes just few minutes after it is started with the agent

[Arun_Dasetty]

Hi Ned,

The provided agent logs are not complete logs, we request you to check the behavior with bit higher values for Xms and Xmx and if issue persists we request you to send logs archive at path <agent_install_dir>/logs the logs requested will help us to assist you better as we could see interceptors applied the heap jvm args allocated and snapshot or other limits hit in logs.

Keep us posted the logs if issue persists after suggested changes. We see 80 MB is very less memory considering it is some app server like tomcat and addition agent usage could have just exposed the low memory allocated to jvm, hope that clarifies.

Regards,

Arun