I believe I've seen that error when the agent and server are out of sync with what each thinks is in inventory.

Did you start the agent with --cleanconfig?

On 07/22/2010 08:50 AM, Alexey Kamenchuk wrote:

Hi Heiko

Yes I use 300

Ive restarted the agent but it didnt help - still same error Cannot say for sure if there was an OOME before, unfortunately logs for previous days have been deleted

Inventory - 2 JBoss servers -xmx256MB All the EJB3 and Hibernate stats metrics are turned off (I have 300+ beans and 200+ hiber entities and dont need those fine-grained metrics yet)

RHQ Agent Measurement subsystem summary: Recent Measurements Currently Schedule Measurements: 572 Failed Collections per Minute: 0 Late Collections per Minute: 0 Measurements Collected per Minute: 52.8 Total Time Collecting Measurements per Minute: 314.8ms

Regards Alexey

On 22.07.2010 15:01, Heiko W.Rupp wrote:

...

Hi,

this is very strange.

Am 22.07.2010 um 09:10 schrieb Alexey Kamenchuk:

...

For one the monitored JBoss servers, Im not able to monitor any of its child resources (cache, tomcat, ejb's etc). Agent logs reveal exception shown below (there are multiple of them in logs, one per child resource) Caused by: java.lang.IllegalStateException: No plugin class loader was specified for org.rhq.core.pc.inventory.ResourceContainer$ComponentInvocationThread@848170f.

Could this be caused by Out of Memory exceptions that have shown up earlier? Do you have large inventory on that specific agent? I guess that you are using plain RHQ 300 ?

Heiko

---

rhq-users mailing list rhq-users@lists.fedorahosted.org https://fedorahosted.org/mailman/listinfo/rhq-users