Software Quality Assurance Forums: Problems connecting SiteScope to WebLogic via JMX

Hi guys, I'm struggling with this issue below. It's about SiteScope JMX connectivity which I intend to use from LoadRunner. (I couldn't find a SiteScope forum) Hoping someone has experience with this connecting SiteScope using JMX to a *busy* domainruntime service MBean on WebLogic ...

1. VuGen 9.0, SiteScope 8.5
2. web(http/html) protocol
3. n/a
4. n/a
5. nil
6. n/a/
7. subscription
8. current
9. winXP sp2
10. service request file, no reply, see below for details

Categorize the problem
Type of case: Technical
Product: sitescope
Product version: 8.5
Sub-product: SiteScope WebLogic
Operating system: Sun Solaris
Protocol: SS Performance/Usability

Problem details
JMX Monitor not working for weblogic domainruntime MBean

Case details
Despite implementing recommended registry settings detailed at KM323003, SiteScope is unable to enumerate the WebLogic domainruntime MBean via JMX typically accessed by the URL:
service:jmx:rmi:///jndi//iiop://host:port/weblogic/management.mbeanservers/domainruntime

I believe the reason is because the SiteScope webapp itself is timing out when trying to receive such large content.

I have noticed that a corresponding XML file for all the WebLogic counters available by that MBean is not being generated in <installdir>\SiteScope\templates.applications\JMX_etc ...

I have been able to get this working using the standard runtime MBean (not domainruntime) for any of the individual managed servers, however this is not practical in a clustered environment where access is typically limited to only the admin server on port 7001.

I have confirmed IIOP connectivity and correct function of the target domainruntime MBean using JConsole. I am assuming this problem is related to the size of the target MBean, which for a busy cluster is going to contain perhaps thousands of counters in the hierarchy.

Post Extras