health monitoring nac database performance and deadlock

Document ID : KB000072574
Last Modified Date : 05/03/2018
Show Technical Document Details
Issue:
Performance problems on the NAC or Database server and/or database deadlocks. 
Environment:
CA Release Automation 4.7. 5x, and 6.x
Cause:
If the problem can be traced to be related to health_* tables in the database, or a large number of messages in the nolio_dm_all.log file related to health monitoring, then the problem could be caused by known deadlock and performance issues with the health monitoring feature. For this reason the health monitoring feature has been deprecated - starting in version CA Release Automation v6.3. In CA Release Automation 6.5 we removed the related dashboards. 
https://docops.ca.com/ca-release-automation/6-2/en/dashboard-and-reporting/manage-and-view-the-health-monitor

If the system is upgraded then the health monitoring might still be enabled and should be disabled. This can be done through the JMX console. 

 
Resolution:
Open the JMX console and navigate to this page (using the user/pass: nolio/nolio). 
http://raserver:20203/mbean?objectname=HealthMonitor%3Ainfo%3DcontrolMonitoring 

Invoke the "setEnable" option to false and verify the status with the "isEnable" option. 

If for some reason the JMX is not available then you can review the following section in the product documentation for disabling JMX. Changing the value reference in the product doc from false to true enables the JMX: https://docops.ca.com/ca-release-automation/6-5/en/installation/ca-release-automation-security/security-configuration#SecurityConfiguration-DisableJMX

If a deadlock has already occurred then a stop/start of the Release Automation management services will be needed to clear that lock. 
 
Additional Information:
The following communities post describes how to increase the interval used to collect the health monitoring data. If, for some strange reason, disabling the health monitoring feature does not help then you can try using this information to increase the interval to a point (maybe once a day or week) where it should not cause problems:
https://communities.ca.com/docs/DOC-231150006