IBM Tivoli Composite Application Manager for Application Diagnostics, Version 7.1.0.1

Monitoring overhead rises after upgrade from ITCAM for WebSphere version 6.1 Fix Pack 4 or lower

The problem: After Tivoli Enterprise Monitoring Agent is upgraded from ITCAM for WebSphere version 6.1 Fix Pack 4 or lower, the performance overhead rises significantly.

The solution: This happens because the Tivoli Enterprise Monitoring Agent automatically starts the baselining process for all applications. In the baselining process, statistical information on request response times is collected; ITCAM uses it to automatically determine response time thresholds for Application Health monitoring.

During the baselining process, Level 2 monitoring is enabled for the application, irrespective of the configured defaults and any Take Actions. This means that the monitoring overhead will also rise to values typical for Level 2 monitoring. The Tivoli Enterprise Portal may show the request monitoring level as Level 1 or Disabled, but the monitoring is still performed at Level 2.

By default, the baselining process runs for seven days. After this time, monitoring returns to the level set by the configured defaults and the Start_Request_Monitoring, Stop_Request_Monitoring and Set_Application_Monitoring Take Actions. Accordingly, the overhead will also return to expected levels.

If the increased overhead is not acceptable, you may use the Tivoli Enterprise Portal to issue the Stop_Baselining Take Action for any affected application. The baselining process will be stopped and the overhead will return to a normal level. However, if you do this, the automatic response time thresholds may not be set properly, so the Application Health "lights" may not reflect the true state of the application.

For more details on automatic threshold setting and the baselining process, see the Online Help available in the Tivoli Enterprise Portal.


Parent topic:

Errors occur during or after upgrading

+

Search Tips   |   Advanced Search