Displaying Different Measure Writing Performance Graphs on the System Health Page

Choose whether to display the measures received / written per period graph on the System Health page that displays actual measures written versus measures that have been received (default), or a graph that tries to predict the system's load (red/amber/green) based on a background calculation.

Note that a calculated prediction may include false assumptions based on estimates of both load and parallel processing and thus sometimes misinterprets the actual system load. This behavior can be seen for example when the graph displays the system as being overloaded, but the backlog of unwritten measures is not actually growing over time. For this reason, it is recommended to use the actual measures received / written per period graph, which gives a better picture of indicating whether or how well the Performance Manager application server is keeping up with its workload over time. This actual graph is the default setting and no action is required to display it on the System Health page.

To change the display to the original, older graph, which shows calculated estimates, proceed as follows:

Open the TMAppServerHomeConf.xml file with a text editor.
The default path for this file is C:\Program Files (x86)\Silk\Silk Performance Manager <version>/conf/appserver on the application server.
Locate the <UseSystemhealthHistory> XML tag. If it does not exist, add it manually.
Set the value to true (default) to display a graph that displays actual measures written versus measures that have been put in a backlog to be written later due to system overload. Set the value to false to display a graph that displays a predicted system load (red/amber/green). Example: <UseSystemhealthHistory>true</UseSystemhealthHistory>.
Save and close the XML file.
Refresh the System Health page to see the changes.