eG Monitoring
 

Measures reported by SiebelStatsTest

Components refer to the various tasks or programs that run on the Siebel server and perform the work requested by the user. For example, the object manager is one of the key components on a Siebel server. In order to effectively measure the end-user experience with a Siebel server, it is essential to keenly observe and analyze the fluctuations in resource usage, responsiveness, and errors encountered by these components. The SiebelStatsTest, executed by an internal agent, enables such an analysis. In the event of any deterioration in the performance of a Siebel server, the metrics reported by this test will enable administrators to figure out whether there are any resource-intensive/error-prone components on the Siebel server, which are impacting its performance.

Measurement Description Measurement Unit Interpretation
CPUTime The total CPU time for component tasks Secs Ideally, the value of this measure should be low. A very high value could indicate that users are executing one/more CPU-intensive tasks on the Siebel server. Further investigation is hence required to zero-in on the resource-hungry components.
Elapsed_time The total elapsed (running) time for component tasks Secs This measure is indicative of time taken by the task to complete its operation.
Think_time The average end-user think time between requests. Secs  
Total_resp_time The total amount of time taken by the components to respond to requests Secs A very high value indicates slow component responsiveness. Response time issues can be caused by high resource utilization or heavy load on the components.
Total_tasks The total number of tasks completed for the server components Number  
Avg_resp_time The average time taken by the components to respond to requests Secs A very high value indicates that the component responds slowly to requests. Response time issues can be caused by high CPU utilization or heavy load on the components.
Avg_conn_time The average connect time for component sessions Secs Ideally, a low value is desired. A high value indicates connection bottlenecks.
Errors The component job ran but encountered an error during operation Number Ideally, this value should be 0.
Tests_attempted The number of test attempted Number  
Tests_failed The number of tests failed Number  
Tests_successful The number of test successful Number