eG Help

Monitoring the clients that are communicating with the Tuxedo application server domain and their experience with the domain, will accurately point administrators to clients who are imposing the maximum load on the domain and the clients whose experience with the domain is below-par. This is exactly what the PepAppClientTest does.

This test automatically discovers the clients who are currently connected to the server domain, reports the number of active sessions for each client, and also reveals the number of transactions of every client that have been committed and aborted. In the process, the test pinpoints those clients who have launched too many sessions on the domain; in the event that the server domain chokes, such clients could be considered guilty of overloading the domain. Additionally, the test reveals which client’s transactions were aborted often, rendering such clients unable to complete their transactions; this in turn may disrupt the business flow and adversely impact experience with the server domain.

Measurement	Description	Measurement Unit	Interpretation
No_of_clients	Indicates the number of sessions currently active for this client.	Number	This is a good indicator of the workload currently imposed by a client on the application server domain. In the event of an overload, you can compare the value of this measure across clients to know which client is contributing to the overload by launching many sessions on the domain. You can also use the detailed diagnosis of this measure to know which users are communicating with the application server domain via the client, the duration of each user's session with the domain, and the current status of the session. With the help of this information, administrators can in no time isolate those sessions that are idle, but are unnecessarily consuming resources and adding to the server load. By closing such sessions, administrators can conserve server resources and reduce load.
Avg_begins	Indicates the average number of transactions started by this client.	Number
Avg_commits	Indicates the average number of transactions committed by this client.	Number
Avg_abort	Indicates the average number of transactions of this client that were aborted.	Number	Ideally, the value of this measure should be low. A high value could indicate the incidence of many transactional errors that are causing the transactions to abort. You may want to compare the value of this measure across clients to know which client’s transactions aborted the most, thus leading administrators to those clients who may not have a had a good experience with the server.