Tracking Siebel Object manger tasks - number of users in Siebel
Siebel deployments use Object Manager components to render the UI. There is generally one task per user in Siebel, or more accurately per open browser session. To track the number of Siebel users is a key statistic, it is the most basic baseline you should have for your deployment, to provide a peak useage information and correlate memory use and CPU use with the number of users.
The code below gives you a method to track the number of tasks per Siebel Component. It can be applied to any Siebel Component, but in this case we are interested in Object Managers that support users. Each deployment of Siebel is different, and may use different Object Managers, depending on the vertical. For example, a call center may use eCommunicationsObjMgr_enu, while a Finance company may use FINSObjMgr_enu. Many deployments customize their own version of object managers. In anycase, Object Managers tend to be the Siebel Components with the most number of tasks running.
Once you determine what Object manager you are using, you need to determine what the "Alias" for it is. Since you have VA2 running that is easy, just look at the "Monitored Components" folder and the Component name and Alias will be listed.
The Siebel method for tracking the number of users is quite simple. It is simply calling the "numoftasksforappsrvrcomp" and passing in the correct parameters. You will have to customize the $enterprise and $appserver parameters to your environment to use this Satistic.
my $enterprise = 'siebel';
my $appserver = 'compaqamd';
my $compalias = 'eCommunicationsObjMgr_enu';
my $entobj = siebsrvobj->newobj($datasession,$debug,$enterprise); #initialize the enterprise object
$retval = $entobj->numoftasksforappsrvrcomp($appserver,$compalias); #get number of running tasks for Component
The output will return the current number of Tasks running for a Siebel component. Often deployment have mulitple Siebel Servers, and the same method "numoftasksforappsrvrcomp" should be executed for multiple application servers and added together. Below is the example of the output for a rather large Siebel deployment, which shows the useage pattern, of when users are logging on and off the system. There are around 4000 Siebel users on this system at various peak times.