Situations and their impact on the cost of monitoring

Situations can have an impact on the resource usage of the OMEGAMON agent (TEMA) tasks.

Referring back to an earlier post, I mentioned the notion of the more I do, the more it will likely cost. The more data I request and the more data I store and/or act on, the will result often times be a higher cost of collection, and potentially greater overhead. The more alerts, the more information I alert on, the more rows of information I potentially alert on, and the larger the number of managed systems I alert on, the result will potentially be a higher cost of alerting. This cost of alerting will often be seen in places such as the TEMA address space.

To easily see how situation processing is impacting a managed system, from the Tivoli Portal you can right click on a managed system and select 'Manage Situations' (see the example). The pop-up that you get will show what situations are distributed to the managed system, plus some other very interesting information about the situations.

There is some very interesting information that this pop-up shows, as well. One column shows the interval that the situation executes on. The tighter the interval, the more work the TEMA has to do to handle the situation. Notice also, that there are several different intervals for the situation. Many of them are running on a 30 second interval, others on 1 minute, others on different intervals. One thing to be aware of is situation optimization. If you have multiple situations referencing the same table of information, the Tivoli infrastructure has the ability to optimize the situation checks, by doing one check versus multiple. However, this will work only if the situations are on the same interval.

Another apsect of situation optimization, is that if a situation that has 'Take Action', it is not eligible for this optimization. If you have many situations with 'Take Action', this will potentially significantly reduce the potential benefit of this function. One suggestion is, if you have a component such as OMNIBus, to consider using the EIF interface, versus 'Take Action' to drive alert notification. Using the EIF option will not inhibit situation optimization.