On time situations are situations whose evaluation strategy is based on time. That means the situation triggers are evaluated and triggered at regular intervals that you defined in the business measures model, for example the situation is triggered every 30 minutes.
The on time situation triggering time is based on the Monitor Server clock time not on the clock time of the runtime engine where the process or activity instance on which the situation is modeled has been created. For example, if the process instance was created at 9:20 based on the runtime engine time, and the Monitor Server received the event of the process instance creation at 10:20 based on the Monitor Server clock, then the on time situation will be triggered and evaluated based on this Monitor Server time (10:20).
The Monitor Server evaluates if there is any on time situations that should be triggered periodically based on the value of the server property named On-time Situation Checking Interval. This property is configured through the System Properties page in the WebSphere® Business Monitor administrative console. The value of this property determines the frequency (in minutes) of checking the existence of any on time situations that should be triggered. If this interval is greater than the recurring duration for an on time situation, then the situation will effectively be triggered when the former interval has elapsed. However, if this interval was smaller than the recurring duration for the situation, then the latter will have precedence and the situation will only be evaluated when the recurring duration has elapsed. To illustrate, suppose the on time situation checking interval is 30 minutes, and a model has two on time situations defined, which have recurring durations of 10 and 50 minutes respectively.
If the system had started at 10 am, and the MC instance was created at 10:40 am, then at 11:00 the first on time situation will be evaluated and potentially fired. At 11:30, both the first and second on time situations will be evaluated.
The administrator can set the interval with the value that best suites his needs with regard to the performance. You cannot set the On-time Situation Checking Interval with a value greater than 1 day.
For example, providing that we have these values: There is an on time situation with a 30 minute recurring duration, and the first triggering time of this situation was on 10:20, the administrator set the On-time Situation Checking Interval property to 60 minutes, and the server started on 10:00, In this case, the on time situation will be evaluated is 11:00.
The triggering of an on time situation is also controlled by a gating condition that can optionally be modelled with the situation, for example an on time situation is triggered every 30 minutes if the value of a specific metric is true. Considering this with the example above, if the gating condition of the on time situation that its trigger time is at 10:50 was supposed to be evaluated to true, but since the checking interval was set to 1 hour, and for some reason, the gating condition of this situation was not true when it was evaluated at 11:00. If the Administrator sets the On-time Situation Checking Interval to 50 minutes then the situation would have been fired at 10:50.
Note that you may choose an On-time Situation Checking Interval that is greater than the recurring duration of the situation, if you are aware that the gating condition for this situation will not change frequently. Meaning that, the evaluation of the gating condition of the on time situation in the example above will not change from 10:50 to 11:00 and thus the situation will then be fired normally. This will thus boost up the Monitor Server performance to consider only those processes which have a fire time less than the server current time.
Moreover, if an on time situation has multiple recurring duration points in time at which the gating condition evaluates to true within the On-time Situation Checking Interval, then at most one firing of the situation will occur within this interval.
In Addition there may be multiple on time situations with a different recurring duration for each of them in the same process. The Monitor Server guarantees that the on time situation with the least fire time gets triggered first. This guarantees an accurate evaluation of an on time situation condition that may be affected by another on time situation.