There are instances that a particular event detector may trigger a policy multiple times within a short time-span. An example of this is if we are using the Interface ED to trigger a policy when the rxload passes a certain threshold. If we are polling every 10 seconds and the bandwidth spike is sustained for a minute then that would cause the policy to run 6 times within 60 seconds. In certain situations we would rather simply have the policy run once during the initial spike and not run again unless a certain length of time has passed. Is there a way we can configure a policy to run only a maximum of X number of times throughout a given time range?
By default, there is only one thread allocated to the default scheduling queue. Therefore, only one policy will run concurrently. It depends on your version of IOS as to whether or not this can be changed.
Joe, thanks for such a quick response. We are running 12.4(24)T1 on most of our equipment. I guess my question is somewhat two-fold: 1)does eem queue policies to run if another policy is currently running, and 2)is there a way that we can limit this activity by timer
Using the example that I mentioned let's say that an event detector triggers a policy to run and then 10 seconds later the same event detector triggers the same policy to run again. We would prefer this policy to be limited to only running 1 time within a 60 second time frame.