What does the evaluation interval on an activity set actually do?

Kevin Bumber
Tera Contributor

I've noticed that there is a field called 'Evaluation Interval' on activity sets which is set to 4 hours by default. What the heck does this mean? I know there's sometimes a minute or so delay between when a parent activity set is closed and a dependent child activity set is spawned based on it's trigger condition of the parent set being closed, but I've no idea where this 4 hour interval comes into play or why it matters. Also, so far as I can tell when the LE activity set engine / workflows come into play there's a 30 second or so timer before it makes another check on whether or not an activity should fire.

So...

Why is there an 'Evaluation Interval' option on an activity set if the evaluation timing is already being handled in the workflow?

If I change an activity set's evaluation interval to 1 minute what happens?

If I change it to 48 hours what happens?

I'm sure there's a good reason for it, but I can't figure out what that might be based off of the documentation noted below.

https://docs.servicenow.com/bundle/london-hr-service-delivery/page/product/human-resources/task/Mana...

Note: The Evaluation interval field is not revealed by default. This field works with the Max activity countfield on the Activities tab under Workflow Properties. See Workflow properties . If you reveal the Evaluation interval field with the intention of changing the default value (four hours), use caution. Frequent updates mean that more events fire and could cause your lifecycle event activity sets to cancel before the Lifecycle Event completes. For workflows associated with Lifecycle Events, the value of the Max activity count field has been increased to accommodate long running business processes that contain multiple workflow activities.

6 REPLIES 6

michaelj_sherid
ServiceNow Employee
ServiceNow Employee

Hi Kevin,

This interval is a catch-all to reprocess (or process initially, depending) any outstanding activities. This polling uses the resources identified in the Iteration control settings (Max Flows and Max Simultaneous) These are parameters that that control how many sub-flows can run at a given time (and overall). Moving them to one minute is not recommended. I will try to explain:

This Max simultaneous property should be set to the max number of Activity Sets that need to be running at one time (Awaiting Trigger, Launched or Finishing up) for the largest Lifecycle Event a customer has.  This may be tricky to calculate, depending on the customer’s setup:

 

Example 1:

 Set A > Set B > Set C >Set D

 Here, Max flows = 4 and Max simultaneous = 2 (1 Set Finishing, 1 Set Triggering)

 

Example 2:

                                 Sets E

                Set B  > Sets F

Set A > Set C

                Set D

 Here, Max Flows = 6, Max Simultaneous = 5 (Set B could be finishing, triggering E and F, Set C and D could be running).

 With complicated flows using date triggers or multiple simultaneous paths, setting the Max Flows the same as Max Simultaneous will guarantee no issues as long as long as your total number for any Lifecycle Event does not exceed that number.

These parameters were increased back in Kingston. Having these parameters too low (i.e. 5 and 10 with polling at 1 minute), will result in what you see where the activities will be cancelled. 

 

Regards,

Mike

Where are the iteration control settings located?

These can be found in the HR Activity Set Launcher Workflow

michaelj_sherid_0-1665600860458.png

 

Hi @michaelj.sheridan ,

is it somehow supposed to be deterministic, when the 'catch-all' is happening / required and when the next AS is triggered right away? I have a bit of a weird situation that in the dev instance it's 'instantaneous', but in the higher instance it takes the 4 hours... :-S

Thanks,

Christian