Get Next Work Capability - Design/Performance Considerations


We would like to use Get Next Work capability of the platform.

We have seen that the system is working fine with relatively low active users and tasks on the workbasket.

Our initial impression is that If we add new users and some more tasks on the workbasket, the general performance declines significantly.

We haven't created and SR for that (yet) as we may miss some design considerations for such high load scenarios.

Our target architecture will have 500 active users on a workbasket with 5-7k task.

Do you have recommendations for such high load scenarios?

Thank you

***Edited by Moderator Marissa to update platform capability tags****


Keep up to date on this post and subscribe to comments

September 20, 2019 - 12:23pm

Things to consider:
- Is the report definition that feeds the workbasket designed in a way that performs well:

  1. The report is limited in how many records are being returned.
  2. Filter criteria is indexed.
  3. No properties that are in the report definition are unoptimized for reporting.

Also, look at the acutal areas where performance degradation using performance tools available (Such as gathering PAL data, using the PDC, reviewing the alert logfile) what's slowing it down?  Is it a database call, an action from the user, etc?