Why is Jira Automation's logged execution time longer than the actual run time?

[Abhilash Reddy Sane]

I'm encountering an issue where the reported execution time in the Jira Automation rule logs is significantly longer than the actual time the rule seems to take when observed manually. For instance, a rule that runs in a few seconds appears to log an execution time of over 20secs.

Has anyone else experienced this discrepancy, and are there best practices for accurately measuring automation performance?

I’m trying to update the priority scores for Jira tickets using an automation rule that follows these steps:

1. Check the total backlog count – the rule first retrieves the total number of tickets in the backlog.

2. Check how many tickets meet specific criteria – it filters for a subset of tickets based on certain conditions (e.g., label, status, etc.).

   1. For each ticket, it evaluates whether the total backlog count is greater than 30.

      If the backlog exceeds 30, the rule uses JQL to identify the ticket and updates its priority score with a boost.

   2. If the backlog is 30 or less, the rule still updates the priority score using the same JQL, but without the boost.

The goal is to dynamically adjust ticket priority based on current workload and specific qualifiers.

[Bill Sheboy · Accepted solution]

Hi @Abhilash Reddy Sane -- Welcome to the Atlassian Community!

The symptoms you describe indicate the rule is encountering one (or more) of the service limits:

https://support.atlassian.com/cloud-automation/docs/automation-service-limits/

For the community to offer more guidance, and possible alternatives, please post the following to provide context:

• what problem is the rule trying to solve; that is, "why do this?"
• what version of Jira are you using: Cloud, Server, or Data Center
• an image of your complete rule in a single image
• an image of the audit log details showing the rule execution
• what is not working as expected, and why you believe that to be the case
  • specifically, for the log you provide, what would you expect to happen?

Kind regards,
Bill

[Walter Buggenhout · Accepted solution]

Hi @Abhilash Reddy Sane and welcome to the Community!

I would rather rely on the reported run time of an automation rule than on what you manually observe.

Automation rules run asynchronously in Jira an may depend on background processes to complete in a consistent state. Not everything that is executed by the rule is also visible in the UI. The Atlassian Platform does consist of a lot of interconnected services that do require background syncing of information that you don't (always) see as a user. I would expect that the automation rule logs log the end of executions as the very last step in the overall process.

Hope this helps!

[Abhilash Reddy Sane]

Thank you very much @Walter Buggenhout , this was very helpful!

However the problem is when i run the same for around 60-80 tickets, it takes around 300seconds, sometime more. This causes the automation to finish running but Jira automatically turns it off, which causes the rule to fail and having to reenable it again manually for every run until the count reduces. I have tried simplifying the rule however, the problem persists.