In Automation Rule, when I get a failure due to a LOOP, how can I see the issue that initiated it?

Question

I have an automation rule in Jira Cloud that occasionally fails.  When I look at the audit log, it says "LOOP", but it doesn't tell me which issue initiated the rule.

I've gone over the rule and related/impacted rules closely, but can't determine why it is giving me the loop error on occasion (most of the time, there's no problem).  I need to find the issue that is triggering the rule that gives the loop error, but it doesn't show it in the Audit log.

How can I find that issue?

Bill Sheboy · Answer

Without seeing all the rules and audit logs involved, I suspect you will need to perform a bit of detailed analysis:

* grab some paper or a tool like Miro or Confluence to create a mind map diagram
* find the most recent LOOP error in the global audit log
* inventory all rules than triggered within 5 minutes before that
* working outwards, from the rule closest to the LOOP error, review the audit log for each triggering* identify which rules triggered based on that rule's actions, or candidate rules that should have triggered

* repeat back to the 5 minute mark
* now pause and review the results

Or, submit a ticket to Atlassian Support to take a look at this scenario.  I would be startled to learn they do not have much better logging than we can see for rule execution.  (Apparently, they can detect the LOOP condition somehow.)

And, if they disclosed what "quick succession" meant, perhaps we could even find potential LOOP causes proactively.

Cash Coyne_tanpr0 · Answer

The rule is used quite often on a daily basis, but occasionally it will have the loop error.  I'm just trying to determine what sequence causes that, but it's very difficult without a solid example.  I know this rule kicks off a succession of other rules, but I didn't think it would be more than 5 based on all the scenarios I tried, so not sure which of the rules is causing the loop error.

Bill Sheboy · Answer

Hi @Cash Coyne

My understanding of that error is it's related to the service limit, where a rule cannot trigger itself or other rules more than 10 times in "quick succession": https://support.atlassian.com/cloud-automation/docs/automation-service-limits/

I have found no documentation on what "quick succession" means as a specific time frame.

I recommend checking your global automation log and the automation performance insights to learn if this specific rule has been having this problem for a while.

Kind regards,
Bill