Automation allowance exceeded

Question

Our Automation limit in Jira has exceeded & all automation is failing.  I need to know which CMB requests require manual intervention.  The audit log shows an issue number which I cannot associate with a CMB submission #

Gabriela - LeanZero · Answer

@Leanne Asanuma, a fair correction from Sami, and I should have been clearer. Atlassian's usage page dates the extra usage billing to December 3, 2026, but it doesn't say whether the pause at 100% of the new organization-wide allowance applies before then, so I'd start where Sami points: Jira Settings > System > Global Automation > Usage, which shows your site's current limit and reset date. If that's where you've run out, that's the number to take to whoever owns the plan, and Platform usage in Atlassian Administration is the view to keep an eye on as the step allowances come in.

Sami Shaik · Answer

@Leanne Asanuma ,

@Trudy P Claspill has the recovery search right, and one correction to the second reply matters for where you go to fix it, because there are two meters and only one of them can be stopping your rules today.

Which limit you actually hit. The organisation-wide step allowance that @Gabriela - LeanZero describes (Atlassian Administration → Insights → Platform usage) is metering now but does not stop or charge anything until 3 December 2026; Atlassian's usage page says the allowances and the pause-at-100-percent behaviour take effect from that date. If rules are failing on 29 September, it is the older per-site rule-execution limit in Jira itself: Settings → System → Global Automation → Usage. That is where the count, the cap and the reset date for your site are shown, and that is the number to send to whoever owns the plan. Two ways out today: reduce what counts (the limit counts runs of multi-project and global rules; single-project rules are unlimited on that model, so rescoping rules that only need one project takes them out of the count), or upgrade the plan tier the cap comes from. Also open the audit log and read the exact error text: "monthly rule executions have been exceeded" is this meter; a THROTTLED entry is a service limit, which is a third thing and resets on its own.

Nothing replays. Executions that failed while over the limit are gone; automation does not queue them and does not re-run them after the reset. So the CMB requests Trudy's searches find are the ones you must re-drive by hand, or better, by giving the rules a way to catch up:

* Add the CMB submission number as a column in your search (it is a custom field on the work item, so Issue Navigator → Columns, or use it directly in JQL: "CMB Submission #" is not EMPTY AND created >= "2026-09-28"). That maps the issue keys in the audit log to your submission numbers without opening each one.
* For each rule that should have run, decide whether re-triggering is safe. Rules on Work item created cannot be re-triggered; give that rule a second trigger, Manual trigger, and run it from the affected work items once the limit is lifted. Rules on a field change or transition can be re-driven by a bulk edit of that field or a bulk transition, but only if the actions are idempotent (a second run must not send a second email or create a second child).
* Before you re-drive anything, confirm the limit is actually lifted, or the catch-up runs will fail too and add to the confusion.

What to change after. Put a Log action as the first step of the rules that matter, so the audit log carries the CMB number in its own text next time and this mapping takes minutes rather than a morning.

Gabriela - LeanZero · Answer

Hi Leanne, Trudy's steps will find the requests you missed, so I'll only add the part about getting automation running again, because the limits changed recently and the fix may sit with someone other than a Jira admin. Atlassian now meters automation as steps pooled across your whole organization, and its usage doc is clear that once extra usage is disabled and you reach 100%, "your automation flows will stop running until usage resets on your usage billing date".

That's an org setting. So I'd ask your organization or billing admin to look at Atlassian Administration > Insights > Platform usage today. They can enable extra usage so flows keep running, raise the limit they've set, or buy a usage pack, and any of those should bring the rules back without waiting for the reset. I'd also check which flows used the most steps this month, because a single busy rule with a branch can drain a lot of steps on its own. Worth a look.