Before
Analysts had to review open occurrences, compare them with incidents already marked as notified, group related occurrences, and decide which incidents required owner notification. They then had to start the notification workflow incident by incident and compile a status summary.
How it works
- Read open occurrences from the configured Jira Cloud and Confluence Cloud source types.
- Group occurrences by incident, exclude incidents already marked as notified, and limit the work queue to 20 incidents per run.
- Send each remaining incident to the owner-notification workflow and return a consolidated run summary.
Workflow details
- Trigger and scope: An on-demand asynchronous webhook starts the run and applies the configured scope of Jira Cloud and Confluence Cloud occurrences. In the demonstration, external connector responses are locally replayed.
- Open occurrence collection: The workflow reads occurrences for each supported source type and keeps those linked to incidents with an open status of TRIGGERED or ASSIGNED. Related occurrences are grouped by incident so each incident is handled once.
- Notification selection: Incident IDs are compared with incidents already marked as notified. Previously notified incidents are excluded, while up to 20 remaining incidents are selected for processing; any excess work is deferred to a later run.
- Routing and results: Each selected incident, together with its occurrences, is passed to the owner-notification workflow. The automation collects each outcome and returns counts for open incidents, already-notified incidents, ignored source occurrences, deferred work, and notification results.
What changed
The workflow automatically filters and groups eligible incidents before notification, sends only unnotified work to the owner-notification workflow, and makes deferred or ignored items visible in the final summary.
Work removed
Reviewing open occurrences one by one, comparing incident IDs with notified tags, grouping occurrences by incident, selecting eligible incidents, initiating owner notifications individually, and compiling notification results.
Expected outcome
Up to 20 newly qualifying incidents can be selected and routed in one run, while already-notified incidents are excluded and excess work is explicitly deferred.
