SLA tracking and alerts
writes gatedWatches how long a ticket has gone without an update and raises an alarm before the promise is broken, rather than after.
Jira Service ManagementSlack
State what it is doing right now
| Running | … |
| Environment | … |
| Cadence | … |
| Last ran | never |
Scope
What sets it off
Every 15 minutes, independently of the pillar polls — the only automation on the short tick.
Reads
- HELP tickets
- SLA policies
- Service holidays
Writes
- A Slack message, when sending is enabled
- A row in sla_notifications on every alert it computes, sent or not
- The clock on case_sla, recording that this threshold has been announced
Limited to
- Business hours and holidays are honoured by the P1, P2 and P3 policies. P0 runs on a round-the-clock calendar and can breach at 3am on a Sunday, deliberately
- Warns once per threshold rather than on every tick, and never drops back to a warning once it has breached
- Clocks pause on the statuses that are waiting on somebody else — Pending Validation, Customer UAT, Waiting on Customer, Resolved and Cancelled
- Closed tickets are never chased
Will not do yet
- Slack sending is off until the recorded alerts have been reviewed
- The switch and cadence on this page do nothing. Unlike the other four, this automation does not read its operator row — it runs every 15 minutes regardless, and only the Slack flag stops it speaking
- No Run now, no dry run, and no run history — the buttons will refuse
- Alerts computed while Slack is off are recorded but never delivered, and they still count as announced. Turning Slack on will not replay them
- Holidays are loaded for 2026 and 2027 only; after that the calendar knows about weekends but not holidays
Actions
This shows which clocks are due and what would be sent to Slack. Nothing is sent. Run now writes, so it stays disabled until the automation is enabled and the environment allows it.