On-call Troubleshooting

Start with the outcome Moneat recorded. Alert, route, grouping, paging, delivery, and incident results are separate signals; one successful step does not prove the next one happened.

SymptomCheck
Slack is installed but actions failOpen Slack integration settings, confirm the installation is healthy, required scopes are present, public callback URLs are reachable over HTTPS, and the signing secret matches.
An external alert never appearsCheck the source's Events tab, ingress token, payload mapping, and delivery error. A source can be configured correctly without having received its first event.
An alert appears but nobody is pagedInspect the matched Alert Route, its order and enabled state, the escalation policy, schedule coverage, notification policy, and delivery result. No matching route means route paging did not run.
An alert group has no incidentConfirm that the winning route enables incident creation. Alert-only groups are valid and do not open incidents; suggested grouping also needs responder confirmation.
Native incident controls are unavailableCheck the capabilities response and the organization's plan or self-hosted license. Native incident response is entitlement-based.
Coverage shows a gap but no one was notifiedCoverage reads do not notify. Review the coverage policy, then use Re-evaluate coverage when you intentionally want policy owners paged.
An incident call cannot be createdCheck that an administrator connected a Google Meet or Zoom host account and selected it as the default. If no managed account is available, attach an existing provider link.
A save returns 409Refresh the current resource and retry with its latest version or revision. Version-checked writes reject stale edits instead of overwriting newer work.

Route and delivery checks

When an alert is missing an expected action, inspect each result independently:

  1. Route — did an enabled route match, and which revision won?
  2. Group — was the alert grouped, suggested, automatic, or left alert-only?
  3. Paging — did the winning route select a schedule, user, team, or escalation policy?
  4. Delivery — did the provider accept and deliver the selected channel?
  5. Incident — was creation enabled, skipped, or blocked by the route outcome?

Use the detailed route outcome table in Routing & escalation and the source delivery history in Alerts & alert sources.

Configuration conflicts

If a save returns a conflict, refresh the current resource and retry against its latest version or revision. Version-checked writes reject stale edits rather than overwriting a newer response configuration.