Escalation Policies
Route alerts to different channels when incidents go unacknowledged.
An escalation policy defines a sequence of notification steps. If the first step goes unacknowledged after a set delay, OpsKitty moves to the next step and notifies a different channel or person. This ensures that critical incidents are always seen, even if the primary on-call contact misses the first notification.
Example policy
- 1
Immediately — notify the on-call engineer via Slack
- 2
After 10 minutes without acknowledgement — page the team lead via PagerDuty
- 3
After 30 minutes — notify the entire engineering team via email
Creating a policy
Go to Organization Settings → Escalation Policies and click New Policy. Add steps in order, specifying the delay before each step activates and the notification channel to use. Assign the policy to an alert rule when configuring the alert.
Acknowledgement
An incident is acknowledged when a team member opens the alert in the OpsKitty dashboard and clicks Acknowledge. Acknowledging stops escalation — no further steps fire. The alert remains in a triggered state until the endpoint recovers and the success threshold is met.