Test an escalation policy

Before you begin

Use a test service/window, tell responders, attach the policy, verify schedule/team targets and endpoints, and decide whether the test should acknowledge first step or intentionally advance.

Open the feature

Open the policy and attached test service, then record current targets/timing.

Configure the test

Create a unique test incident title, expected recipient/channel/timestamps, acknowledgement actor, and cleanup plan. Avoid unintended production paging.

Complete the action

1. Trigger the controlled incident.
2. Observe policy/target resolution and first delivery.
3. Acknowledge or intentionally let the next step advance.
4. Inspect delivery history/timeline.
5. Resolve the test incident.

What OpsKnight does

OpsKnight evaluates current open state, policy generation, runtime target eligibility, scheduler timing, queue/provider delivery, and acknowledgement/terminal suppression.

Verify it worked

Compare expected and actual recipients/channels/timestamps, delivery results, timeline, acknowledgement actor, later-step behavior, and final resolution. Retain objective evidence.

Change or undo the test

Resolve and clearly label the test, remove temporary service/policy/override changes, and notify accidentally paged recipients. Preserve audit/delivery evidence.

Troubleshooting

No delivery: trace service attachment → policy step → target resolution → endpoint preference → queue → provider result.

Next step did not stop: verify acknowledgement state/time and incident/policy generation; inspect scheduler/worker logs/metrics.

Next steps

• Escalation troubleshooting
• Inspect delivery