Configure escalation retries and delays

Before you begin

Know incident response objective, provider delivery characteristics, responder expectations, and total acceptable time before each fallback.

Open the feature

Open Escalation policies → select policy → Steps, then edit the intended step timing.

Configure timing

1. Set the wait/delay before advancement according to product fields.
2. Configure bounded retry behavior only where it improves delivery without duplicate storm risk.
3. Calculate cumulative elapsed time through all preceding steps/retries.
4. Save and review the full timeline.

What OpsKnight does

The escalation planner schedules eligible work for the current incident/policy generation. Acknowledgement or terminal state changes suppress/complete later work according to lifecycle rules. Provider retry and policy advancement are distinct concerns.

Verify it worked

Run a controlled test with timestamps. Confirm first delivery, retry/advance time, acknowledgement stopping behavior, and no unexpected duplicate provider messages.

Change or undo timing

Restore the prior reviewed timing values and retest. Already emitted notifications cannot be recalled; communicate accidental pages.

Troubleshooting

Next step is late: inspect scheduler/queue oldest age, configured cumulative delays, provider retries, and incident acknowledgement state.

Duplicate storm: stop the test through correct incident lifecycle, inspect retry versus policy scheduling, and reduce unsafe repeated paths.

Next steps

• Attach policy to service
• Test policy