# Retry policy

## Retry policy
If your endpoint returns a retryable status — `5xx`, `408`, `429`, or
a network/timeout error — Mercura retries on this schedule:
| Attempt | Wait before this attempt |
|  --- | --- |
| 1 | immediate |
| 2 | 30 seconds |
| 3 | 5 minutes |
| 4 | 30 minutes |
| 5 | 2 hours |
| 6 | 12 hours |

After attempt 6 the delivery is recorded as permanently `failed` and
shows up in the admin UI's recent-deliveries view; Mercura does not
retry further. The `Mercura-Event-Id` stays stable across all
attempts, so a delivery that eventually succeeds after some retries
is the same logical event as the earlier attempts.
A `2xx` response stops the retry chain immediately. Any other `4xx`
(except `408` and `429`) is treated as a permanent partner-side
configuration error and is not retried.

Version: 1.33.0