Docs › Webhooks
Retries & delivery logs
Retry schedule, auto-disable, delivery attempts.
When a message is retried
A delivery is retried when your endpoint times out, can't be reached, or answers 5xx, 408, 409 or 429. Any other 4xx is treated as final.
Schedule
The same as Nash/Svix: attempts after 0 s, 5 s, 5 min, 30 min, 2 h, 5 h, 10 h and 10 h. That is eight attempts over about a day.
Controlling delivery from your side
| You respond with | Effect |
|---|---|
2xx | Delivered. |
410 Gone | The endpoint is disabled. |
Header webhook-delivery: disable | The endpoint is disabled. |
Header webhook-delivery: abort-message | Stops retrying this message only. |
Ten consecutive messages that exhaust their retries auto-disable the endpoint. Re-enable it with PATCH /v1/webhooks/{id} and { "isActive": true }.
Delivery log
curl -s "https://api.trydragonfly.com/v1/webhooks/ENDPOINT_ID/deliveries?limit=50" -H "Authorization: Bearer $DRAGONFLY_API_KEY"Each attempt shows eventType, status, attempts, responseStatus, error, nextAttemptAt and deliveredAt.
Testing your receiver
POST /v1/webhooks/{id}/test replays your most recent order's current state as a delivery.updated message. You must have created at least one order with this key. In sandbox, every simulated status change also produces real webhook messages.
Something unclear or missing? Tell us.