Description
Previously Cinder would retry infinitely if it encountered any non 2xx status when it delivered payloads to a registered webhook. They have recently updated to a more standard implementation that doesn't retry. https://docs.cinder.co/reference/webhooks#reliability
If the receiving server returns a 5xx error response, Cinder will continue retrying to send that webhook for up to one hour. There are 6 retries every 10 seconds then 10 retries every ten minutes. A 4xx response status code will not be retried. In the event of failed delivery, admins will be notified in-app.
We should update our implementation in cinder_webhook to report validation errors, etc, as 4xx instead. We probably don't want to return 400 errors for common occurrences like it being a non-AMO entity, etc, or we'll be hiding genuine errors.
Acceptance Criteria
### Acceptance Criteria
- [ ] cinder_webhook returns 4xx error for improper requests
Checks
┆Issue is synchronized with this Jira Task
Description
Previously Cinder would retry infinitely if it encountered any non 2xx status when it delivered payloads to a registered webhook. They have recently updated to a more standard implementation that doesn't retry. https://docs.cinder.co/reference/webhooks#reliability
We should update our implementation in cinder_webhook to report validation errors, etc, as 4xx instead. We probably don't want to return 400 errors for common occurrences like it being a non-AMO entity, etc, or we'll be hiding genuine errors.
Acceptance Criteria
Checks
┆Issue is synchronized with this Jira Task