Hello, I’m looking for information on if, and how, messages stored in the outbox are retried after an endpoint is restarted.
The scenario I’m trying to deal with is message bus unavailability during send, after the message has been committed to the outbox in a transaction with other database changes. Troubleshooting or other health checks may restart the endpoint and lose any in-memory retries. Would this result in a zombie record?
Hi Sean,
Our documentation goes into great detail on how the outbox works and avoids these kind of problems, have you read it, is there anything there that does not make sense?
Regards
John
Hey John, I have, but I think it may have just clicked for me. The incoming message not being acked will result in a retry, which will deduplicate the message so it’s not processed again but will retry sending the outbound messages. Is this accurate?