We have a lambda which calls a service. To handle the scenario of the service being down.
We want to have a retry queue and a dead letter queue.
If the lambda is not able to send the event to service the first time, it will send it to retry queue.
The same event will be fed into lambda again from the retry queue. If it fails again then we try one more time by sending the event to retry queue.
After that if failure happens again we send it to dead letter queue.
My question is that why we need retry queue? We can try to send the event to the service two times in lambda itself and if we fail both the times, then send just send it to dead letter queue.