Sending communication over the network can fail due to intermittent network outage.To mitigate this case and decrease the number of unsent communications a re-try mechanism should try to resend the failed notification for a number of time based on a recurrent time interval(5min).
Take in consideration Exponential Backoff pattern to handle transient fault.
Retry pattern example:
- Application(Mambu) invokes the notification call —> Request fails and we receive 500 internal server error for e.g.
- Application (Mambu) waits for a short interval and tries again —> the request still fails throwing 500
- Application (Mambu) waits for a longer interval and tries again —> we receive success response 200 and we break the circuit
- In the event we still receive errors and we’ve reached the maximum retry count, Mambu will set the notification State to Failed