For this eventuality, the Open API introduced more recently an Event API TMF688 (accessible
here), which is intended to wrap durable message buses.
You are probably right that point-to-point messaging needs more work, the notifier (message producer) needs to handle retries, etc. Having said that, even when working against a message bus it is possible that the bus is down and you'll need to retry, but it's easier to manage, presumably in one place.
------------------------------
Jonathan Goldberg
Amdocs Management Limited
Any opinions and statements made by me on this forum are purely personal, and do not necessarily reflect the position of the TM Forum or my employer.
------------------------------
Original Message:
Sent: Dec 15, 2021 07:28
From: santhosh thatipally
Subject: API Notifications - why its designed with endpoint address communication
Hi Experts.
I know the event notifications model is discussed and designed accordingly. I would like know the history regarding why the event notifications are designed with registering the communication endpoint address and not the messaging model (not like topic with publisher and subscription model).
In this synchronous event notifications model, how the event producer API will ensure the reliability/guarantee of the event to consumer API?
What if the consumer service is down for some time and the producer API tried for sending event?
------------------------------
Thanks
Santhosh Thatipally
Oracle Corporation
------------------------------