Hi jbpaux
about your different architecture, I think that the focus is on the scenario:
1. original Pierre_Roman scenario: more work on the client side to check if current IP is different from the A Record --> therefore less Azure Function triggers (only when update is needed) --> better choice if client can manage all these checks and in scenarios with a very large number of clients and frequent changes (e.g.: 5.000 devices that update every 10 mins) to avoid overloading Azure Function,
2. your scenario: more work on the Cloud side to check if the caller IP (the current client IP) is different from the A Record --> best choice with less sophisticated client devices that cannot handle complex scripts and can only call the Azure Function URL in a scheduled way every specified time
What do you think about it?