I have a server set up with Nginx to receive whatsapp requests from the webhook and process them with a Flask application. The server has been working for the last 2 months until today 29th of july, currently (and since 17:06:00 GMT-3) it is not receiving any whatsapp requests.

The behavior was the following: The server usually receives 3 or 4 requests per minute as people chat with it. It's running Linux on the other side of the continent, all dockerized with its belts and whistles. The server was restarted a few times (as we always do) and we introduced an update (updated .py files as per normal development process). Everything went smoothly, then 20 or 30 minutes later at 17:01:40 GMT-3 it stopped receiving requests, you would send a message to the whatsapp account and it would only give you one tick, then at around 17:05:00 it received a bunch of requests from WPP Business and those were the last. after 17:06:00 no further requests from whatsapp messages went to the server.

Subscribe requests are received and any other request that is not related to whatsapp is correctly received by the server, we have a web form that works correctly from the exact same Flask app and we have ruled out that it is a problem on our setup. The server does not receive the requests period. Nginx (which is on a different container than the app) doesn't receive whatsapp business requests (otherwise it works perfectly fine)

I have set up the application locally on my PC, using Windows and no bells no whistles, just on my own db and own dev enviroment. I tunneled my app using ngrok creating a temporary hostname for my application, I launch the app and subscribe whatsapp business to my locally ran app, the subscribe request was received, and requests started flowing, for two minutes, then they stopped, again.

I have now subscribed the real production server back to the webhook and it does not receive my messages, it only received ONE request for the last hour, that was an old request (older than 5 minutes) and this request came 1 minute after hooking up the webhook (this is normal behavior, old requests come to the server all the time and we handle them gracefully)