Why is my Linux based containerised Web App's endpoints unavailable with nor errors and 100% healthy?

Question

I have a linux-based containerised Web App that is a FastAPI.
The App starts and works and then for no apparent reason stops working.
There are no errors showing anywhere, no resource/quotas issues, literally no error in Azure or in the console log or docker build alluding to any problem.

I have a B2 plan with ALWAYS ON.

Perhaps my app is running out of 100 minutes on B2, but how would i know?
Its possible its consuming resources when the FastAPI is idle, yes, but until I can analyse this and think of a way to turn off the event loops or whatever to quieten it down, I'm at a loss as to how to discover the root cause!!

Can anyone help?

Yes I have tried AI powered help
I have tried looking at 50xxx errors, sure i have the odd one most of the 504's and gateway timeouts are because its down and at the time the web app is unavailable, but nothing says its down even the /health heartbeat is showing a 200 success status.

I am not using any virtual networking or network restrictions, its all public nothing clever!

Could something be disconnecting, the log stream just shows event loop 200 success health checks.

but my FastAPI endpoint goes dead after a few hours. I haven't timed it, even application insights say my app has been 100% available over last 24 hours, but it isn't!

Help!!!