Sometimes, for no apparent reason, our Live API stops working.
How the situation looks like for us:
Please take into account that we have a Staging App Service slot.
1) Api worked fine, we turned on the logs then in web.config, then maybe it automatically restarted and couldn’t start, or it just died (i.e. if it automatically restarted, we need to find out if the issue is on your side or on our side)
2) We discover that API does not work, shows 503 (service unavailable) error.
3) We go to check web.config, and see that the logs are turned off (it is strange because we have enabled it). We go to the log folder and see them exclusively for the 4th of June (it is 12th of June now)
4) We try to just restart the App Service, it does not work - 502.5 error as far as we remember.
5) We turn on the logs back (any saving of web.config should cause restart), but nothing worked, everything is the same 502.5 (i.e. the previous one) error.
6) We check the logs for today (12th of June), there is the same thing: invalid runtimeconfig.json
7) We turn off the logging
8) We restart App Service and everything works.
Could there be a problem in logs that are switched on? Maybe. But even without them, the API stopped working sometimes. But with them, it also does not really want to start.
But eventually , it starts at some point. It is very strange.
And also, which could theoretically affect our API: we have a slot, and it works without issues.
The question is: can the main App Service stop working due to the slot? Meaning if we setup a slot for QA App Service, for example named QA1, then maybe after some time the app on QA App Service will break as well?
But that would be very strange, because on out other project slots work, we swap them, and everything is ok.
What could be the reasons for this issue?