Has something changed in the App Service Infrastrcuture?
I can be barely keep 4 App Services up for 24 hours. We used to have 8 up until 3-4 weeks ago, and never really had any issues with memory, resources or anything else, We got rid of 3 or 4 of those. There are 2 .net core 3.1 apps, one is a web api, and another is a front end. Dev and prod for each.
We are still in beta mode so we go many hours with zero activity and sites will just stop responding in the middle of the day, and now we get frequent 500.37 startup timeouts.
We have made app changes recently to adopt DI for HTTPCLient and database access, all have improved memory consumption locally. I can't get any of the apps to go above 150M in Visual studio locally. There are extremely erratic memory consumption in the Azure apps, spiking and droppng for no apparent reason.
Maybe we are doing something fundamentally wrong, the webapi app has not changed at all in months, but experiences many of the same issues. The dev app and prod app often times have very different memory consumption statistics. I've reviewed pretty much everything I can find, it appears we are doing everything correctly.
We are on S1 app service plan, and Azure keeps telling us we should scale up, but I'm unsure why that would be the answer, we are well within the resource consumption for that plan.
If anyone has any thoughts on what we can try, happy to share more about our configuration and app.