How to diagnose network issues with Azure webapp (linux-container)

Question

I have two webapps in the same data center in Azure. Both are .NET core apps running in docker containers. Both use the same MongoDB cluster hosted in the same data center (managed by MongoDB Atlas):

Starting yesterday, webapp2 can no longer connect to the mongo cluster.

In order to rule out any firewall issues, I've temporarily allowed incoming network traffic to the "prod" cluster from any IP address, but the issue persists.

What I've been able to establish is that:

• webapp2 cannot connect to the "prod" cluster, even when I try with the exact same connection string as webapp1 uses.
• webapp2 can connect to the "test" cluster without a problem.

Since the Kudu Debug Console just hangs when I run ping and tcpping, I'm pretty much left without any tools to further diagnose the issue.

Does anyone have any tips on what else I might try?

Answer 1

Not only are the IP addresses whitelisted, I've even allowed network traffic from ALL ip addresses, like I wrote.

I might mention that everything magically started working again an hour ago. No code change, and Atlas says they didn't change anything on their end. Really scary!

Are you aware of any intermittent network issues in the west-europe datacenter during the last two days?