Why Does Stopping the IaaS VM Provider Service Cause a 502 Bad Gateway Error on My Web Apps?

A 502 Bad Gateway error generally means that the Application Gateway can't reach any backend servers. That's likely because when the IaaS VM Provider is stopped, the health probes fail, leading the gateway to stop sending traffic. You might need to keep that service running or check its settings.

From what I gather, the IaaS VM Provider is essential for Azure VMs that allows them to communicate back to the data center. It’s possible your IIS setup is configured to start only when the IaaS service is running so that it doesn’t run into issues with partial connectivity. You should verify the dependencies of your services.

It seems like when you stop the IaaS VM Provider service, it's affecting how traffic is routed. Since that service is central to communication with the Azure infrastructure, the websites depend on it to function properly. Maybe check if there are service dependencies that prevent IIS from running when the provider is off.

Interestingly, I think there might be more at play here. I've noticed that the 502 error can still happen even if the IaaS VM Provider is running. I found an event log message indicating TLS client authentication issues. This might be causing intermittent failures, which could explain the inconsistent 502 responses.