Start here. This is the direct spoken answer to practice first.
Why this question matters
Edge incidents are visible and urgent because users cannot even reach the application path. This question tests basic but important production boundary thinking.
I would check whether the problem is DNS resolution, App Service custom domain binding, TLS certificate validity, certificate binding, HTTPS-only behavior, or a proxy in front of App Service. I would test from outside the corporate network and inspect the certificate presented to the client, including hostname and expiration. If only some users are affected, I would consider DNS propagation, caching, regional routing, or a CDN/front-door layer. I would communicate clearly because certificate and domain incidents look severe to users.