How are you handling Azure Frontdoor downtime?

One solid approach is to implement redundant global load balancing solutions. For example, you could set up standby CloudFlare tunnel instances that are ready to take over traffic when Frontdoor goes down. Just flip the DNS from Frontdoor to CloudFlare and make sure everything is synchronized and tested beforehand. If you're already in the CloudFlare ecosystem, it might make sense to treat Frontdoor as your backup solution.

It really depends on your setup, as many Microsoft services rely on Frontdoor, making it tricky to gauge the full impact during an outage. The best move is to ensure you have an equivalent service ready to kick in, like CloudFlare, and build in contingencies that can be activated smoothly when necessary.

Currently, my strategy is pretty simple—wait for the service to come back online. I've proposed a few options, but they were seen as too expensive. Unfortunately, that's just the reality we deal with, and I don't get criticized too harshly for outages beyond my control.

If you're dependent on just one region, you could point traffic directly to your App Gateway. But just be aware, that won't be a foolproof solution if Frontdoor goes down. You might want to think about additional layers to maintain uptime.

We've faced similar issues and eventually transitioned our public ingress to Cloudflare Tunnel. We set up multiple cloudflared instances for each environment, which integrates seamlessly since we’re already using Cloudflare Zero Trust. Right now, Frontdoor is just a backup for us; switching the DNS is a manual process for failover, but we're looking into automating health checks next!