I already had 2 cases where my deployed fly apps crash and fly doesn’t restart them.
How to restart machines on failed health checks?
I am deploying Next.js, which seems to run multiple processes, so the container will not terminate and stay in an invalid state, not responding to any request.
So unfortunately there isn’t a feature that has been implemented yet to restart an app when health checks fail. I know you can do it for V1 apps, but for V2 apps, that’s not something that’s possible just yet!
The only restart feature we have outside of fly machine restart would be the auto stop/start machines, which you can set in your fly.toml file. Which is described here- Automatically stop and start Machines · Fly Docs