I recently switched the repo and subdirectory for this app/builder… and now we’re getting a weird failure where it clones the repo and then bombs with no info.
I can’t find more specific logs anywhere… and fly deploy works just fine. How do you dig into the “real logs” for a deployment to see where this explosion is happening?
I am getting this error on the clone… so maybe a system issue?
INFO [fly api proxy] listening at /.fly/api
Machine created and started in 3.15s
2025/02/27 13:22:34 INFO SSH listening listen_address=[fdaa:11:706b:a7b:379:affe:76d6:2]:22[Fly.io Deployer]{"id":1,"step":"__root__","type":"event:start","time":7.397999999980698e-05,"payload":{"ts":"2025-02-27T13:22:35.484625Z"}}
INFO Starting clean up.
WARN could not unmount /rootfs: EINVAL: Invalid argument
[ 6.863812] reboot: Restarting system
machine restart policy set to 'no', not restarting
I saw a report with EINVAL in it recently, try one of these:
Thanks for the reference!
A lot of those look pretty different, tbh. These are trying to unmount /rootfs? It feels like a low level issue that is not necessarily related to my infrastructure.
I’ll definitely keep digging though! For now fly deploy is a decent workaround, except for the nicety of auto-CI 
Yeah, I’d agree with that.
I’m not familiar with the GitHub integration in Fly, but of course you can do a flyctl deploy in a standard GitHub Actions pipeline.