i have created many apps and associated one machine per app, they are all set to autostop and autostart, most of them works as expected.
today i found the machine from an app get deleted after suspended, how can i find out why the machine is deleted?
if the app id is required for helping the debug, how can i provide it to employee of fly.io?
Hi… I don’t speak for Fly.io at all, but generally for things like that you need one of the Support plans. Occasionally a Fly employee will drift through the community forum here and look into something out of personal generosity (or just curiousity, 
), but that really isn’t the default…
It sounds like you have a fairly involved operation there, so having that would be wise, anyway. Going without support is really only for hobbiest dabblers like myself.
The Fly.io platform is prone to losing single Machines. Generally, you should have ≥2 per app. The most common cause is a failure of one of the underlying physical hosts.
If you can still see the Machine in fly m list, you can try fly m status to get the events list. That’s a little cryptic sometimes, but it might give some hints. (E.g., an attempted migration.)
The app-level logs (fly logs -a app-name) should also be consulted in such situations, as a reflex.
Hope this helps a lilttle!
And cheapskates like myself 
Do you mean it was gone completely from the Fly console? That is unusual, even in the case of a host failure.
True… There was another odd case of a missing Machine reported yesterday, as well.
https://community.fly.io/t/volume-bound-to-nonexistent-machine/25493
Perhaps metadata glitches are worming back in to the system, 
…
thanks so much @mayailurus and @halfer , i am ok to get response in a reasonable time so i am consulting the community for help.
curiously enough, this indeed seems to be the only way to get support from fly.io without a paid plan even we spend hundreds of dollars a month in fly.io.
the machine is gone completed in fly.io Dashboard and cli fly m list -a xxx.
the only hints i can find is the logs saying that the machine is suspended and failed to be wake up:
2025-07-27T07:13:19Z app[id] lax [info]Virtual machine has been suspended
2025-07-27T07:15:34Z proxy[id] lax [info]Starting machine
2025-07-27T07:15:34Z app[id] lax [info]2025-07-27T07:15:34.613696215 [id:main] Running Firecracker v1.12.1
2025-07-27T07:15:34Z app[id] lax [info]2025-07-27T07:15:34.613791917 [id:main] Listening on API socket ("/fc.sock").
2025-07-27T07:15:36Z runner[id] lax [info]machine restart policy set to 'no', not restarting
2025-07-27T07:15:39Z proxy[id] lax [error][PM03] could not wake up machine due to a timeout requesting from the machines API
after this log, my machine is gone without any other clues.
I would be minded to send details via email to billing@... so the engineers have a chance to log it as a bug. If you word it as a bug report rather than a request for technical assistance, I think you’ll sneak past the gatekeepers 
.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.