I was alerted that I had an app on a dead machine.
Some of your apps in FRA are on a host which has suffered a hardware failure and will be down for an extended period.
So, I did a quick scale up/down to get the app running on a new machine (fly scale count 2 && fly scale count 1). Maybe not the most straight forward but it worked.
However, the defunct zombie machine is still listed in the “Machines” overview alongside the one properly working. How can I get rid off it?
Hi Marcel, at the moment you can’t get rid of it because the host is offline, so there’s no way to contact the host server and tell it to remove your machine.
You aren’t getting charged for the resources on a host while it’s offline and the Fly Platform knows that the zombie machine is offline so it won’t attempt to route any requests to it.
If we discover that the hardware failure is irreparable then we do have a way to delete all the metadata about the Fly Machines on that host from the platform. But for the moment we think that we are going to be able to repair that host and once we do that Machine will come back online.
I agree that this is somewhat clunky and it would be better to have a way to destroy the Machine even while the host is down, but that’s not a feature we’ve had time to implement yet. For now, I hope you don’t mind living with that alert on your dash until we get the host back online.
Fantastic feedback, thank you! Does that mean I need to monitor my dashboard (manually) in order to not miss the opportunity to delete the machine in due time i.e. before I get charged for it? Or will Fly kill it automatically because my scale count is 1 and there already is a running instance when the old one reappears?
The former, unfortunately. But now that I know about your situation I will keep it in mind and (unless you respond and tell me not to) destroy that machine for you when the host comes back online.