What to do when your Oracle Cloud Infrastructure instance gets instance
Recently right after doing a regular apt-get dist-upgrade my OCI instance got stuck for a few hours in the “reboot” state.
Using the web GUI to restart didn’t work, and the result was the same through the command line.
After a few hours, the machine did change to “starting” but it also got stuck there till the next morning. Web GUI and command line tool did not help again.
So I started searching… I even created an account on Oracle Communities (which is horribly organize, really hard to get to what you need; plus when I clicked on the right resource (I think), it told me I don’t have permission to access it…. wth?, it’s supposed to be community support)
Anyways… after trying to add a payment method to my account, and being blocked by their anti-fraud system telling me “We’re unable to process your transaction. If you continue to encounter the error, you may contact Oracle Customer Service..”; I tried chatting with Customer Service, which just told me they couldn’t do anything….
So I decided to see what else I could do.
Being on the free tier, and using all my 4 Ampere instances, I decided to terminate the stuck one *without* terminating the boot volume, and provision another one, without a boot volume….
Of course that doesn’t seem to be a use case Oracle thought of, so I had to provision a new instance with its own boot volume, which automatically terminates the new instance as it puts me over the free tier limit.
So my solution was to attach the boot volume from the now terminated instance (ex-stuck) to a running one, run their “iscsi” commands to let Linux know about the volume (Thanks God they don’t differenciate between attaching a boot volume and a regular volume); mount the volume, and then rsync to my laptop.
Next step is to simply terminate the boot volume, that I now have copied all the files I wanted from; Create a new instance and restore the data back to it.
Hope this helps someone 🙂