3

I'm trying to downgrade my EC2 instance root device (SSD) to a Cold HDD. I performed the following:

  1. Stopped my instance
  2. Detached the root volume from my instance via the console (was mounted on /dev/xvda1) (not by force).
  3. Created a snapshot of the detached root volume in the same availability area as the instance.
  4. Downgraded my instance from t2.xlarge to t2.micro
  5. Created a new Cold HDD volume from that snapshot in the same availability area.
  6. Attached the newly created Cold HDD volume as /dev/xvda to the instance
  7. Rebooted the instance

Now I'm getting the problem that the instance stays "pending" when I reboot, and after 30 seconds or so it goes back to "stopped". The reason given is:

Server.InternalError: Internal error on launch

When I reattach the old volume, it reboots fine, so the error is coming from the new volume. Can someone tell me if I'm doing something wrong?

Thanks!

Alberto Deca
  • 165
  • 1
  • 12

1 Answers1

3

Answering my own question here: problem was Cold HDD is not an option to attach as a root of a new instance, but "Magnetic" was. Changed volume type to "Magnetic" in step 5 and it fixed it.

Alberto Deca
  • 165
  • 1
  • 12
  • This is correct, because [booting from `sc1` or `st1` isn't supported](http://stackoverflow.com/a/36735612/1695906). – Michael - sqlbot Feb 06 '17 at 18:39
  • incidentally, this helped me with launching a brand new P2 instance which didn't seem to have supported SSD disk for some reason... thanks :) – Zathrus Writer Jul 13 '17 at 17:27