Proceed into rescue mode, run through the prompts, and execute a shell.
From the shell prompt. You'll want to stop all md devices. *(please forgive the freehand scripting), and then run mdadm --zero-superblock /dev/sd* on the remaining disk drives.
_________________________________________________________________
#!/bin/bash
for i in `ls /dev/md*`; do
mdadm --stop $i
done
for i in `ls /dev/sd*`; do
mdadm --zero-superblock $i
done
That should do it.
_________________________________________________________________
It would be nice if we fix curtin to detect errant superblocks, and possibly clear them. It might also be nice to be able to select some sort of fakeraid functionality from maas for each machine. I haven't fully thought this through, and I'll leave that up to the maas team to sort out.
So we discovered a workaround for this.
It turns out that disks had fakeraid superblocks. When d-i or curtin go to install, they are getting screwed up by this.
So, our solution was to boot the servers with an Ubuntu livecd. We used https:/ /help.ubuntu. com/community/ Installation/ MinimalCD
Proceed into rescue mode, run through the prompts, and execute a shell.
From the shell prompt. You'll want to stop all md devices. *(please forgive the freehand scripting), and then run mdadm --zero-superblock /dev/sd* on the remaining disk drives. _______ _______ _______ _______ _______ _______ _______ _______ __
_______
#!/bin/bash
for i in `ls /dev/md*`; do
mdadm --stop $i
done
for i in `ls /dev/sd*`; do
mdadm --zero-superblock $i
done
That should do it. _______ _______ _______ _______ _______ _______ _______ _______ __
_______
It would be nice if we fix curtin to detect errant superblocks, and possibly clear them. It might also be nice to be able to select some sort of fakeraid functionality from maas for each machine. I haven't fully thought this through, and I'll leave that up to the maas team to sort out.