I was having trouble getting a new RAID array to assemble on startup, and before anyone responded to that post, I tried updating mdadm, but this ended up taking a very long time because of all the other dependencies. Somewhere in there it messed up something with my boot. I'm pretty sure grub was updated onto /dev/sda rather than the RAID array because I messed up (see below). But first, here is what is currently happening:

I have a RAID1 array, /dev/md1, that is mounted at /. I was able to boot until I ran the updates. Now when I boot I get

Gave up waiting for root device. Common problems:
-Boot args (cat /proc/cmdline) #unchanged, I checked
   -Check rootdelay= (did the system wait long enough?) 
   -Check root= (did the system wait for the right device?) #yes, correct UUID
-Missing modules (cat /proc/modules; ls /dev) #think this is my error
ALERT! /dev/disk/by-uui/1d3... does not exist. Dropping to a shell! 

and then I am dropped into ash. blikd shows that all my HDDs are there, and registered as raid members, but there are no RAID arrays.

This is super easy to fix. I can run

mdadm --assemble --scan

which will create the RAID arrays, and cat /proc/mdstat shows they are working fine. Then I merely give exit, and it boots like a champ.

The trouble is I can't get it to do this automatically. It's not trying to assemble the arrays before looking for them.

What I THINK happened: As I was upgrading mdadm I was prompted to update grub, and it told me that I needed to select where to install it as the UUID had changed. I, being an idiot, looked at my fstab and misread a comment that said #/ was on /dev/md1 during installation. I mistook this for .../dev/sda1.... Like I said, idiot (I've barely slept tyring to sort all this stuff out). Anyways, this seemed to make sense to me, so I selected /dev/sda1 (or just sda, I forget) in the prompt and it went on its merry way. Then I rebooted and this happened.

So it appears that I mistakenly put grub on a different partition, and so it's tyring to boot from there rather than creating the RAID arrays and then boot from /dev/md1. There are a bunch of posts about this generic error, but none of them seem relevant to what I'm trying to do. Any ideas?

1 Answer 1


This answer, provided by roaima, solved this problem. I had upgraded mdadm before using this solution, as stated in this post, but I recommend avoiding that if possible, as it obviously led to other errors.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .