One of my servers finally decided to show this message consistently and
stopped booting altogether after the last kernel upgrade.

There are 3 kernels on the system. (2.6.31 - 14/15/16)
After a few hours of different combinations I got recovery kernel 2.6.31-14 to 
boot with:
- root=/dev/vg0/root (instead of /dev/mapper/vg0-root)
- rootdelay=200 (maybe less is possible, but 60 is not enough)
- no 'ro'

If one of the above conditions is not met, the server will show busybox
and not mount /

(The next part may not have importance for this bug.)
If they are met, mountall will give the usual trouble. (/home is on NFS)
Luckily, the recovery kernel does not need /home, and the system can be brought 
in a workable state by executing:
# dhclient3 eth0 (# service networking start will not give an IP)
# Ctrl+D
Now, other LVM volumes will be mounted, and mountall will hang on the NFS part
# Escape (NFS filesystems will not be mounted, even if the network is up)
Now, I could upgrade the system, but to no avail.
(Don't press Ctrl+D again. Mountall will hang without giving the possibility to 
escape it and reboot is the only option.)

After getting the system in a workable state, I noticed that lvm2 was
not installed. Isn't it needed for installation on LVM volumes? I
installed it, but still couldn't boot normally.

-- 
karmic rc: root device sometimes not found
https://bugs.launchpad.net/bugs/460914
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to