deloptes wrote:
> Don Armstrong wrote:
>
>> The error would be useful to know. Most likely one or more of them
>> dropped out of the array for some reason and you're booting off of one
>> which has a lower event count and it won't assemble.
>>
>> But it could be any number of things.
>>
>> The output of mdadm --examine /dev/sd[abcdef]1; when md0 fails to
>> assemble would also be useful.
>
> In my case it is Dell OptiPlex 7xx - I have it under the desk for 2y now -
> but it looks like it is 5y old.
> When I looked into the drives they were detected but md disks seemed to be
> messed and not easy recovarable.
>
> What I observed that only raid0 was loaded but not raid1. After removing
> raid0 and loading raid1 I was able to see at least the partitions of the
> drives but I did not have time to go further, so as I had to do a lot in
> the office and @home I just shut it down. I hope I'll have some time next
> week to play with that. Good that I do not need a remote machine at the
> moment.
>
> I hope this helps
>
> regards
Just FYI
I managed to solve this on the OptiPlex today - it took me 3h and several
reboots.
The thing is that the BIOS is locked by the admins and the built in raid
controller can not be deactivated. Thus obviously when upgrading the array
info was dumped into mdadm.conf from the built in array or may be even
before - I don't know, which lead to broken initrd.
When in the initramfs shell I had to unload all related modules because I
had only sda and sdb (no prtitions). After loading ata_piix I get all the
partitions and loading raid1 + ext4 was enough to start raid and mount
root.
mdadm -A -R /dev/md0 etc
then
mount -o bind /proc /root/proc
sys, run, dev ...
mount /dev/md2 /root
cd /root
exec /sbin/chroot . /bin/sh <<- EOF >dev/console 2>&1
exec /sbin/init ${CMDLINE}
EOF
install 3.16 kernel and update-initramfs, but first remove the wrong ARRAY
info from mdadm.conf
DONE.
messy
regards
