FakeRAID 0+1, Upgrades, Partitions, and Dual-boot

[Johnius]

I've been having trouble with both my computers and upgrading from 12.04 to 14.04. Each computer wouldn't boot after the upgrade. The laptop would boot into the old linux headers and I got the upgrade to take by running it again. The desktop won't even boot into the old headers. And, of course, it's my main computer with all of my data. Naturally, I'm frustrated, but it has to be something I'm doing. So I downloaded 14.04 Live for fresh installs, which I couldn't get to work on a USB stick (no idea why, still won't work) so I gave up and got out a DVD and (stupid) Win7 won't burn ISO DVDs natively anymore (smooth M$, smooth). Finally got a DVD to work and installed on the laptop (which, after the updates, the wifi worked with no hiccups. Props).

But I put the DVD in the desktop and it won't even let me into the partition tables. The desktop is (was) dual booting Windows 7 and Kubuntu 12.04. I upgraded from Muon and rebooted (after several errors on the upgrade) and it just dropped me into BusyBox. I'm running FakeRAID 0+1 on the desktop, with separate Windows, Win Reserved, /boot, /, /home, swap partitions. I don't even know where to start because I've never used a live CD before.

[Johnius]

Did some digging, this is the output of /var/log/syslog that the LiveDVD gives an error message:

Sep 3 03:35:36 kubuntu activate-dmraid: ERROR: either the required RAID set not found or more options required
Sep 3 03:35:36 kubuntu activate-dmraid: no raid sets and with names: "pdc_fgejdfii-0"
Sep 3 03:35:37 kubuntu kernel: [ 338.312214] xor: automatically using best checksumming function:
Sep 3 03:35:37 kubuntu kernel: [ 338.348368] avx : 10498.000 MB/sec
Sep 3 03:35:37 kubuntu kernel: [ 338.580753] raid6: sse2x1 1781 MB/s
Sep 3 03:35:37 kubuntu kernel: [ 338.648867] raid6: sse2x2 3449 MB/s
Sep 3 03:35:37 kubuntu kernel: [ 338.716977] raid6: sse2x4 4500 MB/s
Sep 3 03:35:37 kubuntu kernel: [ 338.716981] raid6: using algorithm sse2x4 (4500 MB/s)
Sep 3 03:35:37 kubuntu kernel: [ 338.716984] raid6: using ssse3x2 recovery algorithm
Sep 3 03:35:37 kubuntu kernel: [ 338.910933] bio: create slab <bio-1> at 1
Sep 3 03:35:37 kubuntu kernel: [ 338.911427] Btrfs loaded
Sep 3 03:35:37 kubuntu kernel: [ 339.038024] JFS: nTxBlock = 8192, nTxLock = 65536
Sep 3 03:35:38 kubuntu kernel: [ 339.250284] SGI XFS with ACLs, security attributes, realtime, large block/inode numbers, no debug enabled
Sep 3 03:35:38 kubuntu partman: No matching physical volumes found
Sep 3 03:35:38 kubuntu partman: Reading all physical volumes. This may take a while...
Sep 3 03:35:38 kubuntu partman: No volume groups found
Sep 3 03:35:38 kubuntu partman-lvm: No volume groups found
Sep 3 03:35:39 kubuntu ubiquity[5087]: debconffilter_done: ubi-partman (current: ubi-partman)
Sep 3 03:35:39 kubuntu ubiquity[5087]: dbfilter_handle_status: ('ubi-partman', 141)
Sep 3 03:35:56 kubuntu ubiquity[5087]: dbfilter_handle_status: answer 524288

I'm going to check out the RAID BIOS and see if there's any trouble there. dmraid found the partitions, but it won't display them and says the volume group doesn't exist, but the computer still runs Win7 fine (you know, for how fine Win7 runs...)

[SteveRiley]

Looks like mdadm has replaced dmraid in *buntu 14.04.

http://askubuntu.com/questions/45551...ror-question-m

[oshunluvr]

If you're using BIOS based RAID, I'd boot back into the older system that can access those files and back them up. Then I'd reformat and use mdadm RAID instead or move your linux system to non-BIOS RAID disks and use mdadm RAID there. IMO, easier and more transferable. One major downside is mdadm RAID is not usable by Windows.

Although, after reading the link Steve provided it seems the incompatibilities may have been resolved. It used to be true that BIOS RAID was incompatible with other BIOS RAID version - meaning if your mobo died, likely so did your array. mdadm RAID does not have that tie-in to BIOS.

Admittedly, my opinion on this matter may be outdated as I stopped using RAID altogether once btrfs made it into the kernel, and I don't use Windows.

[Johnius]

Thank you for the link, StevenRiley. oshunluvr, I tried booting into the older systems and none of them function anymore for various reasons. I keep my data on a separate /home partition, so I can just overwrite / and /boot and it's no big deal. If I can find the partition table...

[Johnius]

Mdadm doesn't find the array, either. I reactivated dmraid and it found the array and the partitions, but it doesn't display correctly in the Disk Setup. The /var/log/syslog says, "Couldn't find valid filesystem superblock". "then /dev/mapper/arrayname not found in os-prober output". I'm certainly /not/ clearing my superblocks. Did that when I didn't have any data to lose and it took me two days to recover.

[Johnius]

Tweaked my tactic on the install:

sudo dmraid -ay

Before attempting the install from the live CD. Cleared the "swap" partitions from the mirror and the stripe (which show up as -1 and -0, respectively) and put everything on the 0+1 (which shows up without the modifiers). Installed just fine... that is, without throwing any errors. Upon boot, GRUB holds up like it should and waits for a selection, but Windows is no longer an option (which should be easy to fix). But then I get

Gave up waiting for root device. Common problems:
Boot args (cat /proc/cmdline)
Check rootdelay=
Check root=
Missing modules
ALERT! /dev/mapper/pdc_fgejdfii4 does not exist. Dropping into a shell!

That last line is obviously the trouble. It existed long enough for me to install to it, what happened?

I ran cat /proc/modules and dmraid doesn't show up, so perhaps that is the problem? Could it have been removed during the install clean up?

[Johnius]

After I drop into BusyBox, a simple

dmraid -ay
exit

fixes the problem and off I go. But I don't know where to put dmraid in any of the config files so that it activates my RAID automatically. I don't feel like babysitting my computer during the boot process.

[oshunluvr]

My best guess is it's detecting your RAID as degraded. You could try update-initramfs or adding bootdegraded-true to your boot parameters.

Have you tried removing quiet splash from the boot parameters to see if an error is occurring that you're not seeing?

Also, running sudo dmraid -r and then rebooting might work, but I'd verify that this won't wipe the data first.

[Johnius]

I've had my RAID degrade. It throws a "RAID degraded: Continue boot? Y/N" style message. I tried update-intramfs and even reinstalled dmraid to no effect. I booted into the recovery headers to watch for errors and didn't see anything other than it can't find the RAID array.

I was able to find the /scripts/init-premount and /scripts/local-top mentioned in the recovery error from the live CD and added dmraid -ay to them. I can't find them from the installed OS to ensure that they still exist.

[Johnius]

Got it!

made a file in

/etc/initramfs-tools/scripts/init-premount

with the script

exec dmraid -ay
exec sleep 1

then in Terminal

sudo update initramfs -k all -uv

Thank you for your help! I always appreciate the Kubuntu forums.

[oshunluvr]

Good job. Glad you got it figured out.