PDA

View Full Version : am I totaly hosed?



Avix
22nd December 2008, 07:46 PM
I think I screwed something up. this morning. this is what I got on my F8 box

reading all physical volumes
no volume groups found
volume group "VolGroup00" not found
unable to access resume device (/dev/VolGroup00/LogVol01)
mount: could not find filesystem '/dev/root'
setuproot: moving /dev failed: no such file or directory
setuproot: error mounting /proc: no such file or directory
setuproot: error mounting /sys: no such file or directory
switchroot: mount failed: no such file or directory

this is (was) my main machine. last backup was a week ago. is there any way to try to recover the data?
(mostly docs and e-mail)

or should I just wipe it and use it for storage?

pwca
22nd December 2008, 07:55 PM
I would use a LiveCD/DVD and once booted on that check the contents of the hard drive paying special attention to:

/boot
xorg.conf
inittab

check to see what's missing or changed.

Using some sort of partition editor like gparted you will want to investigate your partitions, logical volumes and master boot record.


btw... love the signature line. :-)

Hlingler
22nd December 2008, 07:56 PM
So, pray tell: what did you do prior to the meltdown? It looks like the kind of error I would expect from a borked OS upgrade. What, if any, updates/upgrades or other software manipulation happened recently?? Any other important changes???

V

Jake
22nd December 2008, 08:26 PM
Yup. would be handy to know what you was doing prior to it going wrong. if you was playing with your LVM volumes (LVM Manager), then that could very well be the problem.

Vector
22nd December 2008, 10:55 PM
This post may be relevant and helpful:
http://forums.fedoraforum.org/showthread.php?t=208502

stevea
22nd December 2008, 11:04 PM
You're certainly hosed to some extent.\

I think Grim Alex-Grim is wrong. The messages you got are the the in-memory kernel tryng to mount the LVM root file system and failing. So your MBR and /boot partition are fine. It's the LVM that isn't recognized.

DO try booting the live CD (or maybe even the install DVD in rescue mode).
One of these *may* have the "system-config-lvm tool on it, and that would tell you a lot about the status of the LVM.

Otherwise post the result of "fdisk -l" and "pvscan" and "lvscan" from the optical boot.

Vector
22nd December 2008, 11:24 PM
Yeah, i never work with lvm; hence the "*may* help" part of the statement ;)

Avix
23rd December 2008, 12:56 AM
these are all from my F10 box with the F8 drive hooked in.
I tried Knoptix and Gparted Live but all I got was kernal panic.

I'm going to try some of the tools referred to in the other thread


[root@tux ~]# fdisk -l

Disk /dev/sda: 250.0 GB, 250000000000 bytes
255 heads, 63 sectors/track, 30394 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0b7ed900

Device Boot Start End Blocks Id System
/dev/sda1 * 1 25 200781 83 Linux

Disk /dev/sdb: 320.0 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x130072d4

Device Boot Start End Blocks Id System
/dev/sdb1 * 1 25 200781 83 Linux
/dev/sdb2 26 38913 312367860 8e Linux LVM

Disk /dev/sdc: 123.5 GB, 123522416640 bytes
255 heads, 63 sectors/track, 15017 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa15acce3

Device Boot Start End Blocks Id System
/dev/sdc1 1 15018 120627328+ b W95 FAT32

Disk /dev/sde: 8086 MB, 8086618112 bytes
255 heads, 63 sectors/track, 983 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x04dd5721

Device Boot Start End Blocks Id System
/dev/sde1 * 1 984 7897056+ b W95 FAT32
Partition 1 has different physical/logical endings:
phys=(982, 254, 63) logical=(983, 36, 13)
[root@tux ~]#

[root@tux ~]# pvscan
PV /dev/sdb2 VG VolGroup00 lvm2 [297.88 GB / 32.00 MB free]
Total: 1 [297.88 GB] / in use: 1 [297.88 GB] / in no VG: 0 [0 ]
[root@tux ~]#

ACTIVE '/dev/VolGroup00/LogVol00' [292.50 GB] inherit
ACTIVE '/dev/VolGroup00/LogVol01' [5.34 GB] inherit
[root@tux ~]#

stevea
23rd December 2008, 02:37 AM
Hey - good news - your LVM file systems *appear* intact.
I suspect you somehow included a new kernel without building a proper initrd.

Can you boot the previous kernel version revision ? (comment out hiddenmenu in /boot/grug/grub.conf).

[root@tux ~]# pvscan
PV /dev/sdb2 VG VolGroup00 lvm2 [297.88 GB / 32.00 MB free]
Total: 1 [297.88 GB] / in use: 1 [297.88 GB] / in no VG: 0 [0 ]
[root@tux ~]#

ACTIVE '/dev/VolGroup00/LogVol00' [292.50 GB] inherit
ACTIVE '/dev/VolGroup00/LogVol01' [5.34 GB] inherit

Avix
23rd December 2008, 03:27 AM
I'll give that a try Stevea. I ran some live cd stuff (Rip's partition recovery etc) and said the partition was there but empty

Avix
23rd December 2008, 06:32 PM
tried the different kernel, I am beginning to suspect what I did was delete the partition when I thought I was deleting a different partition on a different drive. so far no luck trying to recover the delete.

stevea
23rd December 2008, 07:33 PM
Sorry I'm late to reply.

No - boot a live Fedora CD and try mounting the LVMs.
Yes it is possible you blew away a partition (which is difficult to recover) but gove it a try.
I don't think that's the case though. Did you just add one of the disks recently ? This
problem could be as simple as telling grub the wrong disk ID (hd1,1) or whatever.


You just have a small 200MB ext3 (probably a /boot partition) on /dev/sda1 and the rest of the 250MB is unused.
ON /dev/sdb you have a similar 200MB /boot and a ~300GB LVm with two logical volumes.

You can test the goodness of the LVm like this.
Boot as before and as root do the pvscan.
Make directory mountpoints and mount like this ...
mkdir /tmp/test00
mkdir /tmp/test01
mount /dev/VolGroup00/LogVol00 /tmp/test00
mount /dev/VolGroup00/LogVol01 /tmp/test01

Then look at the directory contents
ls -l /tmp/test00
ls -l /tmp/test01


If the mount fails then maybe the file system or LVM is hosed.
If if mounts and you can see the files in both file systems then it's certainly recoverable.

If it's recoverable then I'll suggest you report what is in the boot/grub/grub.conf file on your disk.
That brings up the question ... id /dev/sda1 or /dev/sdb1 your /boot partition ?

I'm beginning to suspect you'll just need to edit the grub.conf file and specify the correct disk for root.
I suggest you stop the boot at the grub line (hit any char) then use the 'e' edit command to examine the boot params. Then you can match this to either /dev/sda1 or /dev/sdb1' (which should have a grub/grub.conf.

Report the grub.conf file contents. the root= and (hd1,0) type stuff is critical.

Avix
24th December 2008, 11:03 PM
ok, from the top what I was doing.

had my F8 install on the 250 gig drive. had installed F10 in the 350 to play with it and see how it was. did something to it that cacked it so it wouldn't boot, so I decided to just reformat it. used Gparted out of F8 to format what I thought was the 350gig drive. when I rebooted, thats when everything went to a not nice place.

Stevea
[root@tux ~]# mount /dev/VolGroup00/LogVol00 /tmp/test00
[root@tux ~]# mount /dev/VolGroup00/LogVol01 /tmp/test01
/dev/mapper/VolGroup00-LogVol01 looks like swapspace - not mounted
mount: you must specify the filesystem type

(thats with the 250 drive hooked into the F10 350 drive)

wimper....

neogranas
29th December 2008, 06:31 PM
I am having the same issue as Avix. Any suggestions? Err.. Just found that it's only loading /dev/mapper/VolGroup00-LogVol0x so it won't see my other drive. Anyone know how to set up the other VolGroup? At least, I think that is what is needed.

Avix
30th December 2008, 06:40 PM
nice to know I'm not the only one neograns, what were you doing with yours when it went all pear shaped?

neogranas
30th December 2008, 07:07 PM
I got a new hard drive for Christmas. I installed Fedora 10 on the new one while at work, plugged in the old one later hoping to pull data off it so I don't have to reconfig httpd.conf and various other things. But I got around it by taking the server offline, putting in an empty hard drive, moving the files, then putting my new one back in so I could pull it in from there. Very dirty and took a while but it worked. I needed the server up as soon as possible.

I'd still like to know the solution so I don't have to ever do what I did again.