Recovering SME Server with lvm drives
Recovering SME Server with lvm drives
The purpose of this howto is to give to you the abilities to access to your data if the SME Server 8 is broken and can't start in a normal way. you have several methods below, but all are done on a default Raid over LVM, therefore you might need to adapt to your configuration, if necessary.
If your issue concerns a grub issue, you should look to this wiki page
Method with the official SME Server CDROM
I presume that your SME Server is on a RAID over LVM, otherwise you will have to adapt this HOWTO.
- start the system with your official SME Server CDROM
- give at prompt : linux rescue
- set your language and your keyboard
- set to no the start of network interfaces
- set to continue the question about how the system is mounted in /mnt/sysimage
- set to ok
- give at prompt :
chroot /mnt/sysimage su -
Now you have successfully mounted your LVM and you are able to read your data in a chroot environment, you can save them on a usb disk
- to exit
exit exit halt
Method with SystemRescueCd
start the system with your system rescue cd or you usb stick, choose your keyboard settings
then start the server X
startx
open a terminal to verify if your raid is initiated.
cat /proc/mdstat
if you are lucky the output will look like this
# cat /proc/mdstat
Personalities : [raid1]
md99 : active raid1 sdb1[1] sda1[0]
104320 blocks [2/2] [UU]
md100 : active raid1 sdb2[1] sda2[0]
262036096 blocks [2/2] [UU]
unused devices: <none>
so we need to launch the LVM
vgchange -ay
afterward if the LVM is launched without error messages, we can mount the LVM in /mnt
mkdir /mnt/recover mount /dev/main/root /mnt/recover
lvdisplay
Now you have successfully mounted your LVM and you are able to read your data on /mnt/recover, you can save them on a usb disk with the file browser for example.
Method
Let’s try starting the raid and see what we get:
mdadm -E /dev/sdb1
What “mdadm -E /dev/sdb1” command shows it is a part of a raid array, what level, how many members, etc.
user@user-desktop:/mnt$ mdadm -E /dev/sdb2
/dev/sdb2:
Magic : a92b4efc
Version : 00.90.00
UUID : 550e0406:c9ce50d2:825b32e4:4a9d3549
Creation Time : Sat Sep 8 12:15:29 2007
Raid Level : raid1
Used Dev Size : 1991936 (1945.58 MiB 2039.74 MB)
Array Size : 1991936 (1945.58 MiB 2039.74 MB)
Raid Devices : 2
Total Devices : 1
Preferred Minor : 2
Update Time : Sat Sep 8 12:22:05 2007
State : clean
Active Devices : 1
Working Devices : 1
Failed Devices : 1
Spare Devices : 0
Checksum : 22e3837f - correct
Events : 0.991
Number Major Minor RaidDevice State
this 0 8 2 0 active sync /dev/sda2
0 0 8 2 0 active sync /dev/sda2
1 1 0 0 1 faulty removed
user@user -desktop:/mnt$
With it being a raid 1 we only need 1 member to start it.
You can also use any md device to assemble the array. You need to make sure you are using an md device that isn't already in use, to check what isn’t being used type:
cat /proc/mdstat
So we have now found which md device we can use and for our example we will use “md8”
What we will do now is assemble and run the array:
mdadm -AR /dev/md8 /dev/sdb2
If you are running other then raid1, you may need to include additional members from other drives:
mdadm -AR /dev/md8 /dev/sdb2 /dev/sdd2 /dev/sde3
Now see if the array is assembled:
cat /proc/mdstat
See if it detects the physical volumes:
user@user-desktop:~$ pvs PV VG Fmt Attr PSize PFree /dev/md8 main lvm2 a- 1.88G 32.00M user@user-desktop:~$
To activate all known volume groups in the system:
user@user-desktop:~$ vgchange main -a n 0 logical volume(s) in volume group "main" now active user@user-desktop:~$ vgchange main -a y 2 logical volume(s) in volume group "main" now active user@user-desktop:~$
Now we should be able to mount the drive:
user@user-desktop:~$ mount /dev/main/root /mnt/oldsmeserver/ user@user-desktop:~$
Looking good so let’s show where our files are:
user@user-desktop:~$ cd /mnt/oldsmeserver/ user@user-desktop:/mnt/oldsmeserver$ dir aquota.group boot etc lib mnt proc selinux sys var aquota.user command home lost+found opt root service tmp bin dev initrd media package sbin srv usr user@user-desktop:/mnt/oldsmeserver$
Now you have successfully assembled your array and able to recover your data.
Notes:
- If the existing system has lvs already installed and has a volume group called "main" there may be issues.
- If you installed SME Server <7.0 your volume group will be different, to find out your volume group type:
user@user-desktop:~$ vgdisplay --- Volume group --- VG Name main The Volume group name. System ID Format lvm2 [..] user@user-desktop:~$
Method with a ubuntu Cdrom
based on http://www.linuxjournal.com/article/8874?page=0,0
on ubuntu (non lvm), install mdadm and lvm2, attach server drive,
find UUID's
$ sudo mdadm --examine --scan /dev/sdb1 /dev/sdb2 ARRAY /dev/md2 level=raid1 num-devices=2 UUID=895293be:9cfa7672:f1761508:386417bc ARRAY /dev/md1 level=raid1 num-devices=2 UUID=10573599:841f46aa:4c068816:67364324
add ARRAY lines to mdadm.conf
$ cat /etc/mdadm/mdadm.conf # mdadm.conf # # Please refer to mdadm.conf(5) for information about this file. # # by default, scan all partitions (/proc/partitions) for MD superblocks. # alternatively, specify devices to scan, using wildcards if desired. DEVICE partitions ARRAY /dev/md2 level=raid1 num-devices=2 UUID=895293be:9cfa7672:f1761508:386417bc ARRAY /dev/md1 level=raid1 num-devices=2 UUID=10573599:841f46aa:4c068816:67364324
checking
$ sudo pvscan PV /dev/md2 VG main lvm2 [148.94 GB / 64.00 MB free] Total: 1 [148.94 GB] / in use: 1 [148.94 GB] / in no VG: 0 [0 ]
checking
$ sudo lvscan ACTIVE '/dev/main/root' [146.94 GB] inherit ACTIVE '/dev/main/swap' [1.94 GB] inherit
mount, check and copy to safe location ...
$ sudo mkdir /mnt/ga $ sudo mount /dev/main/root /mnt/ga $ sudo ls -la /mnt/ga/var/log/messages lrwxrwxrwx 1 root root 32 2010-03-24 18:00 /mnt/ga/var/log/messages -> /var/log/messages.20*