In this small tutorial we will go over building a basic mirror and migrate that mirror to a new system if need be. This was done using Alpine Linux v3.10 but should be transferrable with any system using LVM >=v2.0.
Useful information was found from tldp.org
Initial system has 3 hard drives (/dev/vd{a,b,c}). The system will be configured using the Logic Volume Manager (LVM) as such,
Your system may report drives as hdX or sdX, my example uses vdX
The main OS will be installed on vda while the mirror will be built using vdb and vdc.
Install your flavour of linux on vda. Once installed, login as root and
format vdb and vdc to use the same partition layout and sizes using the
fdisk command. Issue the following commands (or use sudo if desired)
# fdisk -l /dev/vdb
Disk /dev/vdb: 100 MB, 104857600 bytes, 204800 sectors
100 cylinders, 64 heads, 32 sectors/track
Units: sectors of 1 * 512 = 512 byte
Command (m for help): n
Partition type
p primary partition (1-4)
e extended
p
Partition number (1-4):
Value is out of range
Partition number (1-4): 1
First sector (32-204799, default 32):
Using default value 32
Last sector or +size{,K,M,G,T} (32-204799, default 204799):
Using default value 204799
Command (m for help):
Selected partition 1
Hex code (type L to list codes): 8e
Changed system type of partition 1 to 8e (Linux LVM)
Command (m for help): w
The partition table has been altered.
Calling ioctl() to re-read partition table
#
Do the same steps for the second hardrive. In the end you will have two drives with one partition each, e.g., /dev/vdb1 and /dev/vdc1.
It is important to note that for this mirror both hard drives should be the same size. In this example they are 100 MB.
Now that we have extra drives in the system we want to create a mirror but in order to do that we need to tell LVM what physical drives exist to be managed as physical volumes,
# pvcreate /dev/vdb1 /dev/vdc1
Physical volume "/dev/vdb1" successfully created.
Physical volume "/dev/vdc1" successfully created.
Now you can now use the physical volumes to create a volume group. A volume group manages the data of the physical volumes depending on how it is configured. To create a volume group do the following,
# vgcreate data /dev/vdb1 /dev/vdc1
Volume group "data" successfully created
# vgdisplay
--- Volume group ---
VG Name data
System ID
Format lvm2
Metadata Areas 2
Metadata Sequence No 3
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 1
Open LV 1
Max PV 0
Cur PV 2
Act PV 2
VG Size 192.00 MiB
PE Size 4.00 MiB
Total PE 48
Alloc PE / Size 46 / 184.00 MiB
Free PE / Size 2 / 8.00 MiB
VG UUID s5LJNJ-6fsZ-CXcL-FZcU-hozf-cTfl-cPjKkW
Now that a volume group exists a mirrored logical volume can be created. I will try to use all of the available space shared between two drives (192 MiB/2 = 86 MiB),
# lvcreate -L 86M -m1 -n datamirror data
Rounding up size to full physical extent 88.00 MiB
Logical volume "datamirror" created.
# lvs
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Conver
datamirror data rwi-aor--- 88.00m 100.00
lv_root vg0 -wi-ao---- <3.65g
lv_swap vg0 -wi-ao---- 1.25g
Checking what the logical volume information can be done with lvdisplay,
# lvdisplay data
--- Logical volume ---
LV Path /dev/data/datamirror
LV Name datamirror
VG Name data
LV UUID kXK2nP-j9Yz-fUjU-WB0B-B5qN-qTN1-WqKBlw
LV Write Access read/write
LV Creation host, time alpine, 2019-12-17 14:09:16 +0000
LV Status available
# open 1
LV Size 88.00 MiB
Current LE 22
Mirrored volumes 2
Segments 1
Allocation inheri
Read ahead sectors auto
- currently set to 256
Block device 252:6
Now that we have a mirror /dev/data/datamirror we can create a filesystem. In this case I used the ext4 filesystem
# mkfs.ext4 /dev/data/datamirror
mke2fs 1.45.2 (27-May-2019)
Creating filesystem with 90112 1k blocks and 22528 inodes
Filesystem UUID: a1c0d6b1-dbfa-4286-a8e6-3ab85d2c5f7f
Superblock backups stored on blocks:
8193, 24577, 40961, 57345, 73729
Allocating group tables: done
Writing inode tables: done
Creating journal (4096 blocks): done
Writing superblocks and filesystem accounting information: done
From here you can mount it manually with
# mount /dev/data/datamirror /mnt
Or add it to /etc/fstab to do as you see fit.
Now that we have a system let us re-add the system back to a new installation. Before we do let us grap the information of the old LVM setup
# lvs
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Conver
datamirror data rwi-aor--- 88.00m 100.00
lv_root vg0 -wi-ao---- <3.65g
lv_swap vg0 -wi-ao---- 1.25g
# vgs
VG #PV #LV #SN Attr VSize VFree
data 2 1 0 wz--n- 192.00m 8.00m
vg0 1 2 0 wz--n- <4.90g 0
# pvs
PV VG Fmt Attr PSize PFree
/dev/vda2 vg0 lvm2 a-- <4.90g 0
/dev/vdb1 data lvm2 a-- 96.00m 4.00m
/dev/vdc1 data lvm2 a-- 96.00m 4.00m
Luckily all we need to do deactivate and export the volume group data. The logical volume mirror we created will be preserved!
# umount /mnt
# vgchange --activate n data
0 logical volume(s) in volume group "data" now active
# vgexport data
Volume group "data" successfully exported
Now power down the system and migrate the physcial drives to the new system. On the new system we can scan for physical volumes as they are still preserved on the drives,
Once the drives are put into the new system check to see if they are detected,
# pvscan
PV /dev/vdb1 is in exported VG data [96.00 MiB / 4.00 MiB free]
PV /dev/vdc1 is in exported VG data [96.00 MiB / 4.00 MiB free]
PV /dev/vda2 VG vg0 lvm2 [<4.90 GiB / 0 free]
Total: 3 [<5.09 GiB] / in use: 3 [<5.09 GiB] / in no VG: 0 [0 ]
If pvscan detects the drives and that they are exported volume groups
successfully, we can then import them into the new system and activate them,
# vgimport data
Volume group "data" successfully imported
# vgchange --activate y data
1 logical volume(s) in volume group "data" now active
Now that the volume group data is merged in an activated we can mount as before or use /etc/fstab
# mount /dev/data/datamirror /mnt
# ls
doc lost+found src
data fig script
© 2017–2022 David Kalliecharan