Linux LVM at Disk Image

19 August 2014

When you use some virtualization platform (KVM, Xen, etc.), then it is useful to know how to modify content at disk image of the virtualized Linux. When disk images contains partitions formated with some file system format, then it is simple and straightforward, but when virtualized Linux uses LVM, then it becomes little bit tricky. This tutorial will guide you through this process.

First of all it is necessary to stop virtual machine.

Using Loop Device

There are several ways how to modify content on the disk image, but we will use loop device. We pretend that your image is in /var/lib/libvirt/images. Thus we will switch to this directory:

cd /var/lib/libvirt/images

When your image disk is in tester1.img, then we will create loop device using following command:

losetup /dev/loop0 tester1.img

Be aware that /dev/loop0 is not used by other image.

Command kpartx

Linux installation usually has several partitions and we will map them to other loop devices:

kpartx -a /dev/loop0

Note: we can use kpartx directly at file with disk image (tester1.img).

This will map partitions from disk image to following loop devices:

/dev/loop0p1 (/dev/mapper/loop0p1)
/dev/loop0p2 (/dev/mapper/loop0p2)

We can investigate meaning of partitions using fdisk command:

fdisk /dev/loop0

and then type p:

Command (m for help): p

Disk /dev/loop0: 21.5 GB, 21474836480 bytes
255 heads, 63 sectors/track, 2610 cylinders, total 41943040 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00082a8c

      Device Boot      Start         End      Blocks   Id  System
/dev/loop0p1   *        2048     1026047      512000   83  Linux
/dev/loop0p2         1026048    41943039    20458496   8e  Linux LVM


We can see that device /dev/loop0p2 is partition with LVM. This LVM contains partitions, where we want to modify some files (e.g.: /etc/passwd and /etc/shadow ;-)). First of all we need to know what is the name of volume group on this partition. We will use pvscan for this purpose:


The output of pvscan should contain something like this:

PV /dev/mapper/loop0p2   VG vg_tester1   lvm2 [19.51 GiB / 0    free]

We know now that volume group has name vg_tester1 and we will use it for mapping its logical volumes to loop devices:

vgchange -a y vg_tester1

We can see now that there are two new files in /dev/mapper:


We can mount the block device /dev/mapper/vg_tester1-lv_root:

mkdir /mnt/tester1_root
mount /dev/mapper/vg_tester1-lv_root /mnt/tester1_root

Finally we can edit any file in /mnt/tester1_root.

comments powered by Disqus