Logical Volume Expansion With Linux LVM

Sometimes we just run out of space in filesystems. Whether due to lack of understanding of application storage needs, poor planning or we just need extra space for a development effort, we find it necessary to add disk. Virtualization gives us the ability to quickly provision new virtual disks that behave identically to physical ones. This, in conjunction with the flexibility of Linux Logical Volume Management (LVM), we can easily grow volume groups, logical volumes and filesystems. It also makes it easy to practice the techniques before provisioning additional storage in customer production environments. So let’s add a disk and grow the root filesystem in a Sun VirtualBox / Fedora 10 environment.

I’m using a Fedora 10 VirtualBox guest with a roughly 9GB virtual disk. I will add a 2GB disk to the guest, boot it and then do the following:

Add a virtual disk using the VirtualBox Media Manager

The new disk is added as primary slave and we power up the guest machine. When the guest is up, look for the disk.

[root@F10 ~]# fdisk -l

Disk /dev/sda: 10.7 GB, 10737418240 bytes
255 heads, 63 sectors/track, 1305 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0×00048601

Device Boot Start End Blocks Id System
/dev/sda1 * 1 25 200781 83 Linux
/dev/sda2 26 1305 10281600 8e Linux LVM

Disk /dev/sdb: 2147 MB, 2147483648 bytes
255 heads, 63 sectors/track, 261 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0×00000000

Disk /dev/sdb doesn’t contain a valid partition table

We see the 2GB disk, /dev/sdb. Notice that there’s no partition table on it. Let’s look at the filesystem usage before extending the logical volume.

[root@F10 ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
8.7G 2.1G 6.2G 25% /
/dev/sda1 190M 14M 167M 8% /boot
tmpfs 252M 100K 251M 1% /dev/shm

Initialize this new disk as an LVM physical volume

[root@F10 ~]# pvcreate /dev/sdb
Physical volume “/dev/sdb” successfully created

View the state of LVM physical volumes

[root@F10 ~]# pvscan
PV /dev/sda2 VG VolGroup00 lvm2 [9.78 GB / 32.00 MB free]
PV /dev/sdb lvm2 [2.00 GB]
Total: 2 [11.78 GB] / in use: 1 [9.78 GB] / in no VG: 1 [2.00 GB]

View the current volume group state

[root@F10 ~]# vgdisplay
— Volume group —
VG Name VolGroup00
System ID
Format lvm2
Metadata Areas 1
Metadata Sequence No 3
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 2
Open LV 2
Max PV 0
Cur PV 1
Act PV 1
VG Size 9.78 GB
PE Size 32.00 MB
Total PE 313
Alloc PE / Size 312 / 9.75 GB
Free PE / Size 1 / 32.00 MB
VG UUID KskOHW-jogJ-ikWo-qgN6-KEdW-bQXd-jfWHD0

Extend VolGroup00 using the new physical volume /dev/sdb

[root@F10 ~]# vgextend VolGroup00 /dev/sdb
Volume group “VolGroup00″ successfully extended

View updated volume group status

[root@F10 ~]# vgdisplay
— Volume group —
VG Name VolGroup00
System ID
Format lvm2
Metadata Areas 2
Metadata Sequence No 4
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 2
Open LV 2
Max PV 0
Cur PV 2
Act PV 2
VG Size 11.75 GB
PE Size 32.00 MB
Total PE 376
Alloc PE / Size 312 / 9.75 GB
Free PE / Size 64 / 2.00 GB
VG UUID KskOHW-jogJ-ikWo-qgN6-KEdW-bQXd-jfWHD0

Notice that we now have 64 free extents. So now we need to determine the logical volume that is mounted at ‘/’

[root@F10 ~]# lvdisplay
— Logical volume —
LV Name /dev/VolGroup00/LogVol00
VG Name VolGroup00
LV UUID 9RB7rj-t2kl-3pMa-ssDt-LxU5-woiB-ML2czB
LV Write Access read/write
LV Status available
# open 1
LV Size 8.75 GB
Current LE 280
Segments 1
Allocation inherit
Read ahead sectors auto
– currently set to 256
Block device 253:0

— Logical volume —
LV Name /dev/VolGroup00/LogVol01
VG Name VolGroup00
LV UUID q2INDo-Me5O-G2UB-TKl8-bGLF-bzmG-lNdhhT
LV Write Access read/write
LV Status available
# open 1
LV Size 1.00 GB
Current LE 32
Segments 1
Allocation inherit
Read ahead sectors auto
– currently set to 256
Block device 253:1

We need to extend LogVol00, since it’s the 8.7GB volume

Extend the logical volume by the number of free extents (64)

[root@F10 ~]# lvextend -l +64 /dev/VolGroup00/LogVol00
Extending logical volume LogVol00 to 10.75 GB
Logical volume LogVol00 successfully resized

Resize the filesystem on the logical volume LogVol00

The root filesystem ‘/’ is mounted on this logical volume, but with the ext3 online resize capability in the kernel, there’s no need to unmount it.

[root@F10 ~]# resize2fs /dev/VolGroup00/LogVol00
resize2fs 1.41.4 (27-Jan-2009)
Filesystem at /dev/VolGroup00/LogVol00 is mounted on /; on-line resizing required
old desc_blocks = 1, new_desc_blocks = 1
Performing an on-line resize of /dev/VolGroup00/LogVol00 to 2818048 (4k) blocks.
The filesystem on /dev/VolGroup00/LogVol00 is now 2818048 blocks long.

Let’s see what the root filesystem looks like now

[root@F10 ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
11G 2.1G 8.1G 21% /
/dev/sda1 190M 14M 167M 8% /boot
tmpfs 252M 100K 251M 1% /dev/shm

That was pretty straight forward, and the procedure is nearly identical when provisioning SAN LUNs to Linux hosts for LVM expansion. The difference being the method used to expose new block devices to the Linux host. So before you do that, practice in a virtual machine!

Further reading:
LVM-HOWTO