- 1 LVM Basics
- 2 Rescanning scsi bus
- 3 Extending a logical volume
LVM allows a server filesystem to be created, extended and even moved while running, without the need for a reboot and without needing to unmount the filesystem.
Order of things
Firstly there are physical volumes, these are entire physical disks basically (not sure whether you can use partitions as physical volumes). Physical volumes are part of volume groups.
Volume groups group together physical volumes and you make logical volumes out of them, volume groups basically become 'extent' pools which are large (~4MB normally) blocks.
Logical volumes are the actual file-systems which are formatted as ext2/3/whatever. When you create or extend one you pick which volume group to use and how many 'extents' to use. The number of extents times the size of the extent becomes your file-system size.
So you can see from the drawing below the layers and how they relate to one another.
If we were to want to extend the LVM though it would look like this (green parts are the new ones and yellow parts are extended bits)
Rescanning scsi bus
When you hotplug a hard drive in a virtual machine, the drive will not show in fdisk -l output until you reboot your VM or rescan the SCSI bus.
The rescan will be triggered through sysfs pseudo filesystem.
# for BUS in `ls -1 /sys/class/scsi_host`; do echo "- - -" > /sys/class/scsi_host/$BUS/scan; done
Now, you should be able to see the drive when running:
# fdisk -l
Extending a logical volume
All the commands are along the same lines and use an abbreviated prefix and a full suffix. the prefixes are as follows:
- pv [Physical Volume]
- vg [Volume Group]
- lv [Logical Volume]
And the suffixes (not exhaustive):
The commands that you'll need for an extend operation are (in this order): fdisk, pvdisplay, pvcreate, vgdisplay, vgextend, lvdisplay, lvextend, resize2fs. Lets get started.
Here is an example of a extension that i needed to do on the backup-mirror server. I needed to add 60GB to the logical volume that was created when the system was installed. Unfortunately due to the licensing on the ESX server i couldn't hot-plug a new physical drive into the virtual machine so needed to reboot to do that part, but all the rest should be able to be done while it's running which is the general idea anyway.
Displaying what we have
Firstly we should should get an inventory of what we have in the system as far as physical disks are concerned.
[root@backup-mirror ~]# fdisk -l Disk /dev/sda: 32.2 GB, 32212254720 bytes 255 heads, 63 sectors/track, 3916 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/sda1 * 1 13 104391 83 Linux /dev/sda2 14 3916 31350847+ 8e Linux LVM Disk /dev/sdb: 64.4 GB, 64424509440 bytes 255 heads, 63 sectors/track, 7832 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk /dev/sdb doesn't contain a valid partition table
So we can see from that output that there is two phyiscal disks, one ~30GB and one ~60GB. This is what i expected to see. We should check what physical volumes we have allocated now:
[root@backup-mirror ~]# pvdisplay /dev/hdc: open failed: No medium found --- Physical volume --- PV Name /dev/sda2 VG Name VolGroup00 PV Size 29.90 GB / not usable 24.06 MB Allocatable yes (but full) PE Size (KByte) 32768 Total PE 956 Free PE 0 Allocated PE 956 PV UUID 4pBA3O-TeaQ-hP8J-Vpgy-MXYJ-zY8W-bp9sj2
So we can see that there is one physical volume and it isn't /dev/sdb so we should go ahead and create that now:
Creating the physical volume
[root@backup-mirror ~]# pvcreate /dev/sdb Physical volume "/dev/sdb" successfully created
Now double check with pvdisplay again:
[root@backup-mirror ~]# pvdisplay /dev/hdc: open failed: No medium found --- Physical volume --- PV Name /dev/sda2 VG Name VolGroup00 PV Size 29.90 GB / not usable 24.06 MB Allocatable yes (but full) PE Size (KByte) 32768 Total PE 956 Free PE 0 Allocated PE 956 PV UUID 4pBA3O-TeaQ-hP8J-Vpgy-MXYJ-zY8W-bp9sj2 --- Physical volume --- PV Name /dev/sdb VG Name VolGroup00 PV Size 60.00 GB / not usable 32.00 MB Allocatable yes PE Size (KByte) 32768 Total PE 1919 Free PE 1919 Allocated PE 0 PV UUID OZDal1-EKnd-kqQZ-YiVr-CQ6f-XRwr-Tm0iFe
You can see from the output the 'Total PE', 'Free PE' and 'Allocated PE' sections show how many physical extents are where.
Now we should see what logical volumes we have:
Checking volume groups
[root@backup-mirror ~]# vgdisplay /dev/hdc: open failed: No medium found --- 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 29.88 GB PE Size 32.00 MB Total PE 956 Alloc PE / Size 956 / 29.88 GB Free PE / Size 0 / 0 VG UUID MpnZFf-2Gq1-nUg6-wC42-zlDL-0GzS-k1QSqC
We can now extend our volume group with the following command:
Extending Volume Groups
[root@backup-mirror ~]# vgextend VolGroup00 /dev/sdb /dev/hdc: open failed: No medium found Volume group "VolGroup00" successfully extended
Don't worry about the error message about /dev/hdc. Its just complaining about the cdrom drive not being available. We should now check that the volume group has been extended properly:
[root@backup-mirror ~]# vgdisplay /dev/hdc: open failed: No medium found --- 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 89.84 GB PE Size 32.00 MB Total PE 2875 Alloc PE / Size 956 / 29.88 GB Free PE / Size 1919 / 59.97 GB VG UUID MpnZFf-2Gq1-nUg6-wC42-zlDL-0GzS-k1QSqC
Looks good, now lets check the logical volumes:
Checking logical volumes
[root@backup-mirror ~]# lvdisplay /dev/hdc: open failed: No medium found --- Logical volume --- LV Name /dev/VolGroup00/LogVol00 VG Name VolGroup00 LV UUID k1jp0v-yom4-N37J-SKWT-6SgH-tEjG-KFxr5t LV Write Access read/write LV Status available # open 1 LV Size 28.88 GB Current LE 924 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 KdnN1i-JyRt-LXdl-8j03-msZ7-uDfI-yvVESA 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
When we're extending the logical volume we need to make sure that we extend the right one, check the 'LV Size' of each to make sure and also check the 'mount' output or cat /etc/fstab.
[root@backup-mirror ~]# cat /etc/fstab /dev/VolGroup00/LogVol00 / ext3 defaults 1 1 LABEL=/boot /boot ext3 defaults 1 2 tmpfs /dev/shm tmpfs defaults 0 0 devpts /dev/pts devpts gid=5,mode=620 0 0 sysfs /sys sysfs defaults 0 0 proc /proc proc defaults 0 0 /dev/VolGroup00/LogVol01 swap swap defaults 0 0
We can see from this output that the filesystem mounted on '/' (which is the one we want) is /dev/VolGroup00/LogVol00 which is the logical volume that we want to extend.
Extending logical volumes
You'll need to decide by how many 'PE's you want to extend the logical volume by, normally you'd just use all the free 'PE's in the volume group:
[root@backup-mirror ~]# lvextend -l +1919 /dev/VolGroup00/LogVol00 /dev/hdc: open failed: No medium found Extending logical volume LogVol00 to 88.84 GB Logical volume LogVol00 successfully resized
The command above extends the logical volume /dev/VolGroup00/LogVol00 by 1919 physical extents.
Extending the filesystem
The last thing we need to do is extend the actual filesystem on the logical volume with the following command (this will take some time):
[root@backup-mirror ~]# resize2fs /dev/VolGroup00/LogVol00 resize2fs 1.39 (29-May-2006) Filesystem at /dev/VolGroup00/LogVol00 is mounted on /; on-line resizing require d Performing an on-line resize of /dev/VolGroup00/LogVol00 to 23289856 (4k) blocks . The filesystem on /dev/VolGroup00/LogVol00 is now 23289856 blocks long.