The Basics

list of component:

  • PV (Physical Volume)
  • VG (Volume Group)
  • LV (Logical Volume)
  • PE (Physical Extend)
  • LE (Logical Extend)
  • FS (File Sytem)

LVM2 use a new driver, the device-mapper allow the us of diskยดs sectors in different targets: - linear (most used in LVM). - stripped (stripped on several disks) - error (all I/O are consider in errors) - snapshot (allow snapshot async)

  • mirror (integrate elements usefull for pvmove commande)
  • below example show you a striped volume and linear volume
lvs --all --segments -o +devices
server_xplore_col1   vgdata -wi-ao----   21 striped   1.07t /dev/md2(40229),/dev/md3(40229),/dev/md4(40229),/dev/md5(40229),โ€ฆ
server_xplore_col2   vgdata -wi-ao----    1 linear  219.87g /dev/md48(0)

Basic checks

# Summary 
pvs
vgs
lvs

# Scanner
pvscan
vgscan
lvscan

# Details info
pvdisplay   [sda]
pvdisplay   -m /dev/emcpowerd1 
vgdisplay   [vg_root]
lvdisplay   [/dev/vg_root/lv_usr]

# Summary details
lvmdiskscan
  /dev/sda1 [     600.00 MiB]
  /dev/sda2 [       1.00 GiB]
  /dev/sda3 [      38.30 GiB] LVM physical volume
  /dev/sdb1 [    <100.00 GiB] LVM physical volume
  /dev/sdc1 [     <50.00 GiB] LVM physical volume
  /dev/sdj  [      20.00 GiB]
  1 disk
  2 partitions
  0 LVM physical volume whole disks
  3 LVM physical volumes

Usual Scenario in LVM

  • Extend an existing LVM filesystem:
parted /dev/sda resizepart 3 100%
udevadm settle
pvresize /dev/sda3

# Extend a XFS to a fixe size 
lvextend -L 30G /dev/vg00/var
xfs_growfs /dev/vg00/var  

# Add some space to a ext4 FS
lvextend -L +10G /dev/vg00/var
resize2fs /dev/vg00/var

# Extend to a pourcentage and resize automaticly whatever is the FS type.
lvextend -l +100%FREE /dev/vg00/var -r
  • Create a new LVM filesystem:
parted /dev/sdb mklabel gpt mkpart primary 1 100% set 1 lvm on
udevadm settle
pvcreate /dev/sdb1
vgcreate vg01 /dev/sdb1
lvcreate -n lv_data -l 100%FREE  vg01

# Create a XFS
mkfs.xfs /dev/vg01/lv_data
mkdir /data
echo "/dev/mapper/vg01-lv_data   /data                  xfs     defaults        0 0" >>  /etc/fstab 
mount -a 

# Create an ext4
mkfs.ext4 /dev/vg01/lv_data
mkdir /data
echo "/dev/mapper/vg01-lv_data   /data                  ext4     defaults        0 0" >>  /etc/fstab 
mount -a
  • Remove SWAP:
swapoff -v /dev/dm-1
lvremove /dev/vg00/swap
vi /etc/fstab
vi /etc/default/grub
grub2-mkconfig -o /boot/efi/EFI/redhat/grub.cfg
grubby --remove-args "rd.lvm.lv=vg00/swap" --update-kernel /boot/vmlinuz-3.10.0-1160.71.1.el7.x86_64
grubby --remove-args "rd.lvm.lv=vg00swap" --update-kernel /boot/vmlinuz-3.10.0-1160.el7.x86_64
grubby --remove-args "rd.lvm.lv=vg00/swap" --update-kernel /boot/vmlinuz-0-rescue-cd2525c8417d4f798a7e6c371121ef34
echo "vm.swappiness = 0" >> /etc/sysctl.conf
sysctl -p
  • Move data form disk to another:
# #n case of crash, just relaunch pvmove without arguments
pvmove /dev/emcpowerd1 /dev/emcpowerc1

# Remove PV from a VG
vgreduce /dev/emcpowerd1 vg01

# Remove all unused PV from VG01
vgreduce -a vg01

# remove all PV
pvremove /dev/emcpowerd1
  • mount /var even if doesn’t want:
lvchange -ay --ignorelockingfailure --sysinit vgroot/var
  • Renaming:
# VG rename
vgrename 

# LV rename
lvrename

# PV does not need to be rename

LVM on partition VS on Raw Disk

Even if in the past I was using partition MS-DOS disklabel or GPT disklabel for PV, I prefer now to use directly LVM on the main block device. There is no reason to use 2 disklabels, unless you have a very specific use case (like disk with boot sector and boot partition).

The advantage of having LVM directly are: * simplicity - you do not need to use 2 sets of tools * flexibility - you can use pvmove to move the data from one disk volume to another without downtime, you can use snapshot and thin provisioning * you do not need to run partprobe or kpartx to tell the kernel that you created/resized/deleted a volume. And partprobe / `kpartx`` could fail if partitions are in use. * maybe better performance, compared to using LVM on top of MS-DOS or GPT disklables]