1 Logical Volume Manager (LVM)
2 ----------------------------
7 Most people install {pve} directly on a local disk. The {pve}
8 installation CD offers several options for local disk management, and
9 the current default setup uses LVM. The installer let you select a
10 single disk for such setup, and uses that disk as physical volume for
11 the **V**olume **G**roup (VG) `pve`. The following output is from a
12 test installation using a small 8GB disk:
16 PV VG Fmt Attr PSize PFree
17 /dev/sda3 pve lvm2 a-- 7.87g 876.00m
20 VG #PV #LV #SN Attr VSize VFree
21 pve 1 3 0 wz--n- 7.87g 876.00m
24 The installer allocates three **L**ogical **V**olumes (LV) inside this
29 LV VG Attr LSize Pool Origin Data% Meta%
30 data pve twi-a-tz-- 4.38g 0.00 0.63
31 root pve -wi-ao---- 1.75g
32 swap pve -wi-ao---- 896.00m
35 root:: Formatted as `ext4`, and contains the operation system.
39 data:: This volume uses LVM-thin, and is used to store VM
40 images. LVM-thin is preferable for this task, because it offers
41 efficient support for snapshots and clones.
43 For {pve} versions up to 4.1, the installer creates a standard logical
44 volume called ``data'', which is mounted at `/var/lib/vz`.
46 Starting from version 4.2, the logical volume ``data'' is a LVM-thin pool,
47 used to store block based guest images, and `/var/lib/vz` is simply a
48 directory on the root file system.
53 We highly recommend to use a hardware RAID controller (with BBU) for
54 such setups. This increases performance, provides redundancy, and make
55 disk replacements easier (hot-pluggable).
57 LVM itself does not need any special hardware, and memory requirements
64 We install two boot loaders by default. The first partition contains
65 the standard GRUB boot loader. The second partition is an **E**FI **S**ystem
66 **P**artition (ESP), which makes it possible to boot on EFI systems.
69 Creating a Volume Group
70 ~~~~~~~~~~~~~~~~~~~~~~~
72 Let's assume we have an empty disk `/dev/sdb`, onto which we want to
73 create a volume group named ``vmdata''.
75 CAUTION: Please note that the following commands will destroy all
76 existing data on `/dev/sdb`.
78 First create a partition.
80 # sgdisk -N 1 /dev/sdb
83 Create a **P**hysical **V**olume (PV) without confirmation and 250K
86 # pvcreate --metadatasize 250k -y -ff /dev/sdb1
89 Create a volume group named ``vmdata'' on `/dev/sdb1`
91 # vgcreate vmdata /dev/sdb1
94 Creating an extra LV for `/var/lib/vz`
95 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
97 This can be easily done by creating a new thin LV.
99 # lvcreate -n <Name> -V <Size[M,G,T]> <VG>/<LVThin_pool>
101 A real world example:
103 # lvcreate -n vz -V 10G pve/data
105 Now a filesystem must be created on the LV.
107 # mkfs.ext4 /dev/pve/vz
109 At last this has to be mounted.
111 WARNING: be sure that `/var/lib/vz` is empty. On a default
112 installation it's not.
114 To make it always accessible add the following line in `/etc/fstab`.
116 # echo '/dev/pve/vz /var/lib/vz ext4 defaults 0 2' >> /etc/fstab
119 Resizing the thin pool
120 ~~~~~~~~~~~~~~~~~~~~~~
122 Resize the LV and the metadata pool can be achieved with the following
125 # lvresize --size +<size[\M,G,T]> --poolmetadatasize +<size[\M,G]> <VG>/<LVThin_pool>
127 NOTE: When extending the data pool, the metadata pool must also be
131 Create a LVM-thin pool
132 ~~~~~~~~~~~~~~~~~~~~~~
134 A thin pool has to be created on top of a volume group.
135 How to create a volume group see Section LVM.
137 # lvcreate -L 80G -T -n vmstore vmdata