Manage logical storage with LVM

LVM (Logical Volume Manager) decouples logical volumes from physical disks: you grow a volume online, add a disk on the fly or take a snapshot, without reformatting anything. This guide is the entry point of the LVM path: it shows how to build the PV/VG/LV stack, inspect it and grow a volume online with lvextend -r (the number one use case in production). Advanced operations (thin provisioning, snapshots, disk migration) are covered in dedicated guides, presented before the troubleshooting section. For beginner to intermediate administrators, on Debian/Ubuntu and RHEL. All outputs come from an AlmaLinux test.

What you will learn

• Understand the PV / VG / LV model
• Build the stack with pvcreate, vgcreate, lvcreate
• Grow a logical volume online with lvextend -r
• Create and use a snapshot

Prerequisites

• A free disk or LVM-type partition
• The lvm2 package:

Debian / Ubuntu

sudo apt install lvm2

RHEL / AlmaLinux

sudo dnf install lvm2

The PV / VG / LV model

LVM stacks three layers. Understanding this stacking is understanding LVM:

• PV (Physical Volume): a disk or a partition initialised for LVM (pvcreate).
• VG (Volume Group): a pool of space that aggregates one or more PVs (vgcreate).
• LV (Logical Volume): a slice carved out of the VG, which you format and mount like a partition (lvcreate).

Source · SVG

Esc to close, + and − to zoom, scroll to move around. 100 % − Fit + Close

The benefit: an LV is no longer tied to a specific disk. When the VG runs short of space, you add a disk to it; when an LV fills up, you grow it by drawing from the VG. Without reformatting, without unmounting.

Build the LVM stack

Let us start from a 10 GB /dev/sda1 partition dedicated to LVM. You initialise the PV, create the VG, then a 4 GB LV.

sudo pvcreate /dev/sda1
sudo vgcreate vgdata /dev/sda1
sudo lvcreate -L 4G -n lvdata vgdata

  Physical volume "/dev/sda1" successfully created.
  Volume group "vgdata" successfully created
  Logical volume "lvdata" created.

The logical volume is reachable under two equivalent paths: /dev/vgdata/lvdata and /dev/mapper/vgdata-lvdata. Format and mount it like any volume:

sudo mkfs.xfs /dev/vgdata/lvdata
sudo mkdir -p /data
sudo mount /dev/vgdata/lvdata /data

For persistence, add the entry to /etc/fstab (by UUID through blkid, see the mount guide). LVM's /dev/mapper/... paths are stable, but the UUID remains the safest convention.

Inspect the state: pvs, vgs, lvs

Three concise commands give a summary view of each layer. They are the ones to know day to day (the pvdisplay, vgdisplay, lvdisplay variants go into more detail).

sudo pvs
sudo vgs
sudo lvs

  PV         VG     Fmt  Attr PSize   PFree
  /dev/sda1  vgdata lvm2 a--  <10.00g <6.00g

  VG     #PV #LV #SN Attr   VSize   VFree
  vgdata   1   1   0 wz--n- <10.00g <6.00g

  LV     VG     Attr       LSize Pool Origin Data% ...
  lvdata vgdata -wi-ao---- 4.00g

Reading the essentials: the vgdata VG totals 10 GB of which 6 GB are free (VFree), and the lvdata LV uses 4 GB. These 6 free GB are the headroom available to grow.

Grow a volume online: lvextend -r

This is LVM's reason for being. When /data fills up, you extend the logical volume and its file system in a single command, without unmounting, thanks to the -r option (resize filesystem):

sudo lvextend -r -L +2G /dev/vgdata/lvdata

  Size of logical volume vgdata/lvdata changed from 4.00 GiB (1024 extents) to 6.00 GiB (1536 extents).
  Extending file system xfs to 6.00 GiB on vgdata/lvdata...
  data blocks changed from 1048576 to 1572864
  Logical volume vgdata/lvdata successfully resized.

LVM detects the file system (here XFS) and runs xfs_growfs automatically.

Do not worry if your output is much longer than the one above: lvextend -r is very verbose. On AlmaLinux 10 with lvm2 2.03.36, it produces 21 lines, including the File system xfs found on … mounted at … line, the command it is about to run (xfs_growfs /dev/vgdata/lvdata) and the full geometry returned by it. The student comparing with the guide's block believes they made a mistake: they did not.

On an unmounted file system, the command asks a question

A trap in a non-interactive script. If the volume is not mounted, lvextend -r cannot call xfs_growfs, and asks for permission to mount it itself:

File system xfs found on vgdata/lvdata.
File system mount is needed for extend.
Continue with xfs file system extend steps: mount, xfs_growfs? [y/n]:[n]

Without an answer, it answers no by default, prints File system not extended. and exits with rc=5. The costly part: the volume extension is cancelled too. lvs shows that nothing moved, while a badly written script will have concluded it succeeded.

Two workarounds: mount the volume before extending, which is the normal case in production, or make the answer explicit in a script.

lvs confirms the new size:

  LV     VG     Attr       LSize Pool ...
  lvdata vgdata -wi-ao---- 6.00g

Always -r

Without -r, lvextend grows the logical volume but not the file system: the added space stays invisible. -r chains resize2fs (ext4) or xfs_growfs (XFS) for you. It is the reflex to keep.

Add a disk to a Volume Group

When the whole VG runs short of space, add a new disk to it. You initialise the PV, then aggregate it with vgextend:

sudo pvcreate /dev/sdb
sudo vgextend vgdata /dev/sdb

The VG now has the sum of both disks, and lvextend -r can draw from it. That is how storage is extended beyond the size of a single disk, without migration.

Snapshots: back up or test without risk

A snapshot is an instant copy of an LV at a moment T. Ideal before a risky update or to back up a database online:

sudo lvcreate -s -L 512M -n lvdata-snap /dev/vgdata/lvdata

  Logical volume "lvdata-snap" created.

The snapshot only stores the blocks modified since its creation (copy-on-write), hence its small initial size. You can also use it to roll back after a failed operation (lvconvert --merge), and a full snapshot becomes invalid. Consistent backup, rollback and sizing are detailed in the dedicated LVM snapshots guide.

Shrink and remove

Shrinking an LV is trickier: you must shrink the file system first, then the volume. And XFS does not shrink: only ext4 allows it.

sudo lvreduce -r -L 2G /dev/vgdata/lvdata   # ext4 only, back up first

To remove the stack, unmount then remove in the reverse order of creation: LV, VG, PV.

sudo umount /data
sudo lvremove /dev/vgdata/lvdata
sudo vgremove vgdata
sudo pvremove /dev/sda1

Going further with LVM: thin provisioning, snapshots, pvmove

Once the stack is in place and resizing is mastered, three operations cover most of everyday LVM work. Each one has its dedicated guide, with real outputs and traps to avoid.

Thin provisioning allocates space on demand and lets you overcommit storage without wasting it. LVM snapshots go deeper into online backup and rollback with lvconvert --merge. Finally, pvmove moves data from one disk to another so you can retire it without downtime.

Troubleshooting

LVM failures can almost always be read in vgs and lvs: exhausted free space, a full snapshot or a file system that did not follow the volume. The table starts from the observed symptom.

Symptom	Probable cause	Solution
Added space invisible after lvextend	-r forgotten	sudo lvextend -r or manual resize2fs/xfs_growfs
lvcreate refuses: not enough space	VG full	vgs (VFree), add a disk with vgextend
Snapshot gone / volume frozen	saturated snapshot	recreate larger, watch lvs (Data%)
Shrinking impossible	XFS volume	XFS does not shrink: back up, recreate, restore

Swipe to see more

Recommended reading

Managing LVM (building the stack, extending, snapshots) is a central objective of the RHCSA exam. The RHCSA Red Hat Enterprise Linux 9 Cert Guide covers it with exam-style exercises, persistence after reboot included.

Amazon affiliate link. Buying through this link supports the site at no extra cost to you.

Put it into practice

The lab scenario is the one you will live in production: a volume filling up that must be grown without downtime. You extend a logical volume with lvextend, grow the XFS file system, then write the mount into /etc/fstab by UUID to prove it survives the reboot.

Guided lab Practice

Learn how to do it

Extend a logical volume and prove the mount survives a reboot

$ dsoxlab run l2-lvm-extend-persist

Environment: VM·Duration: ~25 min

How it is checked: the state of the system is checked after a reboot, not the command you typed.

First time? What DSOXLab is and how to install it

DSOXLab is the command line that prepares a disposable environment (an isolated shell or a VM), runs the lab scenario, then checks the real state of the system and computes a score. Nothing is installed or broken on your workstation.

It needs Python 3.11 or later, uv or pipx, and git. The installation guide details these prerequisites; on Windows, or on a Mac without Python, use the appliance instead.

This lab starts virtual machines: it also needs ansible-playbook, a hypervisor (KVM or Incus), Terraform and hardware virtualization (/dev/kvm), which inside a VirtualBox VM means nested virtualization. The first time, dsoxlab start l2-lvm-extend-persist checks all of it before creating the machines. Prepare your machine for vm labs.

Install the CLI and add the catalogue, once:

uv tool install dsoxlab   # or: pipx install dsoxlab
dsoxlab catalog add https://github.com/stephrobert/linux-dsoxlab-training

Before you run the lab, check:

• DSOXLab installed
• catalogue available
• compatible environment (isolated shell or VM)

To check, nothing is probed from this page: run dsoxlab doctor. Read the diagnosis

Run this lab, then validate your work:

dsoxlab run l2-lvm-extend-persist
dsoxlab check l2-lvm-extend-persist

See every lab: dsoxlab list-labs. Track your progress: dsoxlab progress. The full dsoxlab guide covers scoring and the pitfalls.

Once this guided lab is mastered, test yourself under exam conditions: partitions, LVM and swap to complete without hints, against the clock, on RHEL or Ubuntu.

Challenge · exam conditions Proof

Solve it without a step-by-step procedure

Drill, partitions, LVM and swap under exam conditions

$ dsoxlab check drill-storage --target rhel|ubuntu

Environment: VM·Duration: ~25 min ·Pass mark: 70% of the score

Format: no hints, timed, dual-distribution (rhel or ubuntu): one content for both RHCSA and LFCS.

How it is checked: the score is computed without hints and against the clock, as in the exam.

First time? What DSOXLab is and how to install it

DSOXLab is the command line that prepares a disposable environment (an isolated shell or a VM), runs the lab scenario, then checks the real state of the system and computes a score. Nothing is installed or broken on your workstation.

It needs Python 3.11 or later, uv or pipx, and git. The installation guide details these prerequisites; on Windows, or on a Mac without Python, use the appliance instead.

This lab starts virtual machines: it also needs ansible-playbook, a hypervisor (KVM or Incus), Terraform and hardware virtualization (/dev/kvm), which inside a VirtualBox VM means nested virtualization. The first time, dsoxlab start drill-storage checks all of it before creating the machines. Prepare your machine for vm labs.

Install the CLI and add the catalogue, once:

uv tool install dsoxlab   # or: pipx install dsoxlab
dsoxlab catalog add https://github.com/stephrobert/linux-dsoxlab-training

Before you run the lab, check:

• DSOXLab installed
• catalogue available
• compatible environment (isolated shell or VM)

To check, nothing is probed from this page: run dsoxlab doctor. Read the diagnosis

Run this lab, then validate your work:

dsoxlab run drill-storage
dsoxlab check drill-storage

See every lab: dsoxlab list-labs. Track your progress: dsoxlab progress. The full dsoxlab guide covers scoring and the pitfalls.

Key points

• LVM stacks PV → VG → LV: the logical volume is no longer tied to a specific disk.
• pvcreate → vgcreate → lvcreate builds the stack; you format and mount the LV like a partition.
• pvs / vgs / lvs give the state; VFree shows the headroom for growth.
• lvextend -r grows volume and file system online, in one command.
• vgextend adds a disk to the pool to go beyond the size of a single disk.
• Snapshots (lvcreate -s) back up online; watch how full they are.

Knowledge check

Check that the essentials of this guide are acquired. The questions only cover what has just been explained here.

Knowledge check

Check what you have learned with this interactive quiz

6 questions

6 min

70% to pass

How it works

• The timer starts when you click Start
• Multiple choice, true/false and short answer questions
• You can move between questions
• Detailed results are shown at the end

Start the quiz

Starts the quiz and the timer

FAQ: frequently asked questions about LVM

The questions below cover those that come up most often about LVM: its principle, growing a volume and the place of snapshots.

What is LVM on Linux?

How do you grow an LVM volume online?

What is the difference between PV, VG and LV?

Can you shrink an LVM logical volume?

LVM or a classic partition, which one to choose?

How do you create an LVM snapshot?

×

See in the glossary →