Disk Partitions, Volumes, and LVM

Storage devices such as HDDs, SSDs, USB drives, and virtual disks provide raw space. Before that space can be used conveniently, it usually needs to be divided, organized, formatted, and mounted.

Linux storage can be understood in layers.

Physical disk
    |
    v
Partition
    |
    v
Filesystem
    |
    v
Mount point
    |
    v
Usable files and directories

With LVM, there are extra layers that make storage more flexible.

Physical disk or partition
    |
    v
Physical Volume, PV
    |
    v
Volume Group, VG
    |
    v
Logical Volume, LV
    |
    v
Filesystem
    |
    v
Mount point

The main idea is that normal partitions are relatively fixed, while LVM gives you a flexible storage pool that can be expanded, resized, and organized more easily.

Partitions

A partition is a section of a storage device.

A physical disk can be divided into multiple partitions. Each partition can be used for a different purpose.

For example, a disk might contain:

/boot       boot files
/           root filesystem
/home       user files
swap        swap space

A simple partition layout might look like this:

Disk: /dev/sda
+-------------+----------------+----------------+
| /dev/sda1   | /dev/sda2      | /dev/sda3      |
| /boot       | /              | /home          |
| 1G          | 40G            | 200G           |
+-------------+----------------+----------------+

Partitions are useful because they separate storage into independent regions. This can help with organization, booting, security, backups, and system recovery.

Partition Tables

A partition table describes how a disk is divided into partitions.

The two most common partition table types are:

MBR
GPT

MBR is older. GPT is newer and more common on modern systems.

MBR Partitions

MBR stands for Master Boot Record.

MBR has an important limitation: it supports only four primary partitions.

To work around this, MBR uses extended and logical partitions.

MBR Disk
+-------------+-------------+-------------------------------+
| Primary 1   | Primary 2   | Extended Partition            |
|             |             | +----------+ +-------------+  |
|             |             | | Logical  | | Logical     |  |
|             |             | | Part 1   | | Part 2      |  |
|             |             | +----------+ +-------------+  |
+-------------+-------------+-------------------------------+

The extended partition is a container. It does not directly hold a filesystem in the normal way. Instead, it contains logical partitions.

Important point:

Primary, extended, and logical partitions are mainly MBR concepts.

GPT Partitions

GPT stands for GUID Partition Table.

GPT is the modern partitioning scheme. It avoids many MBR limitations.

With GPT:

• No need for extended partitions
• No need for logical partitions
• Supports many partitions
• Better support for large disks
• More robust partition metadata

A GPT disk looks simpler conceptually:

GPT Disk
+-------------+-------------+-------------+-------------+
| Partition 1 | Partition 2 | Partition 3 | Partition 4 |
+-------------+-------------+-------------+-------------+

On modern Linux systems, GPT is usually preferred unless there is a specific compatibility reason to use MBR.

Volumes

A volume is a usable storage area that can contain a filesystem.

A volume may be a simple partition, such as:

/dev/sda1

or it may be a logical volume created by LVM, such as:

/dev/vg_data/lv_home

The important difference is that a partition is tied directly to a section of one disk, while a logical volume can be created from a flexible storage pool.

Partition vs Volume

A partition is a physical division of a disk.

A volume is a usable storage unit that usually contains a filesystem.

A simple partition-based setup looks like this:

Disk
 |
 +--> Partition
        |
        +--> Filesystem
               |
               +--> Mount point

An LVM-based setup looks like this:

Disk or partition
 |
 +--> Physical Volume
        |
        +--> Volume Group
               |
               +--> Logical Volume
                      |
                      +--> Filesystem
                             |
                             +--> Mount point

In everyday use, both a partition and a logical volume can be formatted and mounted.

The difference is in how flexible they are to manage.

Why Use LVM?

LVM stands for Logical Volume Manager.

LVM allows Linux systems to manage storage more flexibly than traditional partitions.

With LVM, you can:

• Combine multiple disks into one storage pool
• Create logical volumes from that pool
• Extend logical volumes when more space is needed
• Reduce some logical volumes carefully
• Move data between disks
• Create snapshots
• Add new disks to existing storage
• Manage storage using meaningful names

For example, instead of being locked into a fixed /home partition size, you could create a logical volume for /home and expand it later if users need more space.

LVM Big Picture

LVM has three main layers:

Physical Volume, PV
Volume Group, VG
Logical Volume, LV

A physical volume is a disk or partition prepared for LVM.

A volume group is a pool of storage made from one or more physical volumes.

A logical volume is a usable storage unit created inside a volume group.

|-----------------|     |-----------------|     |-----------------|
| Physical Volume |     | Physical Volume |     | Physical Volume |
| PV: /dev/sdb    |     | PV: /dev/sdc    |     | PV: /dev/sdd    |
| Size: 50G       |     | Size: 100G      |     | Size: 150G      |
|-----------------|     |-----------------|     |-----------------|
        |                       |                       |
        +-----------------------+-----------------------+
                                |
                                v
|------------------------------------------------------------------|
|                      Volume Group: TEST                          |
|                        Total Size: 300G                          |
|                                                                  |
|  |------------------|     |------------------|                   |
|  | Logical Volume   |     | Logical Volume   |   Free Space      |
|  | vol_name_1       |     | vol_name_2       |   30G             |
|  | Size: 180G       |     | Size: 90G        |                   |
|  |------------------|     |------------------|                   |
|                                                                  |
|------------------------------------------------------------------|

This means:

• /dev/sdb, /dev/sdc, and /dev/sdd are physical volumes
• They are combined into a volume group named TEST
• Logical volumes are created from space inside TEST
• Unused space remains available for future expansion

Physical Volumes

A physical volume, or PV, is a disk or partition initialized for use by LVM.

Examples:

/dev/sdb
/dev/sdc1
/dev/nvme1n1p3

A PV is not normally mounted directly. Instead, it becomes part of a volume group.

To create a physical volume:

sudo pvcreate /dev/sdb

To view physical volumes:

sudo pvs

Example output:

PV         VG      Fmt  Attr PSize   PFree
/dev/sdb   TEST    lvm2 a--   50.00g  0
/dev/sdc   TEST    lvm2 a--  100.00g 10.00g
/dev/sdd   TEST    lvm2 a--  150.00g 20.00g

Interpretation:

Volume Groups

A volume group, or VG, is a storage pool made from one or more physical volumes.

Example:

TEST

A VG combines the capacity of all physical volumes assigned to it.

To create a volume group:

sudo vgcreate TEST /dev/sdb

To add more physical volumes to an existing volume group:

sudo vgextend TEST /dev/sdc /dev/sdd

To view volume groups:

sudo vgs

Example output:

VG     #PV #LV #SN Attr   VSize    VFree
TEST     3   2   0 wz--n- 300.00g  30.00g

Interpretation:

The VFree value is important. It tells you how much space is available for new logical volumes or for extending existing ones.

Logical Volumes

A logical volume, or LV, is the usable storage area created inside a volume group.

It behaves much like a partition. You can format it, mount it, and store files on it.

Examples:

/dev/TEST/vol_name_1
/dev/mapper/TEST-vol_name_1

Both paths usually refer to the same logical volume.

To create a logical volume:

sudo lvcreate -L 20G -n vol_name_1 TEST

This creates a 20 GB logical volume named vol_name_1 inside the volume group TEST.

To view logical volumes:

sudo lvs

Example output:

LV          VG    Attr       LSize   Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
vol_name_1  TEST  -wi-a----- 180.00g
vol_name_2  TEST  -wi-a-----  90.00g

Interpretation:

Filesystems on Logical Volumes

A logical volume is not ready for normal file storage until it has a filesystem.

For example, to create an ext4 filesystem:

sudo mkfs.ext4 /dev/TEST/vol_name_1

Then create a mount point:

sudo mkdir -p /mnt/mydata

Mount it:

sudo mount /dev/TEST/vol_name_1 /mnt/mydata

Check it:

df -h /mnt/mydata

Example output:

Filesystem                    Size  Used Avail Use% Mounted on
/dev/mapper/TEST-vol_name_1    20G   24K   19G   1% /mnt/mydata

Interpretation:

The logical volume is formatted, mounted, and usable.
Files written to /mnt/mydata are stored on the LV.

LVM Naming

LVM names usually follow this pattern:

/dev/VOLUME_GROUP/LOGICAL_VOLUME

Example:

/dev/TEST/vol_name_1

Linux may also show the same LV under /dev/mapper:

/dev/mapper/TEST-vol_name_1

Both refer to the same logical volume.

Basic LVM Creation Workflow

A full LVM setup usually follows this order:

1. Identify disks or partitions
2. Create physical volumes
3. Create a volume group
4. Create logical volumes
5. Create filesystems
6. Create mount points
7. Mount logical volumes
8. Optionally configure /etc/fstab

Example:

sudo pvcreate /dev/sdb /dev/sdc /dev/sdd
sudo vgcreate TEST /dev/sdb
sudo vgextend TEST /dev/sdc /dev/sdd

sudo lvcreate -L 20G -n vol_name_1 TEST
sudo lvcreate -L 12G -n vol_name_2 TEST

sudo mkfs.ext4 /dev/TEST/vol_name_1
sudo mkdir -p /mnt/vol1
sudo mount /dev/TEST/vol_name_1 /mnt/vol1

Extending a Logical Volume

One of the biggest advantages of LVM is that logical volumes can be extended.

If the volume group has free space, you can increase an LV size.

Example:

sudo lvextend -L +10G /dev/TEST/vol_name_1

This adds 10 GB to the logical volume.

However, the filesystem also needs to be expanded.

For ext4:

sudo resize2fs /dev/TEST/vol_name_1

A more convenient method is to use -r, which resizes the filesystem automatically:

sudo lvextend -r -L +10G /dev/TEST/vol_name_1

This is often safer and simpler because it handles both steps:

extend LV
resize filesystem

Reducing a Logical Volume

Reducing a logical volume is more dangerous than extending it.

If done incorrectly, it can destroy data.

The safe order for ext4 is:

1. Back up important data
2. Unmount the filesystem
3. Check the filesystem
4. Shrink the filesystem
5. Shrink the logical volume
6. Mount again
7. Verify data

Example:

sudo umount /dev/TEST/vol_name_1
sudo e2fsck -f /dev/TEST/vol_name_1
sudo resize2fs /dev/TEST/vol_name_1 15G
sudo lvreduce -L 15G /dev/TEST/vol_name_1
sudo mount /dev/TEST/vol_name_1 /mnt/vol1

Important warning:

Do not reduce an LV before reducing the filesystem.
The filesystem must fit inside the smaller LV.

For some filesystems, such as XFS, shrinking is not supported. XFS can be grown online but not shrunk in place.

Deleting a Logical Volume

To delete a logical volume, first unmount it.

sudo umount /mnt/vol1

Then remove it:

sudo lvremove /dev/TEST/vol_name_1

LVM will ask for confirmation.

After removal, the space returns to the volume group as free space.

Check with:

sudo vgs

LVM Snapshots

An LVM snapshot captures the state of a logical volume at a point in time.

Snapshots are useful for:

• Temporary backups
• Consistent backup windows
• Testing risky changes
• Rollback scenarios
• Capturing a filesystem state before updates

A snapshot does not immediately copy all data. Instead, it uses copy-on-write behavior. When data changes on the original LV, the snapshot keeps the old blocks so it can preserve the earlier state.

Original LV before snapshot
        |
        v
Snapshot created
        |
        v
Changes happen on original LV
        |
        v
Snapshot preserves old data blocks

Create a snapshot:

sudo lvcreate -L 1G -s -n snap_vol_name_1 /dev/TEST/vol_name_1

This creates a 1 GB snapshot named snap_vol_name_1.

View snapshots:

sudo lvs

Example output:

LV               VG    Attr       LSize  Origin      Data%
snap_vol_name_1  TEST  swi-a-s---  1.00g vol_name_1  12.00
vol_name_1       TEST  owi-a-s--- 20.00g

Interpretation:

• snap_vol_name_1 is a snapshot of vol_name_1
• Data% shows how much of the snapshot space is used
• If Data% reaches 100%, the snapshot becomes invalid

Restoring from a Snapshot

To merge a snapshot back into the original logical volume:

sudo lvconvert --merge /dev/TEST/snap_vol_name_1

The merge may happen immediately or at the next activation, depending on whether the origin volume is active.

A common safe workflow is:

• Create snapshot
• Make changes
• If changes are bad, unmount original LV
• Merge snapshot
• Reactivate LV
• Mount again

Snapshots are useful, but they are not a full backup replacement. If the disk fails, snapshots on the same disk may be lost too.

Persistent Mounting with /etc/fstab

If you want an LV mounted automatically at boot, add it to /etc/fstab.

Example:

/dev/TEST/vol_name_1   /mnt/vol1   ext4   defaults   0   2

A more robust option is to use a UUID.

Find the UUID:

sudo blkid /dev/TEST/vol_name_1

Example:

/dev/TEST/vol_name_1: UUID="1111-2222" TYPE="ext4"

Then use:

UUID=1111-2222   /mnt/vol1   ext4   defaults   0   2

Test /etc/fstab before rebooting:

sudo mount -a

Partition Types Compared

Primary partition:

• traditional partition type, especially on MBR disks
• can contain a filesystem or bootable OS files
• MBR supports up to 4 primary partitions

Extended partition:

• MBR-only container used to work around the 4-partition limit
• does not directly store normal data
• contains logical partitions

Logical partition:

• created inside an extended partition on MBR disks
• can contain a filesystem
• used when more than 4 partitions are needed

LVM logical volume:

• flexible virtual storage unit created inside a volume group
• can be resized more easily
• can span multiple disks through the volume group
• supports snapshots and advanced management

Modern note:

On GPT disks, primary/extended/logical partition limits are mostly not relevant.
GPT supports many normal partitions without extended partitions.

When to Use Partitions vs LVM

Use normal partitions when:

• The layout is simple
• The system is small
• You do not need resizing flexibility
• You want easier recovery with minimal layers
• You are creating simple boot or EFI partitions

Use LVM when:

• Storage may grow later
• You need flexible resizing
• You want to combine disks into a pool
• You want snapshots
• You manage servers or virtual machines
• You want meaningful volume names

A common Linux server layout might use both:

/boot      normal partition
/boot/efi  EFI system partition
/          LVM logical volume
/home      LVM logical volume
/var       LVM logical volume

Useful LVM Commands

Creation commands:

• pvcreate /dev/sdb
• vgcreate TEST /dev/sdb
• vgextend TEST /dev/sdc
• lvcreate -L 20G -n vol_name_1 TEST
• mkfs.ext4 /dev/TEST/vol_name_1
• mount /dev/TEST/vol_name_1 /mnt/vol1

Resize commands:

lvextend -r -L +10G /dev/TEST/vol_name_1
resize2fs /dev/TEST/vol_name_1
lvreduce -L 15G /dev/TEST/vol_name_1

Snapshot commands:

lvcreate -L 1G -s -n snap_vol_name_1 /dev/TEST/vol_name_1
lvconvert --merge /dev/TEST/snap_vol_name_1

Removal commands:

umount /mnt/vol1
lvremove /dev/TEST/vol_name_1
vgremove TEST
pvremove /dev/sdb

Safe LVM Lab Setup Using Loopback Disks

The following lab simulates disks using regular files.

This is safer than practicing on real disks because it avoids accidentally destroying real data.

Important warning:

• These commands still use sudo and create block devices
• Run them only on a test machine or lab VM
• Carefully copy device names
• Do not use real disks such as /dev/sda unless you intend to erase them

Scenario 1: Create a Complete LVM Setup in a Lab

Simulate three disks, create LVM physical volumes, combine them into one volume group, create a logical volume, format it, mount it, and verify that it works.

Step 1: Create Test Disk Files

mkdir -p ~/lvm-lab

truncate -s 1G ~/lvm-lab/disk1.img
truncate -s 1G ~/lvm-lab/disk2.img
truncate -s 1G ~/lvm-lab/disk3.img

These files act like fake disks.

Step 2: Attach Files as Loop Devices

sudo losetup --find --show ~/lvm-lab/disk1.img
sudo losetup --find --show ~/lvm-lab/disk2.img
sudo losetup --find --show ~/lvm-lab/disk3.img

Example output:

/dev/loop10
/dev/loop11
/dev/loop12

Interpretation:

• The files are now visible as block devices
• In this example, the fake disks are /dev/loop10, /dev/loop11, and /dev/loop12
• Your device numbers may be different

Step 3: Create Physical Volumes

sudo pvcreate /dev/loop10 /dev/loop11 /dev/loop12

Example output:

Physical volume "/dev/loop10" successfully created.
Physical volume "/dev/loop11" successfully created.
Physical volume "/dev/loop12" successfully created.

Check:

sudo pvs

Example output:

PV           VG  Fmt  Attr PSize    PFree
/dev/loop10      lvm2 ---  1024.00m 1024.00m
/dev/loop11      lvm2 ---  1024.00m 1024.00m
/dev/loop12      lvm2 ---  1024.00m 1024.00m

Interpretation:

The loop devices are prepared for LVM.
They are not yet part of a volume group.

Step 4: Create a Volume Group

sudo vgcreate lab_vg /dev/loop10 /dev/loop11

Example output:

Volume group "lab_vg" successfully created

Check:

sudo vgs

Example output:

VG      #PV #LV #SN Attr   VSize  VFree
lab_vg    2   0   0 wz--n- 1.99g  1.99g

Interpretation:

Two 1 GB loop disks are combined into one volume group.
The volume group has about 2 GB of usable space.

Step 5: Create a Logical Volume

sudo lvcreate -L 1G -n data lab_vg

Example output:

Logical volume "data" created.

Check:

sudo lvs

Example output:

LV    VG      Attr       LSize Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
data  lab_vg  -wi-a----- 1.00g

Interpretation:

A 1 GB logical volume named data now exists inside lab_vg.

Step 6: Create a Filesystem

sudo mkfs.ext4 /dev/lab_vg/data

Example output:

Creating filesystem with 262144 4k blocks and 65536 inodes
Filesystem UUID: 12345678-abcd-1234-abcd-123456789abc

Interpretation:

The logical volume now contains an ext4 filesystem.
It can be mounted and used for files.

Step 7: Mount the Logical Volume

sudo mkdir -p /mnt/lvmlab
sudo mount /dev/lab_vg/data /mnt/lvmlab

Check:

df -h /mnt/lvmlab

Example output:

Filesystem               Size  Used Avail Use% Mounted on
/dev/mapper/lab_vg-data  974M   24K  907M   1% /mnt/lvmlab

Interpretation:

The logical volume is mounted and usable.
Files placed in /mnt/lvmlab are stored on the LVM logical volume.

Scenario 2: Simulate a Full Filesystem and Fix It with LVM

Simulate a common storage bottleneck: a mounted filesystem is almost full, but the volume group still has free space.

This is one of the best practical use cases for LVM.

Step 1: Fill the Filesystem

sudo fallocate -l 850M /mnt/lvmlab/bigfile

Check usage:

df -h /mnt/lvmlab

Example output:

Filesystem               Size  Used Avail Use% Mounted on
/dev/mapper/lab_vg-data  974M  851M   58M  94% /mnt/lvmlab

Interpretation:

The filesystem is almost full.
Applications writing to /mnt/lvmlab may fail soon.

Step 2: Check Whether the Volume Group Has Free Space

sudo vgs

Example output:

VG      #PV #LV #SN Attr   VSize  VFree
lab_vg    2   1   0 wz--n- 1.99g  1016.00m

Interpretation:

• The filesystem is almost full, but the volume group has about 1 GB free
• This means the logical volume can be extended

Step 3: Extend the Logical Volume and Filesystem

Use -r to resize the filesystem at the same time:

sudo lvextend -r -L +500M /dev/lab_vg/data

Example output:

Size of logical volume lab_vg/data changed from 1.00 GiB to 1.49 GiB.
Logical volume lab_vg/data successfully resized.
resize2fs 1.46.5
The filesystem is now 390144 blocks long.

Step 4: Verify the Result

df -h /mnt/lvmlab

Example output:

Filesystem               Size  Used Avail Use% Mounted on
/dev/mapper/lab_vg-data  1.5G  851M  542M  62% /mnt/lvmlab

Interpretation:

• The logical volume grew
• The filesystem grew
• The mount point now has more free space
• The capacity bottleneck is resolved

Scenario 3: Simulate “No Space Left in Volume Group”

Show what happens when the filesystem needs more space but the volume group does not have enough free space.

Step 1: Try to Extend Too Much

sudo lvextend -r -L +5G /dev/lab_vg/data

Example output:

Insufficient free space: 1280 extents needed, but only 129 available

Interpretation:

• The logical volume cannot be extended because the volume group does not have enough free space
• This is not a filesystem problem
• This is a volume group capacity problem

Step 2: Confirm with vgs

sudo vgs

Example output:

VG      #PV #LV #SN Attr   VSize  VFree
lab_vg    2   1   0 wz--n- 1.99g  516.00m

Interpretation:

Only about 516 MB is free in the volume group.
A 5 GB extension is impossible without adding more storage.

Step 3: Add Another Physical Volume

Earlier, /dev/loop12 was prepared as a PV but not added to the VG.

Add it now:

sudo vgextend lab_vg /dev/loop12

Example output:

Volume group "lab_vg" successfully extended

Check:

sudo vgs

Example output:

VG      #PV #LV #SN Attr   VSize  VFree
lab_vg    3   1   0 wz--n- 2.99g  1.51g

Interpretation:

• The volume group now includes another physical volume
• Free space increased
• The LV can now be extended further

Step 4: Extend Again

sudo lvextend -r -L +1G /dev/lab_vg/data

Check:

df -h /mnt/lvmlab

Example output:

Filesystem               Size  Used Avail Use% Mounted on
/dev/mapper/lab_vg-data  2.5G  851M  1.6G  35% /mnt/lvmlab

Interpretation:

• Adding a PV increased VG capacity
• The LV and filesystem were extended
• The filesystem now has enough free space

Scenario 4: Simulate and Inspect LVM Layers

Use LVM tools to understand the relationship between PVs, VGs, LVs, filesystems, and mount points.

Check Block Devices

lsblk

Example output:

NAME              SIZE TYPE MOUNTPOINT
loop10              1G loop
└─lab_vg-data     2.5G lvm  /mnt/lvmlab
loop11              1G loop
└─lab_vg-data     2.5G lvm  /mnt/lvmlab
loop12              1G loop
└─lab_vg-data     2.5G lvm  /mnt/lvmlab

Interpretation:

• The logical volume data is built from space in the volume group
• The volume group is backed by multiple loop devices
• The logical volume is mounted at /mnt/lvmlab

Check Physical Volumes

sudo pvs

Example output:

PV           VG      Fmt  Attr PSize    PFree
/dev/loop10  lab_vg  lvm2 a--  1020.00m    0
/dev/loop11  lab_vg  lvm2 a--  1020.00m    0
/dev/loop12  lab_vg  lvm2 a--  1020.00m  500.00m

Interpretation:

All three loop devices are physical volumes.
Some free space remains on /dev/loop12.

Check Volume Group

sudo vgs

Example output:

VG      #PV #LV #SN Attr   VSize  VFree
lab_vg    3   1   0 wz--n- 2.99g 500.00m

Interpretation:

• The volume group contains 3 PVs and 1 LV
• It still has 500 MB free

Check Logical Volume

sudo lvs

Example output:

LV    VG      Attr       LSize Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
data  lab_vg  -wi-ao---- 2.50g

Interpretation:

• The data LV is 2.5 GB
• The o attribute indicates it is open, meaning it is active and mounted or in use

Scenario 5: Create a Snapshot Before a Risky Change

Create a snapshot, make a risky change, and understand how the snapshot protects the earlier state.

Step 1: Create a Test File

echo "original version" | sudo tee /mnt/lvmlab/example.txt
cat /mnt/lvmlab/example.txt

Example output:

original version

Step 2: Create a Snapshot

sudo lvcreate -L 200M -s -n data_snap /dev/lab_vg/data

Example output:

Logical volume "data_snap" created.

Check:

sudo lvs

Example output:

LV        VG      Attr       LSize   Origin Data%
data      lab_vg  owi-aos---   2.50g
data_snap lab_vg  swi-a-s--- 200.00m data   0.01

Interpretation:

• data_snap is a snapshot of data
• It preserves the original state at the time the snapshot was created

Step 3: Make a Risky Change

echo "bad change" | sudo tee /mnt/lvmlab/example.txt
cat /mnt/lvmlab/example.txt

Example output:

bad change

Step 4: Inspect Snapshot Usage

sudo lvs

Example output:

LV        VG      Attr       LSize   Origin Data%
data      lab_vg  owi-aos---   2.50g
data_snap lab_vg  swi-a-s--- 200.00m data   1.25

Interpretation:

• The snapshot is using some space because the original LV changed after the snapshot was created
• If Data% reaches 100%, the snapshot becomes invalid

Step 5: Restore by Merging the Snapshot

Unmount the filesystem first:

cd ~
sudo umount /mnt/lvmlab

Merge the snapshot:

sudo lvconvert --merge /dev/lab_vg/data_snap

Example output:

Merging of volume lab_vg/data_snap started.
lab_vg/data: Merged: 100.00%

Reactivate if needed:

sudo lvchange -ay /dev/lab_vg/data

Mount again:

sudo mount /dev/lab_vg/data /mnt/lvmlab
cat /mnt/lvmlab/example.txt

Expected output:

original version

Interpretation:

• The snapshot merge restored the LV to the earlier state
• The risky change was undone

Scenario 6: Simulate Snapshot Space Exhaustion

Show why snapshot size matters.

Snapshots need enough space to store changed blocks. If too much changes after the snapshot is created, the snapshot can fill up and become unusable.

Step 1: Create a Small Snapshot

sudo lvcreate -L 50M -s -n small_snap /dev/lab_vg/data

Step 2: Write Enough Data to the Original LV

sudo fallocate -l 200M /mnt/lvmlab/change-after-snapshot.bin

Step 3: Check Snapshot Usage

sudo lvs

Example output:

LV          VG      Attr       LSize  Origin Data%
data        lab_vg  owi-aos--- 2.50g
small_snap  lab_vg  swi-a-s--- 50.00m data   100.00

Interpretation:

• The snapshot is full
• A full snapshot may become invalid
• This means the snapshot can no longer reliably represent the original point-in-time state

Lesson

• Snapshot size must be chosen based on how much data may change while the snapshot exists
• Snapshots should usually be temporary

Scenario 7: Clean Up the LVM Lab

Remove the test LVM setup safely.

Step 1: Unmount

cd ~
sudo umount /mnt/lvmlab

Step 2: Remove Snapshots if Any Remain

sudo lvs
sudo lvremove /dev/lab_vg/small_snap

Only remove snapshots that exist.

Step 3: Remove the Logical Volume

sudo lvremove /dev/lab_vg/data

Example output:

Do you really want to remove active logical volume lab_vg/data? [y/n]: y
Logical volume "data" successfully removed.

Step 4: Remove the Volume Group

sudo vgremove lab_vg

Step 5: Remove Physical Volume Labels

sudo pvremove /dev/loop10 /dev/loop11 /dev/loop12

Step 6: Detach Loop Devices

sudo losetup -d /dev/loop10
sudo losetup -d /dev/loop11
sudo losetup -d /dev/loop12

Step 7: Delete Test Files

rm -rf ~/lvm-lab
sudo rmdir /mnt/lvmlab

Check:

lsblk

Interpretation:

The test loop devices, LVM objects, mount point, and files have been removed.
The lab environment is cleaned up.

Common LVM Troubleshooting

Problem: Filesystem Is Full

Symptoms:

applications cannot write files
df -h shows 100% usage
"No space left on device"

Check:

df -h
sudo vgs
sudo lvs

Interpretation:

• If df is full but VFree exists in vgs, extend the LV
• If df is full and VFree is 0, add storage or delete/move data

Fix if VG has free space:

sudo lvextend -r -L +10G /dev/VG_NAME/LV_NAME

Problem: Volume Group Has No Free Space

Symptoms:

lvextend fails
error mentions insufficient free extents
vgs shows VFree as 0

Check:

sudo vgs
sudo pvs

Fix:

sudo pvcreate /dev/newdisk
sudo vgextend VG_NAME /dev/newdisk
sudo lvextend -r -L +10G /dev/VG_NAME/LV_NAME

Problem: Logical Volume Exists but Is Not Mounted

Check:

sudo lvs
lsblk
findmnt

If the LV exists but is not mounted, mount it:

sudo mkdir -p /mnt/data
sudo mount /dev/VG_NAME/LV_NAME /mnt/data

Problem: LVM Devices Not Active

Sometimes LVM volumes exist but are inactive.

Check:

sudo lvs

Activate all volume groups:

sudo vgchange -ay

Then check again:

lsblk

Problem: Snapshot Is Full

Check:

sudo lvs

If snapshot Data% is near 100%, it is at risk.

Fix options:

• Remove the snapshot if no longer needed
• Extend the snapshot if possible
• Merge it if rollback is needed
• Avoid keeping snapshots too long

Remove snapshot:

sudo lvremove /dev/VG_NAME/SNAPSHOT_NAME

LVM Safety Rules

• Always back up important data before resizing
• Extending is usually safer than reducing
• Do not reduce an LV before shrinking the filesystem
• Use lvextend -r when possible
• Check df -h, lvs, vgs, and pvs before making changes
• Do not run pvcreate on a disk containing data you want to keep
• Snapshots are useful, but they are not full backups
• Test /etc/fstab changes with mount -a before rebooting

Challenges

1. Describe the process of reducing the size of an LVM partition. Explain the precautions to take, such as ensuring adequate free space and backing up any critical data on the partition. Outline the commands you would use, including how to first resize the filesystem before reducing the logical volume itself.
2. Explain how to extend an LVM partition to utilize more free space within the volume group. Describe the process and commands required to first extend the logical volume and then resize the filesystem to use the additional space.
3. Describe the process of taking a snapshot of an LVM logical volume, including the reasons why snapshots are useful (such as for backups or testing changes). Provide the commands needed to create a snapshot and explain how to restore data from a snapshot if needed.
4. Detail the steps required to remove a logical volume from an LVM setup. Describe the commands needed for each step and discuss precautions to ensure data safety before, during, and after the removal process, such as unmounting the volume and backing up important data.
5. Explain the steps for adding a new physical disk to an existing volume group in an LVM setup. Describe the process, including the commands to initialize the disk as a physical volume, add it to the volume group, and verify the volume group has expanded.
6. Research and describe the concept of striping in LVM, including how striping can improve read and write performance by distributing data across multiple physical volumes. Explain the process of setting up a striped logical volume and the commands needed to specify the stripe size.
7. Describe how to monitor the status of your LVM setup, including checking for available free space, the health of logical volumes, and the status of volume groups. List the commands you would use, such as lvdisplay, vgdisplay, and pvdisplay, and explain the key information each command provides.
8. Compare a RAID setup with and without LVM. Discuss the benefits that LVM adds to a RAID environment, such as simplified management, easier resizing of logical volumes, and flexibility in adding storage. Provide examples of how LVM complements RAID setups.
9. Describe the process for moving data from one physical volume to another within an LVM setup, using the pvmove command. Explain why this might be necessary, such as for hardware maintenance or upgrading storage, and outline the steps to ensure data remains accessible throughout the move.
10. Research the LVM caching feature, which allows you to use a faster disk (such as an SSD) as a cache for a slower logical volume. Describe how this setup can improve performance, and provide an outline of the commands needed to set up an LVM cache.