Mickael's place

Mounting partitions with Logical Volume Manager: LVM2

Mounting partitions with Logical Volume Manager: LVM2

Mickael's photo
Mickael
·

7 min read

Table of contents

Mounting partitions with Logical Volume Manager: LVM2

Purpose

The aim of this quick topic is to see how we can manage our partition using logical volume 2 on a ubuntu bionic virtual machine.

System configuration: VM Info:

Note: This topic is more focus on how to use lvm rather than how lvm works Keep in mind that sys admin system manipulations require a good understanding of concepts you are dealing with. If you want to know more about LVM please refer to the documentation

Quick overview

lvm purpose is to give the user a partition managment of physical driver using a logical layer. It basically consists of three steps:

  1. Set Physical Volume.
  2. Gathering Physical Volume into a Volume Group.
  3. Split Volume Group into Logical Volume.

Diagram

The diagram below, shows an overview of the process:

lvm_overview.drawio.png

Use case

In this case I have a main partition sda mount on / , and I add a second 2GB disk sdb through the Virtual Box interface.

Note: Even if everything is virtually generated with VBox, what matters here is the OS point of view.

Here we are going to:

  • Associate sdb to a Physical Volume
  • Create a Volume Group
  • Create Logical Volume
  • Mount the Logical Volume
  • Resize the Logical Volume

Physical Volume set up

Let's see what do we have at the beginning:

lvmdiskscan

Output:

    /dev/sda1 [       9,00 GiB]
    /dev/sda5 [    1022,00 MiB]
    /dev/sdb  [       2,00 GiB]
    1 disk
    2 partitions
    0 LVM physical volume whole disks
    0 LVM physical volumes

We create a physical volume associated to sdb

pvcreate /dev/sdb

Output:

    Physical volume "/dev/sdb" successfully created

Volume Group creation

By now, our physical volume has been created, let's create a volume group from it:

vgcreate vg_test /dev/sdb

Output:

    Volume group "vg_test" successfully created

Below some features that could be useful:

  • Size the volume when create it: $ vgcreate -s 8M vg_test /dev/sdb
  • Add physical disk to an existing volume: $ vgextend vg_test /dev/sdc

Create a Logical Volume from this Volume Group

As it is display in the diagram we are going to split the Volume group into Logical Volume. Let's start with a 1GB logical partition

lvcreate -L 1G -n lv_test vg_test
  • -L : size parameter
  • -n : name of the partition

Output:

    Logical volume "lv_test" created

Note: Here we will create only one logical volume, but you can set your own partitionning using the command above. _A useful command, just in case: Create a logical volume from all free space left: $ lvcreate -l 100%FREE -n lv_test_2 vg_test_

  • -n : name of the LV
  • -l 100%FREE : size to all the volume left

Check

Let's run the same command we ran at the beginning

lvmdiskscan

Output

  /dev/vg_test/lv_test [       1,00 GiB]
  /dev/sda1            [       9,00 GiB]
  /dev/sda5            [    1022,00 MiB]
  /dev/sdb             [       2,00 GiB] LVM physical volume
  1 disk
  2 partitions
  1 LVM physical volume whole disk
  0 LVM physical volumes

Our system configuration has been successfully took into account !

Set our new logical partition

Now that our partition has been created we will format and mount it.

Partition formatting

We will create a new ext4 file system

Note: more information about ext4 here

mkfs.ext4 /dev/vg_test/lv_test

Output:

    File system label =
    Operating System Type: Linux
    Block size = 4096 (log = 2)
    Fragment size = 4096 (log = 2)
    "Stride" = 0 blocks, "Stripe width" = 0 blocks
    65536 i-nodes, 262144 blocks
    13,107 blocks (5.00%) reserved for the super user
    First data block = 0
    Maximum number of filesystem blocks = 268435456
    8 groups of blocks
    32768 blocks per group, 32768 fragments per group
    8192 i-nodes per group
    Emergency superblocks stored on the blocks:
    32768, 98304, 163840, 229376

    Allocation of group tables: completed
    Writing of inode tables: completed
    Creation of the journal (8192 blocks): completed
    Writing the superblocks and accounting information of the system
    files: completed

Mount the volume

Let's mount our new file system on /var/log 

mount /dev/vg_test/lv_test /var/log

Check it:

df -h

Output

    Filesystem                   Size  Used Avail Use% Mounted on
    udev                         483M  4,0K  483M   1% /dev
    tmpfs                        100M  860K   99M   1% /run
    /dev/sda1                    8,8G  4,6G  3,8G  56% /
    none                         4,0K     0  4,0K   0% /sys/fs/cgroup
    none                         5,0M     0  5,0M   0% /run/lock
    none                         497M   76K  497M   1% /run/shm
    none                         100M   40K  100M   1% /run/user
    /dev/mapper/vg_test-lv_test  976M  1,3M  908M   1% /var/log

Manipulate the logical volume

One of the assests of lvm over historical way to partition a disk is that it is very easy to manipulate afterwards. Here we going to resize the FS down to 400MB, first we need to unmount the FS.

Unmount the volume

The path /var/log is a specific one cause it cannot be unmount easily. Let's try to unmount it

umount /dev/vg_test/lv_test

Output:

umount: /var/log: device is busy.    
        (In some cases useful info about processes that use
         the device is found by lsof(8) or fuser(1))

e.g: No way :(

In fact /var/log is a directory where a lot of daemons/applications write their logs, so it is constantly writing.

To unmount it we have to:

  1. Edit the fstab file
vim /etc/fstab

Comment the line dedicated to the lv_test partition (at the end of the file)


    # /etc/fstab: static file system information.
    #
    # Use 'blkid' to print the universally unique identifier for a
    # device; this may be used with UUID= as a more robust way to name devices
    # that works even if disks are added and removed. See fstab(5).
    #
    # <file system> <mount point>   <type>  <options>       <dump>  <pass>
    # / was on /dev/sda1 during installation
    UUID=19d6e63f-0932-407c-8ae4-2c5d086eaa6f /               ext4    errors=remount-ro 0       1
    # swap was on /dev/sda5 during installation
    UUID=55df542f-da15-403a-9a68-b9685b9cd114 none            swap    sw              0       0

    ############ Below the line we have to comment #######
    #/dev/vg_test/lv_test /var/log ext4 nodev,nosuid,noexec 0       2

  1. Then a reboot the machine

    reboot

Once the machine is back and you are logged in, check that the file system has been unmount.

df -h

Output

Filesystem      Size  Used Avail Use% Mounted on
udev            483M  4,0K  483M   1% /dev
tmpfs           100M  856K   99M   1% /run
/dev/sda1       8,8G  4,6G  3,8G  56% /
none            4,0K     0  4,0K   0% /sys/fs/cgroup
none            5,0M     0  5,0M   0% /run/lock
none            497M   76K  497M   1% /run/shm
none            100M   28K  100M   1% /run/user

Note: Usually a umount command is enough for this kind of operation, but I found this case interessting and it comes from a true personal story :P

lv_test size modification:

Now the partition has been unmount let's see what do we have:

lvdisplay

Output:

--- Logical volume ---
LV Path                /dev/vg_test/lv_test
LV Name                lv_test
VG Name                vg_test
LV UUID                S69CrW-55YL-gPRi-ddvd-ovEZ-6RRS-tMaHVX
LV Write Access        read/write
LV Creation host, time my-bionic, 2020-02-17 11:02:41 +0100
LV Status              available
# open                 0
LV Size                1,00 GiB
...

Before doing anything we should do a file system integrity check:

e2fsck -ff /dev/vg_test/lv_test

Output:

e2fsck 1.42.9 (4-Feb-2014)
Pass 1: check i-nodes, blocks and sizes
Pass 2: checking the directory structure
Pass 3: verify directory connectivity
Pass 4: check the reference counters
Pass 5: verification of the information in the group summary /dev/vg_test/ lv_test: 27/65536 files (3.7% non-contiguous), 12713/262144 blocks

Once all check are ok, we can manipulate the FS, let's resize it down to 400MB

resize2fs /dev/vg_test/lv_test 400M

Output

resize2fs 1.42.9 (4-Feb-2014)
Resizing filesystem on /dev/vg_test/ lv_test to 409600 (1k) blocks. The /dev/vg_test/ lv_test filesystem is now 409600 blocks in size.

Note: Here we resized the file system but the logical volume is still 1GB sized.

Now our file system is sized to 400MB as we want, we can decrease the logical volume:

lvreduce --size 400M /dev/vg_test/lv_test

Output (note the warning !):

    WARNING: Reducing active logical volume to 400,00 MiB
    THIS MAY DESTROY YOUR DATA (filesystem etc.)

Do you really want to reduce lv_test? [y/n]: y
    Reducing logical volume lv_test to 400,00 MiB
    Logical volume lv_test successfully resized

The logical volume is now sized to 400MB, we can mount again the FS.

mount /dev/vg_test/lv_test /var/log/ -t auto

Let's have a final check on our file system usage:

df -H

Output:

Filesystem                   Size  Used Avail Use% Mounted on
udev                         506M  4,1k  506M   1% /dev
tmpfs                        105M  877k  104M   1% /run
/dev/sda1                    9,4G  4,9G  4,0G  56% /
none                         4,1k     0  4,1k   0% /sys/fs/cgroup
none                         5,3M     0  5,3M   0% /run/lock
none                         521M   78k  521M   1% /run/shm
none                         105M   29k  105M   1% /run/user
/dev/mapper/vg_test-lv_test  500M  2,4M  467M   1% /var/log

And here we are ! Now you have an overview on how use lvm to manipulate partitions, how create file system and mount them.

Tank you a lot for reading this post, see you soon !

Linuxadminstorage