Core dump overflow

Core dump in progress...

LFCS prep - LVM storage management

| Comments

Today we’ll look at logical volume management tasks. LVM improves the traditional partition to disk approach, adding flexibility (extending and reducing space), snapshot capabilities and redundancy.

First, some terminology is in order:

  • physical volumes (PV) are the physical devices that are mapped to be used inside LVM. Their storace unit is called a physical extent (PE)

  • volume groups (VG) are storage pools made from PVs

  • logical volumes (LV) are made up of the VG space

Creating PVs

I am now going to create 2 PVs of 100 MB each. You first need to create 2 LVM partitions. If you look inside fdisk, you see the partition code for LVM is 8e:

1

8e  Linux LVM

So before writing the partition in fdisk, use t to change it to Linux LVM. Then verify that the new partitions were created:

1
2
3
4

lsblk | grep sdb
sdb             8:16   0    1G  0 disk 
├─sdb1          8:17   0  100M  0 part 
└─sdb2          8:18   0  100M  0 part

Now the PVs can be created:

1
2
3

pvcreate /dev/sdb1 /dev/sdb2
  Physical volume "/dev/sdb1" successfully created.
  Physical volume "/dev/sdb2" successfully created.

Creating VGs

Now you can create a VG from the 2 PVs:

1
2

vgcreate vg-zero /dev/sdb1 /dev/sdb2
  Volume group "vg-zero" successfully created

Let’s look at our existing PVs now:

1
2
3
4
5

pvs
  PV         VG      Fmt  Attr PSize  PFree 
  /dev/sda2  rhel    lvm2 a--  29.00g  4.00m
  /dev/sdb1  vg-zero lvm2 a--  96.00m 96.00m
  /dev/sdb2  vg-zero lvm2 a--  96.00m 96.00m

And check an individual PV in more detail:

1
2
3
4
5
6
7
8
9
10
11

pvdisplay /dev/sdb1
  --- Physical volume ---
  PV Name               /dev/sdb1
  VG Name               vg-zero
  PV Size               100.00 MiB / not usable 4.00 MiB
  Allocatable           yes 
  PE Size               4.00 MiB
  Total PE              24
  Free PE               24
  Allocated PE          0
  PV UUID               ViTFfu-jBZE-47Ie-d6qB-WTDt-fpNf-bPC00Y

Now let’s check our VGs:

1
2
3
4

vgs
  VG      #PV #LV #SN Attr   VSize   VFree  
  rhel      1   2   0 wz--n-  29.00g   4.00m
  vg-zero   2   0   0 wz--n- 192.00m 192.00m

And in more detail:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

vgdisplay vg-zero 
  --- Volume group ---
  VG Name               vg-zero
  System ID             
  Format                lvm2
  Metadata Areas        2
  Metadata Sequence No  1
  VG Access             read/write
  VG Status             resizable
  MAX LV                0
  Cur LV                0
  Open LV               0
  Max PV                0
  Cur PV                2
  Act PV                2
  VG Size               192.00 MiB
  PE Size               4.00 MiB
  Total PE              48
  Alloc PE / Size       0 / 0   
  Free  PE / Size       48 / 192.00 MiB
  VG UUID               gUf1XI-znJt-MM2T-V721-PNbe-uxcr-m3bVTe

Creating LVs

Next, I created a LV named inventory with a size of 128 MB, from the vg-zero VG:

1
2

lvcreate -n inventory -L 128MB vg-zero 
  Logical volume "inventory" created.

Look at the existing LVs:

1
2
3
4
5

lvs
  LV        VG      Attr       LSize   Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
  root      rhel    -wi-ao----  26.99g                                                    
  swap      rhel    -wi-ao----   2.00g                                                    
  inventory vg-zero -wi-a----- 128.00m

And check our newly created one:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

lvdisplay /dev/vg-zero/inventory 
  --- Logical volume ---
  LV Path                /dev/vg-zero/inventory
  LV Name                inventory
  VG Name                vg-zero
  LV UUID                jE4XRJ-ddlX-nM8s-NVSr-4cfK-F9O0-JcKecs
  LV Write Access        read/write
  LV Creation host, time rhel7, 2018-01-26 15:36:04 +0200
  LV Status              available
  # open                 0
  LV Size                128.00 MiB
  Current LE             32
  Segments               2
  Allocation             inherit
  Read ahead sectors     auto
  - currently set to     8192
  Block device           253:2

Before using the new LV, you need to create a filesystem on top of it. I went with XFS:

1
2
3
4
5
6
7
8
9
10

mkfs.xfs /dev/vg-zero/inventory 
meta-data=/dev/vg-zero/inventory isize=512    agcount=4, agsize=8192 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=1        finobt=0, sparse=0
data     =                       bsize=4096   blocks=32768, imaxpct=25
         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
log      =internal log           bsize=4096   blocks=855, version=2
         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0

Now prepare a location to mount it:

1

mkdir /mnt/inventory

I want it to mount at boot, so I added the following to /etc/fstab:

1

/dev/vg-zero/inventory     /mnt/inventory                xfs    defaults        0 0

Let’s break down the /etc/fstab fields a bit:

  • the first field represents the device, which can be specified by name, UUID, or label

  • the second field is the mount point of the device

  • the third is the filesystem type

  • fourth is for the mount options. The defaults option specifies the usage of default options: rw, suid, dev, exec, auto, nouser, async

  • the fifth is a backup flag for the dump utility. Use 1 to enable it and 0 to disable it

  • sixth represents the fsck automatic check at boot. Use 0 to disable it, 1 for the root filesystem, and 2 for other filesystems that might need the automatic checking

It’s a good practice to mount the devices you added to fstab before a reboot to ensure there are no errors: mount -a

Extending / reducing VGs

To add more space to a VG, you can add more PVs to it. Let’s see its current space:

1
2

vgdisplay vg-zero | grep Free
  Free  PE / Size       16 / 64.00 MiB

I created a new PV on /dev/sdb3 and added it to the VG:

1
2

vgextend vg-zero /dev/sdb3
  Volume group "vg-zero" successfully extended

The storage size has increased now:

1
2

vgdisplay vg-zero | grep Free
  Free  PE / Size       28 / 112.00 MiB

And to reduce the space of a VG, remove one or more PVs from it:

1
2

vgreduce vg-zero /dev/sdb3
  Removed "/dev/sdb3" from volume group "vg-zero"

To migrate the PEs used on a PV to other PVs in the VG, use pvmove:

1
2
3

pvmove /dev/sdb1
  /dev/sdb1: Moved: 0.00%
  /dev/sdb1: Moved: 100.00%

Resizing LVs

You can use lvresize for extending or shrinking LVs, or lvextend for extending only. There needs to be enough free space in the VG for growing the LV.

1
2
3
4

lvextend -L +50M /dev/vg-zero/inventory 
  Rounding size to boundary between physical extents: 52.00 MiB.
  Size of logical volume vg-zero/inventory changed from 128.00 MiB (32 extents) to 180.00 MiB (45 extents).
  Logical volume vg-zero/inventory successfully resized.

You also need to grow the filesystem to occupy the newly extended LV. Since I picked XFS, I’ll use xfs_growfs for this:

1
2
3
4
5
6
7
8
9
10
11

xfs_growfs /mnt/inventory/
meta-data=/dev/mapper/vg--zero-inventory isize=512    agcount=4, agsize=8192 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=1        finobt=0 spinodes=0
data     =                       bsize=4096   blocks=32768, imaxpct=25
         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
log      =internal               bsize=4096   blocks=855, version=2
         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
data blocks changed from 32768 to 46080

This was just for demo, in practice you want to resize both the LV and the filesystem with a single command:

1
2
3
4
5
6
7
8
9
10
11
12
13
14

lvextend -L 190M -r /dev/vg-zero/inventory 
  Rounding size to boundary between physical extents: 192.00 MiB.
  Size of logical volume vg-zero/inventory changed from 180.00 MiB (45 extents) to 192.00 MiB (48 extents).
  Logical volume vg-zero/inventory successfully resized.
meta-data=/dev/mapper/vg--zero-inventory isize=512    agcount=6, agsize=8192 blks
         =                       sectsz=512   attr=2, projid32bit=1
         =                       crc=1        finobt=0 spinodes=0
data     =                       bsize=4096   blocks=46080, imaxpct=25
         =                       sunit=0      swidth=0 blks
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
log      =internal               bsize=4096   blocks=855, version=2
         =                       sectsz=512   sunit=0 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
data blocks changed from 46080 to 49152

This time I specified an absolute size instead of adding a certain amount.

Removing LVs

When you’re done, you backed up your data or no longer need it, you can delete the LV. Unmount it first:

1
2

umount -v /mnt/inventory/
umount: /mnt/inventory (/dev/mapper/vg--zero-inventory) unmounted

Remove the LV:

1
2
3

lvremove /dev/vg-zero/inventory 
Do you really want to remove active logical volume vg-zero/inventory? [y/n]: y
  Logical volume "inventory" successfully removed

Remove the VG:

1
2

vgremove vg-zero 
  Volume group "vg-zero" successfully removed

Finally, also remove the PVs:

1
2
3

pvremove /dev/sdb1 /dev/sdb2
  Labels on physical volume "/dev/sdb1" successfully wiped.
  Labels on physical volume "/dev/sdb2" successfully wiped.

And don’t forget to delete any remaining fstab entries.

1
2
3
4
5
6
7
8
9

 ___________________________________
/ That secret you've been guarding, \
\ isn't.                            /
 -----------------------------------
        \   ^__^
         \  (oo)\_______
            (__)\       )\/\
                ||----w |
                ||     ||

Comments