Storage Handling Scenario 1
Requirement 1: Initial Storage Configuration
Our task is to set up storage partitions based on the specifications below. Once the partitions are in place, we will create the necessary volume groups and logical volumes to distribute the required storage.
Here are the specifications:
| Type | Mount Point | Size (GB) |
|---|---|---|
| assets | /mnt/assets | 500 |
| assets (replica) | /mnt/assets_replica | 500 |
| media | /mnt/media | 250 |
| tools | /mnt/tools | 250 |
Before rolling up our sleeves and start partitioning, let’s evaluate the available storage space:
$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 1T 0 disk
├─sda1 8:1 0 900M 0 part /boot
├─sda2 8:2 0 8G 0 part [SWAP]
├─sda3 8:3 0 70G 0 part
│ ├─rhel-root 253:0 0 60G 0 lvm /
│ └─rhel-home 253:1 0 10G 0 lvm /home
└─sda4 8:4 0 512G 0 part
sdb 8:16 0 1T 0 disk
sdc 8:32 0 512G 0 disk
sdd 8:48 0 512G 0 disk
sde 8:64 0 512G 0 disk
From the output, we can deduce which disks are available and plan the space allocation based on our storage needs.
| Physical Devices | Size (TB) | Size (GB) | Size (MiB) |
|---|---|---|---|
| /dev/sda | 1 | 1000 | 1,024,000 |
| /dev/sdb | 1 | 1000 | 1,024,000 |
| /dev/sdc | 0.5 | 500 | 512,000 |
| /dev/sdd | 0.5 | 500 | 512,000 |
| /dev/sde | 0.5 | 500 | 512,000 |
We’ll begin by partitioning our devices. As /dev/sda is already partitioned, we can skip the initialization step (mklabel gpt). Let’s first determine the end point of the existing partition.
Given the output:
$ parted /dev/sda unit MiB print
Model: QEMU QEMU HARDDISK (scsi)
Disk /dev/sda: 1048576MiB
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags:
Number Start End Size File system Name Flags
1 1MiB 1025MiB 1024MiB xfs /boot boot
2 1025MiB 9225MiB 8200MiB linux-swap(v1) swap
3 9225MiB 819225MiB 810000MiB LVM partition lvm
we can identify the ending point.
To create a new partition from there:
parted /dev/sda mkpart primary 819225MiB 1331225MiB
parted /dev/sda set 4 lvm on
Let’s verify that the new partition has been created successfully:
$ parted /dev/sda unit MiB print
Model: QEMU QEMU HARDDISK (scsi)
Disk /dev/sda: 1048576MiB
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags:
Number Start End Size File system Name Flags
1 1MiB 1025MiB 1024MiB xfs /boot boot
2 1025MiB 9225MiB 8200MiB linux-swap(v1) swap
3 9225MiB 819225MiB 810000MiB LVM partition lvm
4 819225MiB 1331225MiB 512000MiB LVM partition lvm
Looks good. For the other drives, since they are unallocated, we’ll start our partition numbering from 1:
$ parted /dev/sdb mklabel gpt
Warning: The existing disk label on /dev/sda will be destroyed and all data on this disk will be lost. Do you want to continue?
Yes/No? Yes
Proceeding with the partition creation:
parted /dev/sdb mkpart 1MiB 512000MiB
parted /dev/sdb set 1 lvm on
To initialize your partitions as physical volumes for use with LVM, run the commands below:
pvcreate /dev/sda4 /dev/sdb1/ /dev/sdc1 /dev/sdd1 /dev/sde1
udevadm settle
If executed successfully, the following output should confirm the disk partitions:
$ lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 1T 0 disk
├─sda1 8:1 0 900M 0 part /boot
├─sda2 8:2 0 8G 0 part [SWAP]
├─sda3 8:3 0 70G 0 part
│ ├─rhel-root 253:0 0 60G 0 lvm /
│ └─rhel-home 253:1 0 10G 0 lvm /home
└─sda4 8:4 0 512G 0 part
sdb 8:16 0 1T 0 disk
└─sdb1 8:17 0 512G 0 part
sdc 8:32 0 512G 0 disk
└─sdc1 8:33 0 512G 0 part
sdd 8:48 0 512G 0 disk
└─sdd1 8:49 0 512G 0 part
sde 8:64 0 512G 0 disk
└─sde1 8:65 0 512G 0 part
The detailed partitioning and unpartitioned space on the devices can be summarized as:
| Physical Devices | Physical Volumes | Start (MiB) | End (MiB) | Size (MiB) | Unpartitioned (MiB) |
|---|---|---|---|---|---|
| /dev/sda | /dev/sda1 | 1 | 1,025 | 1,024 | 1,047,552 |
| /dev/sda | /dev/sda2 | 1,025 | 9,225 | 8,200 | 1,039,352 |
| /dev/sda | /dev/sda3 | 9,225 | 819,225 | 81,000 | 958,352 |
| /dev/sda | /dev/sda4 | 819,225 | 1,331,225 | 512,000 | 446,352 |
| /dev/sdb | /dev/sdb1 | 1 | 512,000 | 512,000 | 0 |
| /dev/sdc | /dev/sdc1 | 1 | 512,000 | 512,000 | 0 |
| /dev/sdd | /dev/sdd1 | 1 | 512,000 | 512,000 | 0 |
| /dev/sde | /dev/sde1 | 1 | 512,000 | 512,000 | 0 |
Given the replication requirement for assets, and to maintain redundancy across disks, our Volume Groups (VGs) need to span multiple disks. Here’s the proposed VG and associated physical volume layout:
| Partitions | vg1 | vg2 | vg3 |
|---|---|---|---|
| /dev/sda4 | ✔️ | ||
| /dev/sdb1 | ✔️ | ||
| /dev/sdc1 | ✔️ | ||
| /dev/sdd1 | ✔️ | ||
| /dev/sde1 | ✔️ | ||
| Total (GB) | 500 | 1000 | 1000 |
To create the Volume Groups, execute the below commands:
vgcreate -s 1MiB vg1 /dev/sda4
vgcreate -s 1MiB vg2 /dev/sdc1 /dev/sdd1
vgcreate -s 1MiB vg3 /dev/sdb1 /dev/sde1
Now is the time to create the Logical Volumes to cater our needs.
| Volume Group | Logical Volume | Utilised Space (MiB) | Available Space (MiB) |
|---|---|---|---|
| vg1 | assets_lv | 512,000 | 0 |
| vg2 | assets_replica_lv | 512,000 | 0 |
| vg3 | media_lv | 256,000 | 256,000 |
| vg3 | tools_lv | 256,000 | 0 |
lvcreate -n assets_lv -L 500G vg1
mkfs -t xfs /dev/vg1/assets_lv
lvcreate -n assets_replica_lv -L 500G vg2
mkfs -t xfs /dev/vg2/assets_replica_lv
lvcreate -n media_lv -L 250G vg3
mkfs -t xfs /dev/vg3/media_lv
lvcreate -n tools_lv -L 250G vg3
mkfs -t xfs /dev/vg3/tools_lv
If everything goes well, we should see the below output.
$ lsblk --fs
NAME FSTYPE FSVER LABEL UUID FSAVAIL FSUSE% MOUNTPOINTS
sda
├─sda1 xfs f3300511-982f-4996-99ad-c62b499dd4f3 900M 10% /boot
├─sda2 swap 1 cd1b7f44-49cc-47ac-8896-872af7a19220 [SWAP]
├─sda3 LVM2_member LVM2 001 Ju5BPe-u0je-YiDY-ygKm-UZ6m-cbj3-hdmvFU
│ ├─rhel-root xfs 0447ab1a-c219-4929-9ef5-0a64be0e163a 60.0G 10% /
│ └─rhel-home xfs 0b7b551f-82b2-4158-b49b-34cd8142d28c 10.0G 1% /home
└─sda4 LVM2_member LVM2 001 UUID_VG1
sdb
└─sdb1 LVM2_member LVM2 001 UUID_VG3
└─vg1-assets_lv xfs LVM2 001 0a33c8b6-96f9-4e5a-a51d-dcb43ca487f8
sdc
└─sdc1 LVM2_member LVM2 001 UUID_VG2
└─vg2-assets_replica_lv xfs LVM2 001 2d679c3f-1ad5-4b64-81b3-353dbe3a3d18
sdd
└─sdd1 LVM2_member LVM2 001 UUID_VG2
└─vg3-media_lv xfs LVM2 001 9d2f8a2c-3ef9-49df-bb85-63f7c84b9b10
sde
└─sde1 LVM2_member LVM2 001 UUID_VG3
└─vg3-tools_lv xfs LVM2 001 7e235afc-f625-4d9d-9798-b3b4e4c20f7f
Let’s create mount points with the below reference table.
| Type | Mount Point | Logical Volume | Size (GB) |
|---|---|---|---|
| assets | /mnt/assets | assets_lv | 500 |
| assets (replica) | /mnt/assets_replica | assets_replica_lv | 500 |
| media | /mnt/media | media_lv | 250 |
| tools | /mnt/tools | tools_lv | 250 |
mkdir -p /mnt/assets /mnt/assets_replica /mnt/media /mnt/tools
We can mount these to the Logical Volumes created accordingly.
mount /dev/vg1/assets_lv /mnt/assets
mount /dev/vg2/assets_replica_lv /mnt/assets_replica
mount /dev/vg3/media_lv /mnt/media
mount /dev/vg3/tools_lv /mnt/tools
You can verify if you can access the mount pionts.
ls -l /mnt/assets
mount | grep assets
But this is not persistent, so a reboot will unmount it. To persist the mount we shall use the /etc/fstab file.
$ sudo -i
$ vim /etc/fstab
/dev/vg1/assets_lv /mnt/assets xfs defaults 0 0
/dev/vg2/assets_replica_lv /mnt/assets_replica xfs defaults 0 0
/dev/vg3/media_lv /mnt/media xfs defaults 0 0
/dev/vg3/tools_lv /mnt/tools xfs defaults 0 0
You can use the below command to mount all filesystems mentioned in the /etc/fstab file.
mount -a
To double confirm if the mount points are still mounted, you can also reboot the machine and verify.
systemctl reboot
ls -l /mnt/assets
In the next scenario, we shall see how to extend the Logical Volumes by creating more partitions and adding them to the Volume Groups.