Managing RAID in Linux

In this article I’ll show the various raid levels of a practice way and the basic manage to add or remove disks to our raid device. First of all we have to partition our hard disk with the type of partition code fd, used for Linux RAID software

Create the partition and select the partition type for every hard disk:

root@debian6:~# fdisk /dev/sdb

WARNING: DOS-compatible mode is deprecated. It's strongly recommended to
switch off the mode (command 'c') and change display units to
sectors (command 'u').

Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-5221, default 1):
Using default value 1
Last cylinder, +cylinders or +size{K,M,G} (1-5221, default 5221):
Using default value 5221

Command (m for help): t
Selected partition 1
Hex code (type L to list codes): fd
Changed system type of partition 1 to fd (Linux raid autodetect)

Command (m for help): w
The partition table has been altered!

Calling ioctl() to re-read partition table.
Syncing disks.

• RAID 0 

root@debian6:~# mdadm --create /dev/md0 --level=0 --raid-devices=2 /dev/sdb1 /dev/sdc1

RAID 0 stores the data between two or more disks. With this raid level win in speed of read/write, but if one of the disks fails the data will be lose.

• RAID 1

root@debian6:~# mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sdb1 /dev/sdc1

RAID 1 or mirror stores a copy of a disk in one or more disks, with this way if one disk fail, the other disks can replace the failed disk.

• RAID 4/5 

root@debian6:~# mdadm --create /dev/md0 --level=4 --raid-devices=3 /dev/sdb1 /dev/sdc1 /dev/sdd1

The RAID 4 we have to user a minimum of three disks. This raid level keep the advantage of raid 0 and raid 1, storing the data between the different disks but using a dedicated disk to parity (control errors) to restore the data if one of the disks fails.

RAID 5 it’s the same that RAID 4, but the parity it’s stored between the different disks.

For RAID 4 or 5 the size of device will be (Number of disks -1).

• RAID 6

root@debian6:~# mdadm --create /dev/md0 --level=6 --raid-devices=4 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1

The minimum of disks used for RAID 6 is 4. In RAID 4 or 5 if two disks fails we have lost the stored data. With RAID 6 increase the amount of parity data between disks. The size will be (Number of disks – 2).

• RAID 10

root@debian6:~# mdadm --create /dev/md0 --level=10 --raid-devices=4 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1

RAID 1+0 is a combination of raid 0 with raid 1. The minimum number of disks used is 4.

• Saving our RAID configuration

root@debian6:~# mdadm --detail --scan >> /etc/mdadm/mdadm.conf

• Stopping a RAID device

root@debian6:~# mdadm --stop /dev/md0

• Starting a raid device

root@debian6:~# mdadm --assemble --scan

• Replacing a disk

1.- Select a disk as failure disk before removing it:

root@debian6:/mnt# mdadm --manage --set-faulty /dev/md0 /dev/sdc1

2.- Remove the disk:

root@debian6:/mnt# mdadm --manage --remove /dev/md0 /dev/sdc1

3.- Add the new hard disk, partition it and add to the raid device:

root@debian6:/mnt# mdadm --manage --add /dev/md0 /dev/sdc1

• Expand the raid devices

1.- Partition the new disk and add to the raid device:

root@debian6:/mnt# mdadm --manage --add /dev/md0 /dev/sdf1

2.- Expand the number of partitions:

root@debian6:/mnt# mdadm --grow --raid-devices=5 /dev/md0

3.- Resize the filesystem (ext2/ext3/ext4):

root@debian6:/mnt# umount /dev/md0
root@debian6:/mnt# fsck -f /dev/md0
root@debian6:/mnt# resize2fs /dev/md0

Before the resize:

After:

For more information you can visit: https://raid.wiki.kernel.org