Simple implementation of centos disk management RAID

Create a RAID0 with a size of 10 GB
1. Create two shards for RAID 0

Use/dev/sde1/and dev/sdf1 to create a partition. The partition capacity is 5 GB, And the partition is converted to fd.

123 # fdisk/dev/sde # fdisk/dev/sdf # fdisk-l/dev/sd {e, f}

Figure 1 and 2

Note: Do not format it after saving and exiting, because we cannot format a block device at the bottom of software raid, but format the software raid itself.

2. Let the kernel re-read the Partition Table

123 # partx-a/dev/sde1/dev/sde # partx-a/dev/sdf1/dev/sdf # cat/proc/partitions

Figure 3

3. Create the/dev/sde1 and/dev/sdf1 partitions as RAID 0.

Check whether there are related files under/proc/mdstat before creation.

1 # cat/proc/mdstat

Figure 4

1 # ls-l/dev/md *

Figure 5

Since md0 has not been used, let's start from 0!

1 # mdadm-C-a yes-n 2-l 0/dev/sd {e, f} 1

4. Create an md Device

1 # mdadm-C/dev/md00-a yes-l 0-n 2/dev/sd {e, f} 1

Figure 6

Description of md-related files ------> version number, device file, ID number -----------> ASCII code file, saving the ing relationship between all md devices on the current host, which disk partitions are used as md devices?

1 # ls-l/dev/md *

Figure 7

12 # cd/dev/md # file md-device-map

1 # cat md-device-ma

Figure 8

5. format the md Device

Note: After the md device is created, You Can format the md device. The md device is an independent device.

# Mke2fs-t ext4/dev/md0

1 # mke2fs-t ext4/dev/md0

Figure 9

Vi. Mounting

12345 # mkdir/raid0 # mount/dev/md0/raid0 # cd/raid0/# ls # mount

7. Monitor md0 devices and learn about md

1 # mdadm-D/dev/md0 ------ "displays details of md0

Figure 11

Create a RAID1 with a size of 20 GB

1. Create two 20 partitions respectively in the system

Figure 1

II,

1/# fdisk-l/dev/sd {B, c} ask the kernel to re-read and identify the partition/

Figure 2

3. Create an md Device

1 # mdadm-C/dev/md1-a yes-l 1-n 2/dev/sd {B, c} 1

Figure 3

Check whether the creation is successful

1 # cat/proc/mdstat

Figure 4

4. format the md Device

1 # mke2fs-t ext4/dev/md1-L RAID_1

Figure 5

VI,

Figure 6

Figure 7

Test the redundancy of RAID1.

1234 # cd/mnt # ls; pwd # cp/etc/issue./# mount

Figure 8

2. If a hard disk of RAID 1 breaks down, can the/mnt mount point be used?

For RAID1, if a hard disk is broken, does it really not affect data access?

Iii. Destruction

12 # mdadm/dev/md1-f/dev/sdb1 # mdadm-D/dev/md1

Figure 9

1 # cat issue

Figure 10

Copy the file and try to test the write operation.

Figure 11

Test read Operations

Figure 12

OK, no problem!

3. Remove a bad Disk

1 # mdadm/dev/md1-r/dev/sdb1

Figure 13

4. If the/dev/sd disk has been repaired, how can we add it to its RAID group?

12 # mdadm/dev/md1-a/dev/sdb1 # mdadm-D/dev/md1

Figure 14

1 # mdadm-D/dev/md1

Data Reconstruction successful

The second disk in Raid 1 is damaged.

I,

12 # mdadm/dev/md1-f/dev/sdc1 # mdadm-D/dev/md1

Figure 16

2. Read and Write the/mnt directory

1234 # ls/mnt/# cat/mnt/issue-n # echo "I love linux so much">/mnt/love.txt # cat/mnt/love.txt-n

Figure 17

Figure 18

Figure 19

Three disks in Raid 1 are used as backup disks.

1 # mdadm-C/dev/md2-a yes-l 1-n 2-x 1/dev/sd {B, c, d} 2

Figure 20

# Fdisk-l/dev/sd {B, c, d}

Figure 21

Format the md Device

1 # mke2fs-t ext4/dev/md2

Figure 22

1 mount # mount/dev/md2/mnt/12 # cp/etc/fstab/mnt/# cat/mnt/fstab

Figure 23

Simulate disk corruption

Figure 24

12 # mdadm/dev/md2-r/dev/sdb2 # mdadm-D/dev/md2

Figure 25

Check that the file is intact.

Figure 26

Create a RAID5 with a size of 10 GB

1 # ls/dev/md *

1 # mdadm-C/dev/md3-a yes-l5-n3/dev/sd {B, c, d} 3

Figure 27

Figure

28

1 # cat/proc/mdstat

Figure 29

Create an md Device

1 # mdadm-D/dev/md3

Figure 30

Format the md Device

1 # mke2fs-t ext4/dev/md3

Figure 31

123 # mount/dev/md3/mnt/# cp/etc/issue/mnt/issue # cat/mnt/issue

Figure 32

Damaged simulated/dev/sd Disk

12 # mdadm/dev/md3-f/dev/sdb3 # mdadm-D/dev/md3

Figure 33

Data can be accessed normally.

Figure 34

1 # mdadm-D/dev/md3

Figure 35

What will happen if one piece breaks down? You don't know how to try it !!!!

12 # mdadm/dev/md3-f/dev/sdc3 # cat-n/mnt/issue

Figure 36

Figure 37

1 # mdadm-D/dev/md3

At this point, the RAID-related practices have ended. This experiment is only for learning and communication purposes. Do not imitate it in the actual production environment!

All operations in this article are related to RAID and have nothing to do with me!