centos7 uses mdadm soft raid

1. The syntax of
    mdadm mdadm --detail /dev/md0 View the raid details of /dev/md0
    mdadm --create --auto=yes /dev/md[0-9] --level=[015] --raid- devices=N --spare-devices=N /dev/sdx
--create: parameters for the new raid
--auto=yes: determine the new software disk array devices that follow, namely /dev/md0, /dev/md1, etc.
--level=[015]: Set the level of the disk array
--detail: List the detailed information of the disk array device connected later
--raid-devices=N: Use several disks as the device of the disk array
--spare- devices=N: use several disks as alternate devices
    mdadm --manage /dev/md[0-9] [--add /dev/sdxN] [--remove /dev/sdxN] [--fail /dev/sdxN ]
--add: The following device will be added to this md
--remove: The following device will be removed from this md
--fail: The latter device will be set to an error state

2. A raid5 instance
    fdisk /dev/ sda
    uses the n parameter to create 6 partitions with a size of 300M, dev/sda5-10
    w, and use partprobe to refresh the partition table after saving.
    Then use sda5-9 to form a raid5 array
    mdadm --create --auto=yes /dev/md0 --level=5 --raid-devices=4 --spare-devices=1 /dev/sda{5,6,7 ,8,9}
    View the status of the system software disk array
    cat /proc/mdstat
    format and mount with raid
    mkfs -t ext3 /dev/md0
    mkdir /mnt/raid
    mount /dev/md0 /mnt/raid
    View md0 software disk Array details
    mdadm --detail /dev/md0
   
    emulate rescue raid5
    set /dev/sda6 to fail state (alternate /dev/sda9 will automatically make up)
    mdadm --manage /dev/md0 --fail /dev/ sda6
    At this time, quickly enter the following command to view the process of disk replenishment (the speed should be faster, or you will see the result of the replenishment)
    mdadm --detail /dev/md0
    cat /proc/mdstat
   
    After waiting for a while, repeat the above input You can see the result after adding two commands
    mdadm --detail /dev/md0
    cat /proc/mdstat
   
    Delete the wrong /dev/sda6 and add /dev/sda10 as a raid5 spare disk mdadm --manage /dev/md0 --add /dev/ sda10
    --remove /dev/sda6 View the uuid
   
of
    /dev/md
    mdadm --detail /dev/md0 | grep -i uuid
    You can see the following line of
    UUID: 048383b6:c69991e2:f9c2bef7:39d45ef1
   
    Next, set mdadm.conf
    vi /etc/mdadm.conf
    to it Add a line of
    ARRAY /dev/md0 048383b6:c69991e2:f9c2bef7:39d45ef1
   
    to set the boot to automatically mount
    vi /etc/fstab
    Add a line to the end
    /dev/md0 /mnt/raid ext3 defaults 1 2

    Test the mount set in /etc/fstab Device
    umount /dev/md0
    mount -a
    df /mnt/md0
   
4. Close the software raid
    first unmount /dev/md0
    umount /dev/md0
    and then delete the statement previously added in /etc/fstab to
   
    directly close /dev/md0
    mdadm --stop /dev/md0
   
    to see if the raid still exists
    cat /proc/mdstat
   
    delete /etc/mdadm.conf The statement written before in
    vi /etc/mdadm.conf