Managing RAID with mdadm on Linux

mdadm is a tool for managing RAID arrays on Linux systems. Here are some basic tasks you can perform with mdadm:

1. Create a RAID array

To create a RAID array, use the mdadm --create command. For example, to create a RAID 5 array with three disks (/dev/sd[b-d]):

1

sudo mdadm --create /dev/md0 --level=5 --raid-devices=3 /dev/sdb /dev/sdc /dev/sdd

2. Add a disk to an existing RAID array

To add a disk to an existing RAID array, use the mdadm --add command. For example, to add /dev/sde to /dev/md0:

1

sudo mdadm --add /dev/md0 /dev/sde

3. Remove a disk from a RAID array

To remove a disk from a RAID array, first mark the disk as faulty, then remove it. For example, to remove /dev/sdc from /dev/md0:

1
2

sudo mdadm --fail /dev/md0 /dev/sdc
sudo mdadm --remove /dev/md0 /dev/sdc

4. Monitor the RAID array status

To check the status of your RAID array, use the mdadm --detail command. For example, to see the details of /dev/md0:

1

sudo mdadm --detail /dev/md0

5. Repair a RAID array

To repair a RAID array, you need to add a new disk and reassign the array. For example, after adding /dev/sdf:

1

sudo mdadm --add /dev/md0 /dev/sdf

The rebuild process will begin automatically.

6. Save RAID configuration

To ensure that the RAID configuration is loaded at boot, save the current configuration to the mdadm config file. This can be done with the mdadm --detail --scan command and redirecting the output to the configuration file:

1

sudo mdadm --detail --scan | sudo tee -a /etc/mdadm/mdadm.conf

Here’s an example output of the configuration:

1
2
3

ARRAY /dev/md/md2 metadata=1.2 name=md2 UUID=a1079a27:f9562ae5:a195fe0d:fbe4a305
ARRAY /dev/md/md3 metadata=1.2 name=md3 UUID=30a7e838:c1e7d6ea:03255854:5b607a73
ARRAY /dev/md/srvdatastoreplt01:6 metadata=1.2 name=srvdatastoreplt01:6 UUID=602da39c:111b2659:8fdccdaf:9f4998b4

Then, update initramfs:

1

sudo update-initramfs -u

7. Stop a RAID array

To stop a RAID array, use the mdadm --stop command. For example, to stop /dev/md0:

1

sudo mdadm --stop /dev/md0

8. Delete a RAID array

To delete a RAID array, first stop the array and then remove the device. For example:

1
2

sudo mdadm --stop /dev/md0
sudo mdadm --remove /dev/md0

9. Check current RAID configuration

To check the current RAID configuration, you can use:

1

sudo mdadm --examine /dev/sd[b-d]

Here’s an example output from examining a RAID device:

1

sudo mdadm --examine /dev/nvme1n1p2

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

/dev/nvme1n1p2:
          Magic : a92b4efc
        Version : 1.2
    Feature Map : 0x0
     Array UUID : a1079a27:f9562ae5:a195fe0d:fbe4a305
           Name : md2
  Creation Time : Mon Apr 29 09:34:15 2024
     Raid Level : raid1
   Raid Devices : 2
...
   Device Role : Active device 1
   Array State : AA ('A' == active, '.' == missing, 'R' == replacing)

Integrity Check

mdadm also allows checking the integrity of a RAID device by running a consistency check.

1. Check the current status of the RAID

Before performing a consistency check, review the current status of the RAID array:

1

cat /proc/mdstat

2. Start the RAID check

To start a consistency check on a RAID array, you can use the following command, replacing /dev/mdX with the name of the RAID device (e.g., /dev/md0):

1

echo check > /sys/block/mdX/md/sync_action

3. Monitor the progress

You can monitor the progress of the check using:

1

cat /proc/mdstat

4. Review the results

After the check completes, review any errors or issues with:

1

mdadm --detail /dev/mdX

Optional: Repair detected errors

If errors were found during the check, you can force a repair by running:

1

echo repair > /sys/block/mdX/md/sync_action

Summary of Commands

  • View RAID status:
1

cat /proc/mdstat
  • Start consistency check:
1

echo check > /sys/block/mdX/md/sync_action
  • Monitor progress:
1

cat /proc/mdstat
  • View RAID details:
1

mdadm --detail /dev/mdX

For more detailed options and advanced usage, consult the mdadm manual:

1

man mdadm