mdadm

mdadm (Multiple Devices Admin) is a Linux utility used for managing and monitoring software RAID arrays. It is an essential tool for configuring, creating, managing, and troubleshooting RAID arrays using the MD (Multiple Devices) subsystem in Linux. mdadm works with both RAID devices managed by the kernel's RAID functionality and RAID1, RAID5, RAID6, RAID10, and RAID0 configurations.

Key Features of mdadm:

  1. Creating RAID Arrays:

    • mdadm can be used to create new software RAID arrays, specify RAID level (RAID 0, 1, 5, 6, etc.), and assign devices to the array.

  2. Managing RAID Arrays:

    • It supports operations such as adding, removing, or replacing disks in a RAID array. Additionally, it helps in managing RAID components, setting RAID array parameters, and more.

  3. Monitoring RAID Arrays:

    • mdadm can monitor the health of RAID arrays, alerting users to potential failures or degraded arrays. It can also check for resync/rebuild progress.

  4. Resync and Rebuild:

    • When a disk fails in a RAID array, mdadm helps rebuild the array by resyncing the data. It also manages the process of replacing failed disks and restoring redundancy.

  5. RAID Array Maintenance:

    • It can trigger operations such as array reshaping, expanding, or shrinking, as well as performing various maintenance tasks.

  6. Logging and Alerting:

    • mdadm can send email alerts for RAID status changes, array failures, and rebuild completions.

Common mdadm Commands and Their Usage:

  1. Creating a RAID Array:

    • To create a RAID 1 array with two devices (/dev/sda and /dev/sdb):

      mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sda /dev/sdb

    • This command creates a RAID 1 array (/dev/md0) using two devices.

  2. Viewing RAID Array Information:

    • To check the status of RAID arrays:

      mdadm --detail /dev/md0

    • This provides detailed information about the array /dev/md0, including the devices used, RAID level, and array health.

  3. Assembling a RAID Array:

    • To assemble a previously created RAID array, you can use:

      mdadm --assemble /dev/md0 /dev/sda /dev/sdb

    • This command is used when reassembling a degraded or missing RAID array after booting.

  4. Adding a New Disk to a RAID Array:

    • To add a new disk (/dev/sdc) to an existing RAID 5 array (/dev/md0):

      mdadm --add /dev/md0 /dev/sdc

    • This adds /dev/sdc to the existing RAID array, which will then start the resynchronization process to integrate the new disk into the array.

  5. Removing a Disk from a RAID Array:

    • To remove a failed or unused disk from an array:

      mdadm --remove /dev/md0 /dev/sda

    • This command removes /dev/sda from the RAID array /dev/md0.

  6. Monitoring RAID Array Status:

    • You can use mdadm to monitor the status of your RAID array continuously. To monitor the status and automatically send alerts on failures:

      mdadm --monitor --scan --daemonise

    • This runs mdadm in monitoring mode, which checks the array's health and will send alerts if any problems arise.

  7. Checking the Health of All RAID Arrays:

    • To check the health of all RAID arrays:

      mdadm --examine --scan

  8. Growing a RAID Array:

    • To expand the size of a RAID array after adding new disks (e.g., adding /dev/sdc to a RAID 5 array):

      mdadm --grow /dev/md0 --raid-devices=4 /dev/sdc

    • This increases the number of devices in the array, initiating a reshaping process.

  9. Stopping and Assembling RAID Arrays:

    • If you want to stop a RAID array (e.g., /dev/md0), use:

      mdadm --stop /dev/md0

    • This will stop the RAID array. It is commonly used before removing the array from the system.

  10. Rebuilding a RAID Array:

  • If a disk has failed and needs rebuilding, mdadm can monitor and manage this:

    mdadm --manage /dev/md0 --add /dev/sda

  1. Creating RAID Configuration File:

    • To generate the configuration file for RAID arrays:

      mdadm --detail --scan > /etc/mdadm/mdadm.conf

Example Output from mdadm:

$ mdadm --detail /dev/md0
/dev/md0:
        Version : 1.2
  Creation Time : Sun Mar  7 10:25:39 2021
     Raid Level : raid5
     Array Size : 9766481920 (9311.94 GiB 10000.16 GB)
  Used Dev Size : 4883240960 (4655.97 GiB 5000.08 GB)
   Raid Devices : 3
  Total Devices : 3
    Persistence : Superblock is persistent

    Update Time : Wed Mar  9 12:53:01 2021
          State : clean
 Active Devices : 3
Working Devices : 3
 Failed Devices : 0
  Spare Devices : 0

           Name : server:0  (local to host server)
           UUID : c29f1f6e:0eaa7d5b:fd5bb16d:b38a30b4
         Events : 184

    Number   Major   Minor   RaidDevice State
       0     8        1        0      active sync   /dev/sda
       1     8       17        1      active sync   /dev/sdb
       2     8       33        2      active sync   /dev/sdc

Key Fields:

  • Version: Version of the RAID superblock used.

  • Creation Time: When the array was created.

  • RAID Level: RAID type (RAID 5 in this example).

  • Array Size: Total size of the array.

  • Used Dev Size: Size of each individual device in the array.

  • Active Devices: Number of devices currently active in the array.

  • State: The current state of the array (e.g., clean, resync, degraded).

  • Name: The name of the RAID array.

  • UUID: Unique identifier for the array.

  • Devices: Information about each device in the array, including its status.

Conclusion:

mdadm is a powerful tool for creating, managing, and monitoring software RAID arrays in Linux. It provides administrators with the necessary capabilities to manage disk redundancy, array health, and rebuilding processes, while also offering monitoring and alerting features to ensure that RAID arrays remain in optimal condition. The flexibility and reliability of mdadm make it the go-to tool for managing software RAID in Linux systems.

PreviouslvresizeNextmkfs.gfs2

Last updated