Configuring RAID Arrays

1. Understanding RAID Levels

RAID Level	Description	Redundancy	Performance
RAID 0	Striping without parity. Data is split across multiple disks.	None	High read/write
RAID 1	Mirroring. Data is duplicated on two or more disks.	Full	Low read, high write
RAID 5	Striping with distributed parity. Requires at least 3 disks.	One disk	Medium read, medium write
RAID 6	Striping with dual distributed parity. Requires at least 4 disks.	Two disks	Medium read, medium write
RAID 10	Combination of mirroring and striping. Requires at least 4 disks.	Full	High read, high write

Note: RAID 0 offers no redundancy and is not recommended for critical data.

Warning: RAID levels with parity (5, 6) can be slow during rebuilds.

2. Creating RAID with mdadm

Example: Create a RAID 1 Array

sudo mdadm --create /dev/md0 \
  --level=1 \
  --raid-devices=2 \
  /dev/sdb1 \
  /dev/sdc1

Option	Values	Description
--create	`/dev/md0`	Device name for the RAID array
--level	`0`, `1`, `5`, `6`, `10`	RAID level to use
--raid-devices	`2`	Number of devices in the array
--name	`myarray`	Name for the array (optional)

Note: Ensure disks are unmounted and have no existing filesystems before creating the array.

Warning: Incorrect RAID configuration can lead to data loss. Always backup critical data before proceeding.

RAID Level Summary

0: Striped array, no redundancy
1: Mirrored array, high redundancy
5: Distributed parity, good read performance
6: Distributed dual parity, better fault tolerance
10: Striped mirrors, best performance and redundancy

3. Viewing RAID Status

Example: Check RAID status with mdadm

sudo mdadm --detail /dev/md0

Property	Description	Notes
State	Current state of the array (active, degraded, etc.)	Shows if array is healthy or has issues
Devices	Number of devices in the array	Shows total and active devices
Array Size	Total size of the RAID array	Sum of all device sizes
Used Dev Size	Size of each individual device	Shows size of each member disk
UUID	Unique identifier for the array	Used to identify and manage arrays

Note: The mdadm --detail command provides detailed information about the RAID array's configuration and status.

Example: List all RAID arrays

cat /proc/mdstat

Example: Monitor RAID rebuild progress

watch -n 1 cat /proc/mdstat

4. Adding Disks to RAID

Step	Action	Description
1	`mdadm --manage /dev/mdX --add /dev/sdY`	Add a new disk to an existing RAID array. Replace `mdX` with the array device and `sdY` with the new disk.
2	`watch -n 1 cat /proc/mdstat`	Monitor the rebuild process to ensure the disk is being added and data is syncing.
3	`mdadm --detail /dev/mdX`	Verify the status of the RAID array to confirm the new disk is active and rebuilding.

Example: Adding a Disk to RAID 1 Array

# Add a new disk to an existing RAID 1 array
mdadm --manage /dev/md0 --add /dev/sdb

# Monitor the rebuild process
watch -n 1 cat /proc/mdstat

Note: Ensure the new disk is of equal or greater size than the existing disks in the array to maintain redundancy.

5. Removing Failed Disks

Step	Action	Description
1	`mdadm --manage /dev/mdX --remove /dev/sdY`	Remove a failed disk from the array. Replace `/dev/mdX` with the array device and `/dev/sdY` with the failed disk.
2	`mdadm --manage /dev/mdX --add /dev/sdZ`	Add a new disk to replace the failed one. Ensure `/dev/sdZ` is a healthy disk.
3	`cat /proc/mdstat`	Verify the array rebuild process has started and is progressing.

Example: Removing and Replacing a Disk

# Check current array status
mdadm --detail /dev/md0

# Remove failed disk
mdadm --manage /dev/md0 --remove /dev/sdb

# Add new disk to array
mdadm --manage /dev/md0 --add /dev/sdc

# Monitor rebuild progress
watch cat /proc/mdstat

Note: Ensure the new disk is of equal or greater size than the failed one to maintain array integrity.

Warning: Removing a disk from an array that is not marked as failed can lead to data loss. Always confirm the disk status before removal.

6. Stopping RAID Arrays

Example: Stopping a RAID Array

# Stop the RAID array
mdadm --stop /dev/md0

# Or stop all active arrays
mdadm --stop --scan

Command	Description	Notes
mdadm --stop <array>	Stops a specific RAID array	Array must be inactive before stopping
mdadm --stop --scan	Stops all currently active RAID arrays	Useful for cleanup before system shutdown

Note: Ensure all data is safely written to disk before stopping arrays to prevent data loss.

Warning: Stopping a RAID array while it's in use can cause data corruption or loss. Always verify that no processes are using the array before stopping it.

7. Assembling RAID Arrays

Example: Assemble a RAID 1 Array

sudo mdadm --assemble /dev/md0 /dev/sdb1 /dev/sdc1

Option	Values	Description
--assemble	`--assemble`	Assemble a RAID array from existing devices
--scan	`--scan`	Scan for and assemble all arrays defined in configuration
--force	`--force`	Force assembly even if devices are not clean

Note: Ensure that the array configuration matches the devices being assembled to avoid data corruption.

Warning: Using --force can lead to data loss if the array is not properly aligned or has inconsistencies.

Example: Assemble with Configuration File

sudo mdadm --scan --assemble

Assembling Process Summary

Identify the RAID devices and their configuration.
Use mdadm --assemble to reconstruct the array from its component disks.
Verify that the array is correctly assembled and healthy.

8. Saving RAID Configuration

Example: Save RAID configuration using mdadm

# Save RAID configuration to /etc/mdadm/mdadm.conf
mdadm --detail --scan &gt; /etc/mdadm/mdadm.conf

# Verify the configuration file
cat /etc/mdadm/mdadm.conf

Command	Description	Notes
mdadm --detail --scan	Scans all RAID devices and generates configuration entries	Used to create the mdadm.conf file
mdadm --examine	Examines a RAID device for metadata and configuration details	Helps in identifying devices and their status

Note: The /etc/mdadm/mdadm.conf file should be backed up to ensure that RAID arrays can be properly reassembled after a system reboot.

Warning: Ensure that the /etc/mdadm/mdadm.conf file is correctly updated after any changes to the RAID array configuration to avoid data loss or misconfiguration.

9. Monitoring RAID Health

Example: Check RAID Status with mdadm

mdadm --detail /dev/md0

Property	Description	Notes
State	Current state of the array (active, degraded, failed)	Indicates overall health status
Devices	Number of devices in the array	Should match expected count
Active Devices	Number of active devices	Should match expected count
Failed Devices	Number of failed devices	Should be zero for healthy arrays
Spare Devices	Number of spare devices available	Can be used to replace failed drives

Example: View RAID Events with journalctl

journalctl -u mdmonitor --since "1 hour ago"

Note: Use journalctl to check for RAID-related events or errors in systemd logs.

Example: Monitor RAID with /proc/mdstat

cat /proc/mdstat

Field	Description	Interpretation
md0	RAID array name	Array is active and running
[UU]	Device status (U = up)	All devices are healthy
[UUU]	Device status (U = up)	All devices are healthy

10. Growing RAID Arrays

Example: Adding a new drive to an existing RAID 5 array

# Check current array status
mdadm --detail /dev/md0

# Add new drive to the array
mdadm --add /dev/md0 /dev/sdc

# Monitor rebuild progress
watch -n 1 cat /proc/mdstat

Command	Description	Notes
mdadm --add <array> <device>	Adds a new drive to an existing array	Drive must be same size or larger than existing drives
mdadm --grow <array> --raid-devices=<count>	Changes the number of devices in an array	Can be used to expand RAID 5 or 6 arrays
mdadm --grow <array> --size=<size>	Changes the size of an array	Use with caution, may cause data loss if not properly aligned

Note: Growing a RAID array requires sufficient space on the new drive to accommodate the expanded array size. Ensure that all drives in the array are healthy before attempting to grow.

Warning: Expanding a RAID array can be risky and may lead to data loss if not done correctly. Always backup your data before performing any array expansion operations.