20 Essential Linux Disk Management Techniques Every Sysadmin Must Master

1. Basic Inspection and Information Retrieval

Use lsblk -o NAME,SIZE,FSTYPE,MOUNTPOINT to display a tree view of disks, partitions, and mount points, enabling quick identification of storage layout.

Identify device UUIDs and filesystem types with blkid /dev/sda1, which helps avoid mount errors caused by changing device names and is essential for reliable /etc/fstab configuration.

Monitor disk health via SMART by running smartctl -a /dev/sda, which reports temperature, bad‑sector count, and other critical metrics for early fault prediction.

2. Partitioning and Formatting

Choose fdisk for traditional MBR partitions (≤2 TB) and parted for GPT partitions (>2 TB). Example GPT creation:

parted /dev/sdb mklabel gpt mkpart primary xfs 0% 100%

Use mkfs.xfs -f /dev/sdb1 for forced XFS creation or mkfs.ext4 -i 8192 /dev/sdb2 to adjust inode density based on expected file count.

3. Mounting and Automatic Mount

Mount by UUID to ensure stability across device name changes:

# Get UUID
blkid /dev/sdb1
# /etc/fstab entry
UUID=1234-5678 /data xfs defaults 0 0

Temporary mounts with performance options: mount -o noatime,nodev /dev/sdc1 /mnt/tmp Advanced /etc/fstab parameters, e.g., network storage: nas:/share /mnt/nfs nfs rw,hard,intr 0 0 and disk‑error protection:

/dev/sdb1 /data ext4 defaults,nofail 0 2

4. Space Analysis and Cleanup

Quick directory size overview: du -h --max-depth=1 /var | sort -hr Log management:

journalctl --vacuum-size=200M   # limit journal size
logrotate -f /etc/logrotate.conf   # force rotation

Find large files efficiently:

find / -type f -size +500M -exec ls -lh {} \;

5. LVM Advanced Management

Create physical volume, volume group, and logical volume:

pvcreate /dev/sdb          # physical volume
vgcreate data_vg /dev/sdb   # volume group
lvcreate -L 10T -n data_lv data_vg   # logical volume

Online expansion (four steps):

lvextend -L +5G /dev/data_vg/data_lv   # extend LV
resize2fs /dev/data_vg/data_lv          # EXT4 resize
xfs_growfs /data                         # XFS resize
vgextend data_vg /dev/sdc                 # extend VG

Create a consistent snapshot for backup:

lvcreate -L 1G -s -n db_snap /dev/data_vg/db_lv

6. Filesystem Deep Techniques

Understand inode (metadata storage) and block (actual data). Monitor inode usage with df -i, especially in small‑file scenarios.

Repair a filesystem safely:

umount /dev/sdb1   # must unmount first
fsck -y /dev/sdb1   # run consistency check

Never run fsck on a mounted filesystem to avoid metadata corruption.

7. RAID Configuration

Create a software RAID‑1 array:

mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sdd /dev/sde

Common RAID levels:

RAID 0 : striping, high performance, no redundancy.

RAID 1 : mirroring, strong redundancy, half capacity.

RAID 5 : distributed parity, balanced performance and redundancy.

8. Performance Monitoring and Optimization

Gather device‑level I/O statistics: iostat -dx 1 Identify top I/O‑consuming processes: iotop A utilization percentage (>80 %) indicates a saturated disk.

9. Best‑Practice Summary

Partition strategy : separate /home to isolate user data.

Filesystem selection : XFS for large files, EXT4 for general use.

LVM adoption : mandatory in production to avoid fixed partition sizes.

Space‑alert mechanisms : combine df, du, and find for proactive monitoring.

Backup first : always verify backup availability before any disk operation.