Understanding Linux File Systems

Understanding Linux File Systems

A file system determines how data is organized on a storage device — how directories are structured, how metadata (permissions, timestamps, ownership) is stored, how the system recovers from crashes, and how efficiently it handles different workloads. Choosing the right file system for a workload matters more than most people realize: the wrong choice can cost performance, limit file sizes, or cause data loss in a crash.

What a file system does

Storage device (SSD/HDD)
      |
      ├── Superblock         — File system metadata: type, size, state
      ├── Inode table        — Per-file metadata: size, permissions, timestamps, block pointers
      ├── Directory entries  — Maps filenames to inode numbers
      └── Data blocks        — Actual file content

When you access /var/log/syslog:
1. Kernel reads /var inode  → finds /var/log directory
2. Reads /var/log inode     → finds syslog file's inode number
3. Reads syslog inode       → finds data block addresses
4. Reads data blocks        → returns file content

How Linux organizes file systems

Linux uses a single unified directory tree. All file systems — whether on a local disk, NFS share, or USB drive — are mounted at a point in this tree. There is no concept of drive letters (C:, D:) as on Windows.

# See the current mount hierarchy
findmnt

# See all mounted file systems with usage
df -hT

df -hT output on a typical Ubuntu server

Filesystem     Type      Size  Used Avail Use% Mounted on
/dev/sda1      ext4       50G   15G   33G  32% /
/dev/sda2      ext4      200G   80G  110G  43% /home
tmpfs          tmpfs      16G     0   16G   0% /dev/shm
/dev/sdb1      xfs       500G  120G  380G  24% /data

Common Linux file systems compared

Checking your current file systems

# Show file system type for all partitions
lsblk -f

# Show detailed info about a specific file system
sudo tune2fs -l /dev/sda1 | head -20    # ext4
sudo xfs_info /dev/sdb1                 # XFS
sudo btrfs filesystem show /data        # Btrfs

# Check file system health
sudo fsck -n /dev/sda2    # -n = dry run, read-only check (on UNMOUNTED fs only!)

# Show inode usage (running out of inodes = "no space" even with free blocks)
df -i

df -i showing inode usage

Filesystem       Inodes  IUsed  IFree IUse% Mounted on
/dev/sda1       3276800 250000 3026800    8% /
/dev/sda2       3276800 3276790      10  100% /var/spool/mail

⚠️ WARNING: Running out of inodes (IUse% 100%) causes "No space left on device" errors even when disk usage is low. This happens when a directory contains millions of small files (mail spools, session files, cache directories). Use df -i to check inode usage alongside disk usage when diagnosing "no space" errors.

Choosing the right file system

Conclusion

Ext4 is the right choice for most Ubuntu systems — it is stable, well-understood, and has excellent tools. Choose XFS for database or log storage with large files and high write throughput. Choose Btrfs when you need built-in snapshots or checksumming (like on a NAS). Always check both disk space (df -h) and inode usage (df -i) when diagnosing "no space left on device" errors — they have different causes and different fixes.