df

Overview

The df (disk free) command displays filesystem disk space usage information. It shows available and used space for mounted filesystems, essential for system monitoring and disk management.

Syntax

df [options] [filesystem...]

Common Options

Option Description
-h Human-readable sizes (K, M, G)
-H Human-readable with powers of 1000
-T Show filesystem type
-i Show inode information
-a Include dummy filesystems
-l Local filesystems only
-x type Exclude filesystem type
-t type Include only filesystem type
--total Display grand total
--sync Sync before getting usage info

Key Use Cases

  1. Check available disk space
  2. Monitor filesystem usage
  3. System health monitoring
  4. Capacity planning
  5. Troubleshoot disk full issues

Examples with Explanations

Example 1: Basic Usage

df -h

Shows disk usage in human-readable format

Example 2: Specific Filesystem

df -h /home

Shows usage for filesystem containing /home

Example 3: Show Filesystem Types

df -hT

Displays filesystem types along with usage

Example 4: Inode Information

df -hi

Shows inode usage instead of block usage

Understanding Output

Default columns: - Filesystem: Device or filesystem name - 1K-blocks: Total space in 1K blocks - Used: Used space - Available: Available space - Use%: Percentage used - Mounted on: Mount point

Example output:

Filesystem      Size  Used Avail Use% Mounted on
/dev/sda1        20G   15G  4.2G  79% /
/dev/sda2       100G   45G   50G  48% /home

Common Usage Patterns

  1. Check root filesystem:

    df -h /

  2. Monitor all local filesystems:

    df -hl

  3. Find full filesystems:

    df -h | awk '$5 > 90 {print}'

Filesystem Types

Common filesystem types: - ext4: Linux native filesystem - xfs: High-performance filesystem - btrfs: Advanced Linux filesystem - ntfs: Windows filesystem - vfat: FAT32 filesystem - tmpfs: Temporary filesystem in RAM - nfs: Network filesystem

Advanced Usage

  1. Exclude specific types:

    df -h -x tmpfs -x devtmpfs

  2. Show only specific type:

    df -h -t ext4

  3. Include totals:

    df -h --total

Monitoring and Alerting

  1. Check for full filesystems:

    df -h | awk '$5 > 95 {print "WARNING: " $6 " is " $5 " full"}'

  2. Monitor specific threshold:

    USAGE=$(df / | awk 'NR==2 {print $5}' | sed 's/%//')
if [ $USAGE -gt 80 ]; then
    echo "Root filesystem is ${USAGE}% full"
fi

Performance Analysis

  • Fast operation
  • Reads filesystem metadata
  • Minimal system impact
  • Real-time information
  • Good for automated monitoring

Best Practices

  1. Regular monitoring of disk space
  2. Set up alerts for high usage
  3. Use human-readable format
  4. Monitor both space and inodes
  5. Exclude irrelevant filesystems

Scripting Applications

  1. Disk space monitoring:

    #!/bin/bash
THRESHOLD=90
df -h | awk -v thresh=$THRESHOLD '
NR>1 && $5+0 > thresh {
    print "WARNING: " $6 " is " $5 " full"
}'

  2. System health check:

    check_disk_space() {
    echo "=== Disk Space Report ==="
    df -h | grep -vE '^Filesystem|tmpfs|cdrom'
    echo ""
    echo "=== Critical Usage (>90%) ==="
    df -h | awk '$5 > 90 {print $0}'
}

Integration Examples

  1. With cron for monitoring:

    # Check disk space every hour
0 * * * * df -h | awk '$5 > 90 {print}' | mail -s "Disk Alert" admin@domain.com

  2. System status dashboard:

    echo "System Status - $(date)"
echo "Disk Usage:"
df -h | grep -v tmpfs
echo ""
echo "Inode Usage:"
df -hi | awk '$5 > 80 {print}'

Inode Monitoring

  1. Check inode usage:

    df -hi

  2. Find filesystems with high inode usage:

    df -hi | awk '$5 > 80 {print "High inode usage: " $6 " (" $5 ")"}'

  3. Monitor both space and inodes:

    df -h && echo "--- Inodes ---" && df -hi

Troubleshooting

  1. Filesystem shows 100% but space available
  2. Inode exhaustion
  3. Stale NFS mounts
  4. Permission issues
  5. Filesystem corruption

Network Filesystems

  1. Show only local filesystems:

    df -hl

  2. Include network filesystems:

    df -h -t nfs -t cifs

  3. Exclude network filesystems:

    df -h -x nfs -x cifs

Output Formatting

  1. Custom format:

    df -h | awk '{printf "%-20s %8s %8s %8s %s\n", $1, $2, $3, $4, $6}'

  2. CSV format:

    df -h | awk 'NR>1 {printf "%s,%s,%s,%s,%s,%s\n", $1,$2,$3,$4,$5,$6}'

  3. JSON format:

    df -h | awk 'NR>1 {printf "{\"fs\":\"%s\",\"size\":\"%s\",\"used\":\"%s\",\"avail\":\"%s\",\"use\":\"%s\",\"mount\":\"%s\"}\n", $1,$2,$3,$4,$5,$6}'

Automation Examples

  1. Cleanup trigger:

    #!/bin/bash
for fs in $(df -h | awk '$5 > 85 {print $6}'); do
    echo "Filesystem $fs is getting full"
    # Trigger cleanup scripts
    cleanup_logs.sh "$fs"
done

  2. Capacity planning:

    df -h | awk '
NR>1 {
    gsub(/%/, "", $5)
    if ($5 > 70) print $6 " will need attention soon (" $5 "%)"
}'