Optimizing Disk Space Usage on Linux Servers: A Practical Guide

Introduction

Running out of disk space is a frequent issue for Linux servers, leading to service disruptions, application crashes, or failed deployments. As a DevOps engineer, optimizing disk space is a critical skill to ensure system reliability. This article provides a step-by-step guide to identifying and resolving disk space issues on Linux servers using built-in tools and automation, helping you maintain healthy servers and impress potential employers.

Why Disk Space Optimization Matters

Full disks can cause:

• Application failures due to inability to write logs or data
• Slow performance from excessive I/O
• Failed CI/CD pipelines or backups
  Proactively managing disk space demonstrates your ability to maintain production systems and prevent outages.

Prerequisites

• Root or sudo access to a Linux server (e.g., Ubuntu, CentOS)
• Basic familiarity with Linux commands
• Tools like df, du, find, and ncdu (install if needed)

Step-by-Step Guide to Optimizing Disk Space

1. Check Current Disk Usage

Start by identifying how much disk space is used and where:

df -h

This displays disk usage in a human-readable format. Focus on filesystems nearing 100% usage (e.g., / or /var).

Example output:

Filesystem      Size  Used Avail Use% Mounted on
/dev/sda1       100G   85G   15G  85% /

2. Identify Large Files and Directories

Use du to find space-hogging directories:

du -h / | sort -rh | head -n 10

This lists the top 10 largest directories. Focus on common culprits like /var/log, /tmp, or /home.

For a more interactive tool, install and use ncdu:

# Ubuntu
sudo apt install ncdu
# CentOS
sudo yum install ncdu
ncdu /

Navigate with arrow keys to identify large files or directories.

3. Clean Up Old Logs

Log files in /var/log often consume significant space. Check their sizes:

ls -lh /var/log

To clean up:

• Compress Old Logs: Use gzip to compress logs older than a week:

  find /var/log -type f -name "*.log" -mtime +7 -exec gzip {} ;

• Clear Large Logs: Truncate active logs safely (e.g., for syslog):

  sudo truncate -s 0 /var/log/syslog

• Use logrotate: Configure /etc/logrotate.conf to automate log rotation. Example for /var/log/myapp.log:

  /var/log/myapp.log {
    daily
    rotate 7
    compress
    missingok
    notifempty
  }

  Test the configuration:

  sudo logrotate -f /etc/logrotate.conf

4. Remove Unused Packages

Unnecessary packages can accumulate, especially on long-running servers. On Debian/Ubuntu:

sudo apt autoremove
sudo apt autoclean

On CentOS/RHEL:

sudo yum autoremove
sudo yum clean all

These commands remove unused dependencies and clean cached package data.

5. Delete Temporary Files

Temporary files in /tmp or /var/tmp can pile up. Safely remove old files:

find /tmp -type f -mtime +7 -delete
find /var/tmp -type f -mtime +7 -delete

Verify no critical processes are using these files before deletion:

lsof /tmp

6. Automate Cleanup with a Script

Create a script to automate disk cleanup. Save as cleanup_disk.sh:

#!/bin/bash

# Configuration
LOG_DIR="/var/log"
TMP_DIR="/tmp"
DAYS_TO_DELETE=7
LOG_FILE="/var/log/disk_cleanup.log"

# Ensure log file exists
touch "$LOG_FILE"
if [ $? -ne 0 ]; then
  echo "ERROR: Cannot write to $LOG_FILE" >&2
  exit 1
fi

# Log function
log() {
  echo "$(date '+%Y-%m-%d %H:%M:%S') $1" >> "$LOG_FILE"
}

# Clean temporary files
find "$TMP_DIR" -type f -mtime +$DAYS_TO_DELETE -delete
log "Deleted temporary files older than $DAYS_TO_DELETE days in $TMP_DIR"

# Compress logs
find "$LOG_DIR" -type f -name "*.log" -mtime +$DAYS_TO_DELETE -exec gzip {} ;
log "Compressed logs older than $DAYS_TO_DELETE days in $LOG_DIR"

# Remove old compressed logs
find "$LOG_DIR" -type f -name "*.gz" -mtime +30 -delete
log "Deleted compressed logs older than 30 days in $LOG_DIR"

# Clean package cache (Ubuntu)
if command -v apt >/dev/null 2>&1; then
  apt autoremove -y && apt autoclean
  log "Cleaned apt package cache"
fi

# Clean package cache (CentOS)
if command -v yum >/dev/null 2>&1; then
  yum autoremove -y && yum clean all
  log "Cleaned yum package cache"
fi

# Report disk usage
DISK_USAGE=$(df -h / | tail -1 | awk '{print $5}')
log "Disk usage after cleanup: $DISK_USAGE"

Make it executable:

chmod +x cleanup_disk.sh

7. Schedule the Script with Cron

Run the cleanup daily at 3 AM. Edit the crontab:

crontab -e

Add:

0 3 * * * /path/to/cleanup_disk.sh

8. Test the Script

Run manually to verify:

./cleanup_disk.sh

Check /var/log/disk_cleanup.log for output:

2025-07-16 15:00:00 Deleted temporary files older than 7 days in /tmp
2025-07-16 15:00:01 Compressed logs older than 7 days in /var/log
2025-07-16 15:00:02 Disk usage after cleanup: 60%

Common Issues and Fixes

• Permission Errors: Run the script as root or with sudo to access system directories.
• Accidental Deletion: Double-check find commands with -print before -delete (e.g., find /tmp -type f -mtime +7 -print).
• Logrotate Failures: Verify /etc/logrotate.conf syntax with logrotate -d.
• Package Manager Errors: Ensure apt or yum commands are run with sufficient privileges.

Best Practices

• Monitor Disk Usage: Use tools like Prometheus or Nagios to alert before disks fill up.
• Backup Critical Data: Archive important logs to a remote location before deletion.
• Test Incrementally: Test cleanup commands on a non-production server first.
• Log Actions: Always log cleanup activities for auditing and debugging.

Conclusion

Optimizing disk space on Linux servers is a vital DevOps skill that prevents outages and ensures system reliability. This guide and script provide a reusable solution for identifying and resolving disk space issues, showcasing your expertise in system administration. Save this guide for your next server maintenance task, and keep your systems running smoothly!