How to Fix a Full Disk on a Linux Server

A full disk is one of the most common causes of server emergencies. When a Linux server disk fills to 100%, MySQL stops writing to databases, PHP cannot create temporary files, email bounces, and logs cannot be written. Your sites go down silently with no obvious error message to the visitor.

Here is how to diagnose and fix it quickly.

Step 1: Find What Is Using the Space

Run these commands in order to find the large files:

# Find largest directories
du -sh /* 2>/dev/null | sort -hr | head -15

# Drill into the largest one (e.g. /var)
du -sh /var/* 2>/dev/null | sort -hr | head -15

# Find individual large files
find / -type f -size +100M 2>/dev/null | sort -k5 -rn | head -20

The 6 Most Common Causes

1. Log Files That Were Never Rotated

The most common cause. Nginx, Apache, MySQL, and PHP error logs grow indefinitely without log rotation configured. A single busy site can generate gigabytes of access logs per month.

Fix:

# Check log sizes
du -sh /var/log/* | sort -hr | head -10

# Clear a specific log safely (do NOT delete it)
truncate -s 0 /var/log/nginx/access.log

# Restart nginx to re-open log file
systemctl restart nginx

Prevent recurrence: Configure logrotate. Check /etc/logrotate.d/ — most services have a config here. If missing, create one.

2. Old Backup Files Accumulating

aaPanel and cPanel backup tasks create files in /www/backup or /home/*/backup. If retention is set to "keep all" they grow indefinitely.

Fix: Go to aaPanel → Cron → edit your backup tasks → set retention to 7 days. Then manually delete old backup files from the backup directory.

3. MySQL Binary Logs

MySQL binary logs (used for replication and recovery) accumulate rapidly on busy databases. They live in /var/lib/mysql/ as files named mysql-bin.000001, etc.

# Check binary log size
du -sh /var/lib/mysql/mysql-bin* 2>/dev/null

# Purge logs older than 7 days (run in MySQL)
mysql -u root -p -e "PURGE BINARY LOGS BEFORE DATE_SUB(NOW(), INTERVAL 7 DAY);"

# Disable binary logs if not needed (add to /etc/mysql/my.cnf)
# skip-log-bin

4. PHP Session Files

PHP stores session files in /tmp or /var/lib/php/sessions. On a busy site with no session cleanup, millions of small files accumulate.

find /var/lib/php/sessions -type f -mtime +7 -delete

5. WordPress Upload Bloat

Unoptimised media uploads, multiple image sizes generated by WordPress, and plugin-generated files accumulate in wp-content/uploads. A single high-traffic blog can generate 10-20GB over a few years.

Fix: Audit /wp-content/uploads for very large files, remove duplicates, and implement a media cleanup plugin. Consider offloading media to cloud storage (S3, Cloudflare R2).

6. Core Dumps

When processes crash, Linux can write large core dump files. Check /var/crash or /tmp for files ending in .core or core.*

find / -name "core.*" -o -name "*.core" 2>/dev/null | xargs ls -lh 2>/dev/null

Setting Up Disk Monitoring

Never find out about a full disk from a down site. Set up an alert:

# Add to crontab — alert when disk over 80%
*/30 * * * * /usr/local/bin/check-disk.sh

Alternatively, aaPanel has built-in monitoring alerts under Monitor → Disk Usage — configure email alerts at 80% threshold.