cPFence_daily_cron.sh

#!/bin/bash
# Script by cPFence Team, https://cpfence.app
# 
# Description:
# This script performs routine server maintenance tasks:
# - Truncates user error logs larger than 5MB and server logs larger than 100MB, keeping only the last entries using `sponge` (ensure `sponge` is installed on the server).
# - Retains only the last 10 days of system journal logs.
# - Updates and upgrades packages non-interactively and checks if a reboot is required.
# - Clears APT cache to save space.

# Truncate error logs > 5MB for users on server
/usr/bin/find /var/www \
  -path "*/cpf_logs/*" -prune -o \
  \( -name "*.log" -o -name "*_log" \) -size +5M -type f -exec sh -c '
    for file; do
      tail -c 5M "$file" | /usr/bin/sponge "$file"
    done
  ' sh {} +

# Truncate server logs > 100MB
/usr/bin/find /var/log \( -name "*.log" -o -name "*_log" \) -size +100M -exec sudo sh -c '
  for file; do
    tail -c 100M "$file" | /usr/bin/sponge "$file"
  done
' sh {} +

# Keep only the last 10 days of logs in the journal
sudo journalctl --vacuum-time=10d > /dev/null 2>&1

# Set non-interactive environment for package installations
export DEBIAN_FRONTEND=noninteractive

# Define your command
CMD="apt-get"

# Run the update, upgrade, and autoremove in a non-interactive manner and check if a reboot is required
update_output=$(${CMD} update -y > /dev/null 2>&1 && \
               ${CMD} full-upgrade -y > /dev/null 2>&1 && \
               ${CMD} autoremove -y > /dev/null 2>&1 && \
               [ -f /var/run/reboot-required ] && echo "Reboot required" || echo "No reboot required")

# Output the result
echo "$update_output"

# If a reboot is required, display more info about the reason
if [ -f /var/run/reboot-required ]; then
    echo "The following packages require a reboot:"
    cat /var/run/reboot-required.pkgs
fi

# Check for errors
if [ $? -eq 0 ]; then
    echo "Script executed successfully."
else
    echo "Error occurred during script execution."
fi

# Clear APT cache
apt-get clean

Not sure why.. but I keep seeing leftovers in /var/www/*/.mysql_backup_enhance
Maybe something like: find /var/www/*/.mysql_backup_enhance/* -type d -mtime +1 and deleting any files? Just a suggestion based on what we do.

Just a FYI, you should probably leave at least 30 days worth of logs in the journalctl. How much space are you saving with this script? Doesn't seem like the space savings is worth losing the logs.