Server Maintenance Checklist: What to Do Daily, Weekly and Monthly

A server that nobody looks at works fine until the day it does not. A short routine of checks catches most problems while they are still small. Here is a checklist you can follow yourself, or hand to the person who looks after your server.

Daily: two minutes

• Is it up? Open your site and check that your monitoring shows green.
• Did last night's backup finish? Look at the backup log or the report email. A backup job that fails silently is the most common surprise.
• Any alerts? Read the alert emails instead of filing them away.

Weekly: fifteen minutes

• Install security updates. On Ubuntu or Debian use apt update && apt upgrade. On AlmaLinux or Rocky use dnf upgrade. Reboot when a kernel update asks for it, at a quiet time.
• Check disk space. Run df -h and look for anything above 80%. Logs and old backups are the usual culprits.
• Check memory and load. free -h and uptime show whether you are close to the limit.
• Review failed services. systemctl --failed should list nothing.
• Look at failed logins. A sudden rise in failed SSH attempts is worth a look; fail2ban or a firewall rule can slow it down.

Monthly: an hour

• Test a restore. Restore one site or database from a backup to a test place. A backup you have never restored is only a hope.
• Update your CMS and plugins. Outdated plugins are the number one way websites get hacked.
• Review users and keys. Remove accounts and SSH keys of people who no longer need access.
• Check certificates. Make sure SSL certificates renew on their own and see the expiry dates.
• Check the firewall. Only the ports you need should be open to the internet.
• Look at performance. Slow database queries and growing logs tell you where to tune before visitors complain.

Every quarter

• Review the plan. Is the server still the right size? Right-sizing saves money and prevents crashes.
• Review your recovery plan. If the server vanished today, how long would it take to be back and who does what?
• Check the operating system's support dates. An end-of-life system stops getting security fixes.

Keep a log

Write down what you changed and when, even in a simple text file. When something breaks the next week, the first question is always "what changed?".

Too much to do?

This checklist takes a few hours a month if everything goes well, and much more when it does not. If you would rather spend that time on your business, our managed services take over the monitoring, patching and backups for a fixed monthly price, with a person to call when something breaks.