Skip to main content Link Menu Expand (external link) Document Search Copy Copied
Aikes Docs

Server updates and upgrades

Update a server

1. Check load, usage and logs

  • disk space (df -h)
  • cpu load (htop)
  • memory (htop)
  • container list (docker ps)
  • dmesg (dmesg -T)
  • syslog (tail -f -n500 /var/log/syslog)

  • logged in users (w and lastlog)

  • check if external monitoring is active
  • check the firewall (nmap -A | grep open)
  • check the backups

2. Reboot

Before we change anything we want to know if we can still reboot. So that if that goes wrong we know it is not the updates. Also have a quick look at the application or verify that the server is doing what it is supposed to do.

3. Snapshot / backups

A full snapshot of the disk or a quick run of the backups. So that if something goes wrong we can easily go back to the way it was. This depends on the server and the kind of updates we’re doing. For container hosts this is not needed but for a complicated Ubuntu install it definitely is.

4. Update

On Ubuntu and Debian use the following command as root:

apt update && apt dist-upgrade && apt autoremove && apt clean && date && reboot

Or without the questions if you are comfortable with that:

apt update && apt dist-upgrade -y && apt autoremove -y && apt clean && date && reboot

Or with screen on unstable connections:

sudo screen sh -c "apt update && apt dist-upgrade && apt autoremove && apt clean && date && reboot"

The date is there so it outputs the time before the reboot, if you are waiting 5 minutes for it to come back you know something is wrong.

Use ssh -o 'ConnectionAttempts 999' <yourserver> to reconnect automatically.

5. Final checks

  • check if the application or the server is still doing its job
  • check if automatic security updates are enabled
  • check the load, usage and logs
  • register the update in the log for compliance