Linux server guides
Practical answers for people who run Ubuntu, Debian, AlmaLinux and Rocky Linux servers. Every command is checked on a real server before it's published, and every page lists its sources.
Patching
- Ubuntu end of life dates, plus Debian, RHEL, AlmaLinux, Rocky and CentOSUbuntu, Debian, RHEL, AlmaLinux, Rocky Linux and CentOS end of life dates from each vendor, what is past EOL in November 2026, and what to do next.
- Unattended-upgrades on Ubuntu and Debian: set it up and prove it's workingSet up unattended-upgrades on Ubuntu and Debian servers, then prove it's patching: the settings, timers, origins and logs to check, and the silent failures.
- apt list --upgradable: how to check for security updates on Ubuntu and DebianList pending updates with apt list --upgradable, pick out the security fixes and see which CVEs each one closes, on Ubuntu and Debian, plus Ubuntu Pro ESM.
- sudo apt-get upgrade vs apt upgrade vs full-upgrade: what each one does on a serverWhat sudo apt-get upgrade, apt upgrade and full-upgrade really do on a server: removals, kept-back packages, new kernels, a dry run first and a safe order.
- “The following packages have been kept back”: why apt does it and how to fix it safelyWhy apt says “the following packages have been kept back”: phased updates, new dependencies, holds or removals. Find the cause, then use the safe fix for it.
Kernels and reboots
- How to check the kernel version in Linux (running vs installed)Check the Linux kernel version with uname -r, then compare it with the newest installed kernel to see if a reboot is still needed. Tested on Ubuntu 24.04.
- How to update the kernel on Ubuntu and confirm it's runningUpdate the Ubuntu kernel the supported way with apt, pick GA or HWE, reboot, confirm the new kernel runs, and remove old kernels safely. Tested on 24.04.
- Linux reboot command: reboot, schedule and verify a server restartWhich Linux reboot command to use: reboot, shutdown -r or systemctl reboot. Schedule or cancel a reboot, and prove the server came back by its boot ID.
Services and logs
- journalctl tail: follow a service's logs and find out why it failedTail and follow systemd logs with journalctl: -f, -n, -u, --since, -b -1 and -p err, then find why a unit failed and keep logs across reboots.
- systemctl list services: see what's running, failed or enabledList services with systemctl: running, failed, enabled or masked. Then check one with status, is-active and show, and prove a restart really worked.