Skip to content

Linux for DevOps

By Updated

Linux is the operating system under almost everything you'll run: cloud instances, Kubernetes nodes, containers, and CI runners. This track covers the administration and troubleshooting skills you'll use every week, with commands you can run on any modern distribution. Examples use Ubuntu 24.04 LTS and note where RHEL-family systems differ.

What You'll Learn

  • How the filesystem, users, and permissions control access
  • How processes, signals, and systemd services behave, and how to read their logs
  • How to manage disks, filesystems, and LVM without losing data
  • How to secure SSH and sudo, keep packages patched, and diagnose a slow server
  • How to recover a server that won't boot, fix SELinux and AppArmor denials, tune the kernel, and configure networking

Read in This Order

  1. Filesystem and Permissions — the directory layout, ownership, modes, special bits, ACLs, and umask
  2. Processes and Signals — process states, ps and /proc, signals, zombies, priorities, and open files
  3. systemd and journald — units, writing a service, overrides, timers, sandboxing, and journalctl
  4. Storage, Disks, and LVM — block devices, filesystems, fstab, LVM, growing cloud disks, and "disk full" incidents
  5. Users, sudo, and SSH Hardening — accounts, groups, sudoers, key-based SSH, and a hardened sshd configuration
  6. Packages and Patching — apt and dnf, pinning, unattended security updates, and reboot handling
  7. Performance Troubleshooting — the USE method and a 60-second triage for CPU, memory, disk, and network
  8. Boot Process and Recovery — GRUB, the initramfs, rescue and emergency mode, a broken fstab, and cloud instance recovery
  9. SELinux and AppArmor — reading denials and fixing labels, ports, booleans, and profiles instead of disabling them
  10. Kernel Tuning — sysctl, the settings worth changing, kernel modules, and resource limits
  11. Network Configuration — ip, static addresses with Netplan and nmcli, DNS, hostnames, and safe changes over SSH

The Commands You'll Use Most

Task Command
What's using the disk? df -h, du -xh --max-depth=1 / \| sort -h
What's running? ps aux --sort=-%cpu \| head, systemctl list-units --failed
Why did the service fail? systemctl status app, journalctl -u app -e
What's listening? ss -tulpn
Who can read this file? ls -l, stat, getfacl
Is the box overloaded? uptime, vmstat 1, iostat -xz 1
Why was access denied by policy? sudo ausearch -m AVC -ts recent, sudo aa-status
What are the addresses and routes? ip -br addr, ip route
What did the last boot log? journalctl -b -1 -p err

Try It Interactively

The Real Outages page covers the incidents behind this track: the 2012 leap second, Dirty Pipe, the xz backdoor, and Ubuntu's silently stalled updates.

How This Connects to the Rest of the Site

Linux concept Shows up again in
Processes, namespaces, cgroups Docker: Linux namespaces and cgroups
systemd services Podman Quadlet units, Kubernetes node services (kubelet, containerd)
Permissions and users Container USER directives, Kubernetes securityContext
SSH and sudo Ansible connectivity and become
Performance tools Docker production troubleshooting

Next

Start with Filesystem and Permissions.