Hands-On: Advanced Linux

Knowing how to start a service, read its logs and partition a disk gets you through a normal day on a Linux server. It does not get you through the day a service keeps crashing under a memory limit, a filesystem will not mount, or an operator suddenly cannot run the one command they need. Hands-On: Advanced Linux picks up where Hands-On: Learn Linux stops and puts you in a disposable Linux machine in your browser, where every exercise is graded on the state you leave the system in.

You start by writing systemd units from scratch: choosing how a daemon is started, loading its settings from a file, restarting it on failure, and creating many instances from one template. You then override shipped units without editing them, wire dependencies, activate services from sockets, paths and timers, and cap memory and share CPU by weight through cgroups, including working out why the kernel killed a process. Later chapters lock services down with sandboxing directives, filter the journal by its fields, and go beyond permission bits with ACLs, immutable files, file capabilities and narrowly scoped sudo rules. You also inspect processes through /proc, persist kernel parameters, isolate a process with namespaces, partition and tune ext4, recover a damaged filesystem, harden SSH and audit installed packages.

The course closes with incident labs, where a deliberately broken system has to be diagnosed and repaired end to end. It suits anyone who already works comfortably at the Linux command line and wants the administration depth that comes next.