CloudSprocket Labs · Linux Server Rescue
Learn Linux troubleshooting on servers that are really broken
Diagnosis is a skill you build through repetitions, and repetitions need incidents.Linux Server Rescue hands you realistic tickets on a disposable systemd server, then checks your repair by its outcome.
Tutorials teach building. Work demands diagnosis.
Most Linux learning material walks you through installing and configuring software on a healthy machine. Operations work is the other way round: a service that worked yesterday is failing now, the symptom is vague, and someone needs the cause found and fixed safely. On-call rotations and SRE interviews test exactly that, and reading about it does not build it.
Production is a poor place to practise. The consequences are real and the safest move is usually a runbook, not an experiment. What builds the muscle is a real target with a known fault, freedom to investigate, and a fast route back to a healthy state.
A rescue loop you can repeat until it is instinct
The lab gives you one real systemd host calledrelay on Ubuntu, Debian or Rocky Linux. Every incident starts from a healthy baseline and arrives as an operational ticket: a symptom and a constraint, never a lesson plan. You collect evidence inside the host, make the smallest safe repair and prove the service is healthy again.
Three ordered hints escalate only as far as you ask, and the verifier checks the outcome rather than your commands, so any safe route to a healthy server passes. The lab is free, MIT licensed and disposable: it runs in Docker on your machine or in a GitHub Codespace, and reset always returns a clean server.
./lab up ubuntu
./lab break 01
./lab shell
./lab verify 01Three of the incidents waiting for you
- 01
Service failure
Trace a systemd service trapped in a restart loop, find the bad override and restore the healthy service without weakening its restart policy.
systemctl · journalctl · unit overrides
- 02
Full filesystem
Follow an application write failure to the correct filesystem, measure what consumed it and recover capacity without moving the data.
df · findmnt · du
- 03
DNS ghost
Find a stale local name-service override that shadows the lab network's DNS answer, restore the intended resolution path and recover the upstream systemd check.
getent · dig · nsswitch
The full set spans services, storage, permissions, DNS, process load, networking and change recovery, and it grows with each release. Thelab page lists every incident with its difficulty.
Learn the method, not a command library
The skill that transfers between incidents is a repeatable investigation: stabilise, observe, narrow, repair, prove. Theintroduction to Linux server rescue walks through that five-step method and the signals each step relies on. Practising it here pays off later too, because container and Kubernetes debugging rest on the same host-level instincts.
Take your first ticket
The quick start takes you from a clean machine or a browser tab to your first verified repair. No accounts, no session timer, no fees, and the server is allowed to break.