openskills.info
Course catalog
Course Preview

Systems Administration

Systems administration is the work of keeping an organization’s computers and services available, secure, consistent, and recoverable. It connects business needs to the day-to-day care of servers, endpoints, accounts, software, networks, data, and operational records across their full lifecycle.

itInfrastructure and operations

Don't Panic — Systems Administration

Systems administration is the work of keeping computers useful after the exciting moment when they first boot and everyone assumes the job is over. It is not merely typing commands at a server. It connects a service requirement to the machines, access, configuration, evidence, backups, and recovery paths that keep the service available later, which is when the machines begin to develop opinions.

The durable shape is a control loop. First, know what exists: the inventory records assets, owners, and dependencies. Then state what should exist: a baseline is the approved configuration for a class of systems. Compare that desired state with actual state, change carefully, observe the result, correct configuration drift, test recovery, and update the record. A command belongs inside that loop; it is not the loop wearing a hat.

This is why a host that is patched, monitored, and apparently calm can still be unmanaged. If nobody owns it, if its backup has never been restored, or if its service depends on a management path that has failed, the reassuring green status is doing its best but does not know the whole story. Automation helps compare and reconcile state across many systems. It also repeats a poor decision with exceptional punctuality, so staged changes need a success signal and a stop condition.

When something breaks, begin with evidence. Telemetry is the measurements and event data collected during operation; logs tell you what happened, metrics show change over time, inventory supplies ownership, and configuration shows the intended state. None is the complete answer. Follow the consumer path from name resolution through network policy, the listener, service process, dependencies, and data. A healthy host can still provide an unreachable service, which is the sort of detail that keeps incident timelines pleasantly full.

Recovery is not the same as repair. Repair changes a failed component; recovery restores the required service or data, perhaps by another path. A completed backup job proves a copy exists. A restore test checks credentials, keys, dependency order, data integrity, and the consumer path. The recovery point objective sets the tolerated age of lost data, while the recovery time objective sets the target time to restore service. Both are targets, not tiny legal guarantees from the universe.

Read the Intro for the full operating model and where its responsibilities meet other teams. Use the Slides when you need the relationships at a glance. Keep the Cheatsheet nearby during a change, a service-path investigation, or a restore test. The Field Notes concentrates on the awkward operational judgment that turns a safe-looking fleet action into a safe one.

Where this skill leads

Relevant careers

See how this topic contributes to broader role-level skill maps.

Sources