Skip to content
IT Atlas

Practice

Troubleshooting guides

Each guide is an ordered sequence of checks that narrows the fault instead of guessing. The order matters: the early steps are the ones that most often end the investigation.

8 guides

A method that works on anything

  1. 1. Establish scope. One user, one device, one site, or everyone? Scope eliminates most causes immediately.
  2. 2. Establish timing and change. What changed, and when did it last work? Most incidents follow a change.
  3. 3. Work bottom up through the layers. Link, address, name resolution, routing, transport, authentication, application.
  4. 4. Read the log the system already wrote before forming a theory.
  5. 5. Change one variable at a time, and record what you changed.
  6. 6. Fix the cause, not just the symptom, then write it down so the next person does not repeat the investigation.