Memory hook: Protect data, isolate the cause, prove the fix.
Must remember
Identify the problem and scope, ask about recent changes, establish a theory, test it, plan the fix, implement under authorization, verify and document. Start with simple evidence: power, connections, input selection and error messages. Do not replace expensive parts before checking a loose cable.
No power differs from no POST, no boot and no display. Verify outlet/adapter/power indicators, then POST diagnostics, firmware device detection and boot configuration. A black screen may still come from a running system with the wrong display path. Intermittent shutdown under load suggests heat or power but requires measurements.
Use known-good compatible components to isolate faults. Memory diagnostics, storage health data and vendor self-tests support a theory; one passed test cannot prove absence of every intermittent fault. Multimeters require appropriate training/range and electrical-safety precautions; never probe hazardous equipment casually.
Network failures can originate at the physical link, VLAN, address, gateway, DNS or application. A loopback plug tests a limited path, while a cable tester checks wiring/continuity; neither proves end-to-end service success. Wi-Fi analyzers expose radio/channel evidence.
Preserve data and recovery keys before storage, firmware or OS changes. Document serials, replaced parts, settings and results. Dispose of batteries/electronics through approved processes. Use ESD protection and lifting practices, and keep liquids/dust away from sensitive equipment.
Finish by reproducing the user's original workflow and checking side effects. Explain the change in plain language, obtain required acceptance and update the ticket so the next technician can understand the evidence.
Choose under exam pressure
| Requirement | Choice and reason |
|---|---|
| System powers but does not POST | Use board/vendor diagnostics and minimum compatible hardware isolation. |
| User files at risk | Prioritize approved backup/recovery before destructive repair. |
| Intermittent fault | Correlate load, temperature, power and logs rather than guessing. |
Traps
- No display is not identical to no power.
- A successful reboot is not complete verification of every affected function.
Active recall
1. Why ask about recent changes?
They identify plausible causes and rollback paths.
2. What should precede destructive storage repair?
An authorized data-protection/recovery plan.
3. What does a cable continuity test not prove?
Correct addressing, DNS, policy or application behavior.
4. Why use a known-good compatible part?
It isolates a component without introducing a new incompatibility.
5. What belongs in the closing ticket?
Symptoms, evidence, cause, actions, verification and relevant configuration/part changes.