Computer ScienceFoundation18 min read

System Troubleshooting

Finding the fault by narrowing it down, not by guessing

This topic appears in:

01

A method, not a list of fixes

Troubleshooting questions are not asking you to memorise every possible fault. They are asking whether you can work through a problem systematically, and the marks follow the method.

The approach is the same one used in every diagnostic discipline: find out exactly what is wrong, form a theory about why, test that one theory, and only then change something. Changing several things at once is the classic beginner error, because when it works you still do not know what was broken — and if it does not, you now have three new variables.

  • 1. Identify the problem. What exactly happens, and when? Ask what changed recently — new software, a new cable, a power cut.
  • 2. Form a theory. What single cause would explain all the symptoms? Start with the simplest and most likely.
  • 3. Test the theory. Change one thing only, so the result means something.
  • 4. Fix, then verify. Confirm the original symptom has gone and that nothing else has broken.
  • 5. Document. Record the fault and the fix, so the next person — or you in six months — does not start from nothing.

Check the simplest cause first

A monitor showing nothing is far more often unplugged than faulty. A printer that will not print is more often out of paper than broken. Working from the cheapest, most likely explanation upwards is not a lack of ambition — it is what keeps you from replacing a working graphics card. Examiners award marks for starting with power, cables and connections.

02

Is it hardware or software?

The single most useful early question, because it halves the search. Two tests usually settle it.

Does the fault survive a restart into a different environment? A problem that disappears when you boot from a USB drive is almost certainly software. Does the fault appear on other hardware? A file that fails to open on three machines is a file problem, not a machine problem.

SymptomLikely causeFirst thing to try
No power at allpower supply or socketcheck the cable and try another socket
Beeps but no displayRAM or graphics card seated badlyreseat the RAM
Very slow, fan loudoverheating or too little RAMclean the fan, check memory usage
Random crashesfaulty RAM or overheatingrun a memory test
Will not boot into the OScorrupt system files or a failing diskboot from recovery media
One program crashes onlythat program or its datareinstall the application
No networkcable, driver or configurationcheck the cable, then ping the router
03

Preventive maintenance

Most faults are cheaper to avoid than to diagnose, and the syllabus expects you to know the routine tasks.

On the software side: keep the operating system and applications updated so that security holes are closed, run antivirus scans, remove software no longer used, and take regular backups. On the hardware side: keep vents and fans clear of dust, since overheating causes both slowdowns and permanent damage; use a surge protector or UPS in areas with unstable power; and handle components by their edges, using an anti-static wrist strap where possible.

The 3-2-1 rule for backups

Keep three copies of anything important, on two different kinds of media, with one of them off-site. A backup sitting on the same machine as the original is not a backup — a theft, a fire or a ransomware infection takes both. This rule appears in exam answers about data loss and is worth quoting by name.

04

Working through a fault

Exam questions give you a scenario and ask for the steps. Write them as steps, in order, each one testing a single possibility.

Worked example

A user reports that their computer turns on — fans spin, lights come on — but the monitor stays completely blank. Describe how you would diagnose the fault.

  1. Confirm the symptom: is the monitor's own power light on?A dark monitor with no power light is a monitor or socket problem, not a computer problem. This one question splits the search in half.
  2. Check the video cable at both ends, and that the correct input source is selected on the monitor.Cables and input selection are the most common causes and cost nothing to check.
  3. Try the monitor on a different computer, or a different monitor on this one.Swapping one component at a time is the only way to tell which of the two is at fault.
  4. If the monitor is fine, listen for beep codes at startup and reseat the RAM and the graphics card.Beeps are the machine reporting a fault before any display is available. Badly seated RAM is a very common cause of exactly this symptom.
  5. Document what was found and what fixed it.The last step is a mark in most schemes, and it is the one students omit.

Check monitor power and input, then the cable, then swap to isolate, then reseat RAM and graphics card — one change at a time.

Before you leave this chapter

  1. Identify → theorise → test → fix → verify → document.
  2. Change one thing at a time, or the result tells you nothing.
  3. Check power, cables and connections before suspecting a component.
  4. Swapping a suspect part into a known-good machine isolates the fault.
  5. Backups follow 3-2-1: three copies, two media, one off-site.
05

Narrowing it down question by question

A good first question halves the search. A bad one tells you nothing you did not already know, and you have spent time to learn it.

For each common symptom there is one cheap question that splits the possibilities cleanly in two. Asking it first is the difference between a five-minute diagnosis and an afternoon of swapping parts.

Each step is a single check that eliminates half of what is left. Notice the first question in every case costs nothing — no part is replaced until three free checks have failed.

Practice questions

6 questions · 20 marks · full working on every one

Try each one on paper first, then open the working. The marks are shown where they are actually awarded, because that is where they are actually lost.

Short questions

3 · 6 marks

Two marks each, in the style of the short-question section of the paper. Answer in two or three lines.

SQ1[2 marks]
Why should only one change be made at a time when troubleshooting?
Model answer

So that the outcome identifies the cause. If several things are changed together and the fault disappears, you do not know which change fixed it — and if it does not, you have introduced new variables that must themselves be undone.

Examiner tip. Give both halves: what you learn when it works, and what you lose when it does not.

SQ2[2 marks]
State two software-based preventive maintenance tasks.
Model answer

Any two of: installing operating system and application updates; running antivirus scans; taking regular backups; uninstalling unused software; clearing temporary files.

Examiner tip. The question says software-based, so cleaning fans and checking cables earn nothing here. Read the qualifier.

SQ3[2 marks]
A computer has started shutting down unexpectedly during heavy use. Suggest a likely cause and one check.
Model answer

Overheating is the most likely cause, since the fault appears under load. Check whether the fans are running and whether the vents are blocked with dust, and monitor the CPU temperature.

Examiner tip. "Under heavy use" is the clue the examiner has planted. A fault that only appears under load points to heat or to power, not to software.

Solved numericals

2 · 8 marks

Full working, one step per line, with the marks shown where they are awarded.

N1[4 marks]
Describe, in order, the stages of a systematic troubleshooting process.
Full working
  1. Identify the problem — establish exactly what happens and what changed recently[1]
  2. Establish a theory of the probable cause, starting with the simplest and most likely[1]
  3. Test the theory by changing one thing, and implement the fix if confirmed[1]
  4. Verify that the fault is gone and nothing else has broken, then document the fault and the solutiondocumentation is regularly omitted and regularly marked[1]

Identify, theorise, test, fix, verify, document.

Examiner tip. Give them in order and name each stage. An unordered list of sensible actions usually scores half.

N2[4 marks]
Explain how you would determine whether a fault is caused by hardware or by software.
Full working
  1. Boot the machine from alternative media, such as a USB drive with a live operating system[1]
  2. If the fault disappears in that environment, the installed software is at fault; if it persists, the hardware isthe interpretation is the mark, not the action[1]
  3. Test the suspect component in a known-good machine, or a known-good component in this machinecomponent swapping[1]
  4. Check whether the same fault occurs in more than one application — a fault in only one program points to that program rather than the hardware[1]

Boot from other media, swap components, and check whether the fault is confined to one application.

Examiner tip. Every test here works by changing one variable and seeing whether the fault follows it. Saying so explicitly shows the examiner you understand why the tests work.

Long questions

1 · 6 marks

Theory and numerical together, as they appear in the long-question section.

LQ1[6 marks]
A school laboratory has repeated problems: machines run slowly by the end of term, and one machine recently lost a teacher's files when its disk failed.
  1. Suggest two likely reasons for the machines slowing down over a term.
  2. Describe a backup strategy that would have protected the lost files.
  3. Suggest one hardware maintenance task the school should schedule.
Mark scheme
  1. Accumulated temporary files, downloads and unused installed software filling the disk[1]
  2. Too many programs configured to start automatically, consuming RAM, and dust build-up causing the CPU to slow itself to avoid overheatingany second sensible cause[1]
  3. Follow the 3-2-1 rule: three copies of the data[1]
  4. On two different types of media, with one copy kept off-site or in cloud storage[1]
  5. Backups should be automatic and scheduled rather than left to the user, and should be tested by restoring from theman untested backup is not known to work[1]
  6. Hardware maintenance: clean dust from fans, vents and heatsinks each term, and check that all fans still spinaccept surge protection or checking disk health[1]

(a) accumulated files and startup programs, plus dust causing thermal throttling (b) 3-2-1, automated and tested (c) scheduled cleaning of fans and vents

Examiner tip. The detail that lifts a backup answer to full marks is testing the restore. A backup nobody has ever restored from is an assumption, not a safeguard.