Predict what a VPN changes, then verify each prediction
Task
Write down what you expect a VPN to change about a machine's view of the network, then measure each one -- because the surprises are where the model is wrong, and the model is what answers the exam's scenario questions.
Steps
- BEFORE connecting, write
lab/vpn/predictions.csvwith headeritem,before,predicted_aftercovering: apparent public address, default gateway, configured DNS servers, reachability of a device on your own local network, and the route table's default entry. - Capture the actual before-state for each item into
lab/vpn/before.txt. - Connect, capture the after-state into
lab/vpn/after.txt, and filllab/vpn/actual.csvwith headeritem,before,after,matched. - Record in
lab/vpn/tunnel.mdwhether the connection is split or full tunnel, and state the evidence -- the route table, not the vendor's description. - Disconnect, confirm the before-state returns, and record it in
lab/vpn/restored.txt. Where you had no VPN available, writelab/vpn/skipped.mdnaming which steps were not run and why.
Verify
awk -F, 'NR>1 && NF>=3 {n++} END {print n" prediction(s)"}' lab/vpn/predictions.csv
test -s lab/vpn/before.txt && echo "before captured"
awk -F, 'NR>1 && NF>=4 {n++} END {print n" item(s) compared"}' lab/vpn/actual.csv 2>/dev/null || test -s lab/vpn/skipped.md && echo "skip documented"
grep -Eic 'split|full' lab/vpn/tunnel.md 2>/dev/null || echo "tunnel note skipped"
Five predictions written before any measurement, a captured before-state, and either a completed comparison or a documented skip. A lab that cannot be run completely should say so precisely rather than be reported as done.
Notes
The prediction file must be written first. Filling it in after the measurement turns a test of your model into a transcription exercise, and the whole value here is finding out where the model was wrong.
This is an independent study companion for CompTIA A+ Core 1 220-1201 and is not produced by or endorsed by CompTIA.