Reason about delay and capacity

Reason about delay and capacity — guided practice

Calculate a simplified transfer bound and distinguish it from latency.

Use network-notes.md or paper. All addresses and outputs are fictional. Room A has laptop A at 192.0.2.10/24, printer P at 192.0.2.20/24 and a switch. Router R has A-side 192.0.2.1/24 and B-side 198.51.100.1/24. Room B has server S at 198.51.100.20/24 and its switch. A uses default gateway 192.0.2.1. A supplied DNS record maps catalog.example to 198.51.100.20. For this /24 exercise, matching first three octets means same subnet. No live network access, capture, scanning or device configuration is needed.

1. 1. Read a link card

   A fictional bottleneck carries 20 megabits per second; a payload contains 100 megabits. Ignore overhead.

   You should see: Units are bits, not bytes.

   Consistent units are needed for division.

2. 2. Calculate a bound

   Divide payload size by bottleneck rate.

   100 / 20 = 5

   You should see: 100/20=5 seconds minimum serialization time under these assumptions.

   A rate translates quantity into time.

3. 3. Add a stated delay

   The case adds a fixed 0.2-second delay and no other costs.

   You should see: Simplified total=5.2 seconds.

   A startup delay is distinct from sustained throughput.

4. 4. Explain real limits

   List shared traffic, overhead and loss as reasons an actual run may take longer.

   You should see: The calculation is a model, not a performance guarantee.

   Naming omitted factors prevents false precision.