Computer networks: practice questions with explanations

1What a computer network is

A network is a group of devices that exchange data: laptops, phones, servers, printers. The devices on it are called hosts.

A network inside one building or campus is a LAN (local area network). A network across cities or countries is a WAN (wide area network). The internet connects millions of networks.

Data does not travel as one long stream. It is cut into small pieces called packets, and each packet finds its own way to the destination.

Devices can only understand each other if they follow the same rules. Such a set of rules for the format and order of messages is a protocol. TCP, IP and HTTP are protocols.

Most services follow the client–server model: a client, such as your browser, sends a request, and a server sends back the answer.

Two words describe how fast a connection is. Bandwidth is the maximum data rate, for example 100 Mbit/s. Latency is the delay before the data arrives, measured in milliseconds.

2Layers and encapsulation

Networking is split into layers. Each layer solves one problem and uses the layer below it. The OSI model has 7 layers; the TCP/IP model used on the internet groups them into 4 or 5.

• 1 Physical: sends raw bits as electrical, light or radio signals over cables and air.
• 2 Data link: delivers frames between devices on the same local network, using MAC addresses. Ethernet and Wi-Fi live here.
• 3 Network: delivers packets across networks, using IP addresses. Routers work here.
• 4 Transport: delivers data between programs, using port numbers. TCP sends segments; UDP sends datagrams.
• 5–7 Session, presentation, application: the services people use, such as HTTP for the web, DNS for names and SMTP for email.

When data is sent, every layer adds its own header in front of the data from the layer above. This is called encapsulation. The receiver removes the headers in reverse order.

TCP sets up a connection first (the three-way handshake SYN, SYN-ACK, ACK). It numbers the data, waits for acknowledgements and resends what is lost. It is reliable and ordered.

UDP sends without a connection and without resending. It is faster and is used where delay matters more than a lost packet: live video, games, and short DNS queries.

3Addresses, devices and helper protocols

Three kinds of address work together, one per layer:

• MAC address (layer 2): a 48-bit hardware address of a network card. It is used only inside the local network.
• IP address (layer 3): an IPv4 address has 32 bits, an IPv6 address 128 bits. It says where a host is on the internet.
• Port number (layer 4): picks the program on the host. Well-known ports run from 0 to 1023: HTTP 80, HTTPS 443, SSH 22, DNS 53, SMTP 25.

The devices follow the same layers. A hub (layer 1) copies every signal to all ports. A switch (layer 2) sends a frame only to the port of the destination MAC address. A router (layer 3) forwards packets between networks.

A few helper protocols make this work:

• DHCP gives a host its IP address, mask, default gateway and DNS server when it joins the network.
• DNS translates a name such as example.com into an IP address, like a phone book.
• ARP finds the MAC address that belongs to an IP address on the local network.
• ICMP carries error and test messages. The ping command uses ICMP echo requests.
• NAT lets many devices with private addresses (10.x, 172.16–31.x, 192.168.x) share one public address.

The default gateway is the router a host sends every packet to when the destination is outside its own subnet.

4The method step by step

Most questions, from definitions to troubleshooting cases, can be answered with the same steps.

1. Find the layer. Is the question about signals, frames and MAC addresses, packets and IP addresses, ports and connections, or an application?
2. Name the device or protocol that works at that layer: hub, switch or router; Ethernet, IP, TCP or UDP; DNS, DHCP or HTTP.
3. For a fault, list what still works and what fails. Something that works proves every layer below it is fine.
4. Look for the lowest layer that can explain everything that fails. That is the most likely cause.
5. Check your answer against the other facts in the question: it must explain all of them.

For short calculations: transmission delay is packet size divided by link rate, and propagation delay is distance divided by signal speed.

5A worked example

Follow every step. Try to predict the next one before you open it.

6Common mistakes

• Mixing up DNS and DHCP. DNS translates names into addresses; DHCP hands out addresses.
• Mixing up a switch and a router. A switch uses MAC addresses inside one network; a router uses IP addresses between networks.
• Thinking UDP is “bad TCP”. It is chosen on purpose when low delay matters more than resending lost data.
• Confusing bandwidth (capacity) with latency (delay).
• Putting port numbers at the network layer. Ports belong to the transport layer; IP addresses to the network layer.
• Forgetting the default gateway. Without it, a host reaches its own subnet but nothing beyond it.

7How exam questions look

• Easy: key terms and facts. The number of OSI layers, what DNS or DHCP does, default port numbers, the size of an IPv4 address.
• Medium: applying the concepts. The layer of a device, why a protocol is chosen, what ARP, NAT or the TCP handshake does, small subnet and delay calculations.
• Hard: cases and analysis. Troubleshooting scenarios, TCP behaviour under loss and congestion, routing protocols, attacks such as ARP spoofing or SYN floods, and TLS.

Every question has four options. Eliminate the options that belong to the wrong layer first; often only one is left.