Menu

Earn Premium with Referrals

Invite your friends and earn Premium rewards through our referral program.

See how it works and start inviting friends.

Network Fundamentals & Models
CN

Network Fundamentals & Models

Practice questions on OSI and TCP/IP models, network topologies, nodes, links, communication concepts, and networking fundamentals.

1. Which layer of the OSI model handles physical addressing, node-to-node data transfer, and operates using MAC addresses and switches?

That’s the Data Link Layer (Layer 2).

Its job is node-to-node (device-to-device) delivery on the same local network. It frames the data, handles error detection, and uses physical addressing — the MAC address burned into each network interface.

Why switches belong here: a switch is a Layer 2 device. It reads MAC addresses and forwards frames to the specific port where the destination device lives, rather than broadcasting everywhere.

LayerNameAddress type
3NetworkIP
2Data LinkMAC
1PhysicalNone (bits)

2. What is the primary operational difference between a Layer 2 Switch and a Layer 1 Hub?

The difference is intelligence.

  • Hub (Layer 1) — blindly broadcasts incoming data to every port. No understanding of addresses, just electrical signal repetition.
  • Switch (Layer 2) — learns which MAC address lives on which port, builds a MAC address table, and forwards frames only to the intended destination.

The consequences:

  • A hub creates one big collision domain — traffic from every port collides, wasting bandwidth.
  • A switch creates a separate collision domain per port and keeps traffic isolated.

Example: if 4 devices on a hub talk, all hear everything. On a switch, only the target device receives its frames.

  • Node — any physical device on the network that can communicate: computers, routers, switches, phones.
  • Link — the transmission path connecting two nodes: copper cable, fiber optic, or a wireless channel.
[Node A] ─────── link (cable/fiber/wireless) ─────── [Node B]

The link is the medium; the node is the endpoint. A network is simply a collection of nodes joined by links.

4. Which communication mode allows simultaneous bidirectional data transmission between two endpoints?

Full-Duplex.

  • Simplex — one-way only (e.g., keyboard → computer).
  • Half-Duplex — two-way, but only one side transmits at a time (e.g., walkie-talkie).
  • Full-Duplex — both directions simultaneously (e.g., telephone call, modern switched Ethernet).
Simplex:     →  only one direction
Half-duplex: ↔  but one at a time
Full-duplex: ⇄  both at once

Modern switches use full-duplex: a device can send and receive at the same time.

5. Which physical network topology connects all endpoints directly to a single central hardware device, ensuring that an individual cable break only affects a solitary node but creating a single point of failure?

Star Topology.

        [Node]

[Node] ──[Hub]── [Node]

        [Node]

All devices connect to one central hub or switch.

  • Good: one cable break only kills that one node.

  • Bad: the central device is a single point of failure — if the hub dies, the whole network dies.

  • Mesh — every node connects to many others (maximum redundancy).

  • Bus — all nodes share one backbone cable.

  • Ring — each node connects to exactly two neighbors in a loop.

Star is the standard for modern LANs because the central switch is cheap and failures are easy to isolate.

6. How is a Hybrid Topology defined in enterprise network design?

A hybrid topology combines two or more distinct base topologies into one network.

      Star-Bus hybrid
   ┌────┴────┐        ┌────┴────┐
 [H1] [H2] [H3]  bus  [H4] [H5] [H6]
   (star workgroup)   (star workgroup)

Common examples:

  • Star-Bus — star workgroups hanging off a shared bus backbone.
  • Star-Ring — star workgroups connected in a ring at the core.

Why use one? Different parts of an organization have different needs. A department might want a star for fault tolerance, while the backbone between buildings is a ring or bus. Hybrid gives you the strengths of each where they matter — at the cost of complexity.

7. What is a Multi-Homed Host in network infrastructure architecture?

A multi-homed host is a device with two or more active network interfaces connected to separate networks at the same time.

      Internet           Internal LAN
        │                     │
   ┌────┴────┐           ┌────┴────┐
   │   NIC 0 │  Server   │ NIC 1   │
   └─────────┘           └─────────┘

Why run multi-homed?

  • Redundancy — if one link fails, traffic moves to the other.
  • Load balancing — spread traffic across links.
  • Routing — a box that bridges two separate networks.

A classic example is a firewall or proxy with one NIC on the internet-facing network and another on the trusted internal network.

8. What is the function of a Pseudo TTY (Teletypewriter) in modern server operating systems?

A pseudo TTY (PTY) is a software-emulated virtual terminal that behaves like a real physical terminal (TTY).

SSH client ──→ [SSH daemon] ──→ PTY (virtual terminal)

                              [shell / command interpreter]

Why it’s needed: programs like SSH and Telnet need to talk to a shell (like bash), and the shell expects a terminal device. A PTY fakes that — it presents the shell with a terminal interface while the actual bytes travel over a network connection.

Without PTYs, you couldn’t run interactive programs over SSH. It’s the plumbing that makes remote command-line access work.

9. What is the primary role of a Backbone Network?

A backbone is the high-capacity central pathway that interconnects different network segments — buildings, floors, or geographic regions.

        [Floor 3 switch]──┐
        [Floor 2 switch]──┼──[BACKBONE core]──┼── [Remote site router]
        [Floor 1 switch]──┘                     └── [Data center]

Characteristics:

  • High speed (fibre, 10/40/100 Gbps)
  • Carries aggregated traffic between segments
  • The core of a hierarchical network

Every floor/branch/local network feeds its traffic up into the backbone, which moves it between segments. If the backbone dies, the segments are isolated from each other — so it’s built for redundancy.

10. What are the 7 layers of the OSI model and how does it compare to the TCP/IP model?

The OSI model has 7 layers, top to bottom:

  1. Application — user-facing services (HTTP, FTP, SMTP, DNS).
  2. Presentation — encoding, encryption, compression.
  3. Session — establishes and manages connections/sessions.
  4. Transport — end-to-end delivery, segmentation, reliability (TCP/UDP).
  5. Network — logical addressing and routing (IP).
  6. Data Link — physical addressing, framing, error detection (MAC, switches).
  7. Physical — bits over the medium (cables, radio, hubs).

The TCP/IP model condenses this to 4 layers: Application (OSI 5–7), Transport (4), Internet (3, = IP), and Network Access/Link (OSI 1–2). TCP/IP was the practical model that actually got built; OSI was the theoretical ideal.

OSITCP/IPExample protocols
ApplicationApplicationHTTP, DNS, SMTP
Presentation(merged)
Session(merged)
TransportTransportTCP, UDP
NetworkInternetIP, ICMP
Data LinkLinkEthernet, ARP
PhysicalLinkCables, Wi-Fi

The interview one-liner: OSI is the conceptual 7-layer reference; TCP/IP is the 4-layer model the real internet follows. Remember the layer order with a mnemonic like “Please Do Not Throw Sausage Pizza Away” (Physical→Application).

11. What is encapsulation and what are the PDU names at each layer?

Encapsulation is how data gains headers as it moves down the stack — each layer wraps the data from the layer above with its own header.

The PDU (Protocol Data Unit) name changes at each layer:

Data (application)  →   HTTP request
Segment (transport) →   TCP adds port numbers + sequence
Packet (network)    →   IP adds source/dest IP
Frame (datalink)    →   Ethernet adds MAC addresses + FCS
Bits (physical)     →   electrical/optical signals on the wire

So the PDU ladder is data → segment → packet → frame → bits (TCP/IP terms; OSI calls the network PDU a packet and datalink a frame, transport a segment).

On the receiving side, the process reverses: each layer removes its header (decapsulation) and passes the payload up. The interviewer favourite: “what is a TCP segment wrapped in an IP packet wrapped in an Ethernet frame called at each step?” — the answer is the ladder above. Encapsulation is why a single message can traverse many networks without each device needing to understand the other layers.

My Private Notes

Notes are auto-saved locally to this device.