Networking 104:
UNDERSTANDING IP ADDRESSES

Primary Functions

• Identification – Uniquely identifies a device (host) on a network.
• Location addressing – Helps in routing packets from the source to the destination.

IP Address Format

When data travels across a network, it's broken into packets. The IP layer ensures each packet is tagged with:

This system allows routers and devices to decide the best path for forwarding each packet. Importantly, IP doesn't care if a packet arrives or not—it just handles addressing and routing. Reliability is the job of other protocols (like TCP).

OSI Model Context

IP operates at the Network Layer (Layer 3) of the OSI model, providing logical addressing and routing services between different networks.

Checking IP Address in Linux

You can check your IP address in Linux using the ifconfig command:

ifconfig

Why Was IP Needed?

Before the IP protocol, there was no standardized way for machines across different networks to identify each other and communicate. IP was developed to:

• Enable communication across heterogeneous systems
• Support scalability beyond local networks
• Establish a uniform addressing scheme
• Provide network-layer packet delivery

With the internet growing from a few hosts to billions of devices, IP became the backbone of modern digital communication.

Types of IP: IPv4 vs IPv6

IPv4 vs. IPv6: Why Two Versions?

IPv4 was the first widely deployed version of the Internet Protocol, but with only about 4.3 billion unique addresses, it couldn’t scale to support the explosive growth of devices. To address this, IPv6 was developed, offering a virtually unlimited address space and modern enhancements such as built-in security, automatic configuration, and improved mobility support.

IPv4 and IPv6 are not directly interoperable; they use completely different addressing formats and packet structures. Some dual-stack systems can handle both, but in general, IPv6 is not backward compatible with IPv4.

IPv4 is simpler, more widespread, and better supported by legacy systems. IPv6, while more complex, enables true end-to-end connectivity, eliminates NAT in many cases, and is better suited for the future of networking.

Breaking Down an IP Address (IPv4)

An IPv4 address is a 32-bit binary number, written in human-readable decimal form as four "octets" separated by dots. Each octet represents 8 bits, making a total of 32 bits:

What Is an Octet? An octet is simply 8 bits (1 byte). Each octet can represent a number from 2² (1) to 2⁸ – 1 = 255. So each octet can hold values from 0 to 255, giving 256 total possible values per octet.

Why 2⁸? Each bit has 2 possible values: 0 or 1. So:

• 1 bit → 2 values
• 2 bits → 2² = 4 values
• 3 bits → 2³ = 8 values
• ...
• 8 bits (an octet) → 2⁸ = 256 values (from 0 to 255)

IPv4 Address Classes (A–E)

Class Descriptions:

• Class A: Starts with binary 0, reserved for large organizations or ISPs, first octet: 1–126 (127 is loopback)
• Class B: Starts with binary 10, balanced between network and host bits, often used by universities, medium orgs
• Class C: Starts with binary 110, supports many small networks, default choice for small businesses
• Class D: Starts with binary 1110, reserved for multicast (1-to-many) communication
• Class E: Starts with binary 1111, reserved for research, experimental use only

Each IP address is typically divided into:

• Network part: Identifies the network
• Host part: Identifies the device on that network

Depending on the subnet mask, the bits are split to define what portion is network vs. host.

Subnet Masks for IP Classes

Formula for usable hosts: 2^host bits − 2 (network + broadcast reserved)

What Does a Subnet Mask Do?

• Separates the IP address into network and host portions
• Helps routers and devices determine if the destination device is local (same subnet) or remote (needs to go through a gateway/router)

Custom Subnetting: Making Networks Fit Your Needs

Subnetting is the process of dividing a large network into smaller, more manageable pieces called subnets. Instead of sticking to the default class sizes (like Class C's 254 hosts), you can create custom subnet sizes to better match the number of devices you need to support.

How It Works Every subnet mask tells us how many bits are used for the network and how many for hosts. When we borrow bits from the host part, we create more subnets but get fewer hosts per subnet.

Now your single /24 network becomes four smaller ones:

Public vs. Private IP Addresses

IP addresses are divided into two broad categories: public (globally routable) and private (used inside local networks). Devices on a private network need a public IP to communicate over the internet usually handled via NAT (Network Address Translation).

• Public: Globally unique IPs assigned by your ISP, visible on the internet.
• Private: IPs reserved for local use within LANs, not routable on the public internet.

Special IPv4 Address Ranges

DHCP Discovery Process

When a device joins a network, it sends a broadcast to find available DHCP servers. The first DHCP server to respond offers an available IP and configuration options.

Security Considerations

However, DHCP lacks authentication. There's no built-in mechanism to verify the identity of the DHCP server or client. This makes it vulnerable to spoofing attacks where a rogue DHCP server can assign incorrect settings to mislead or intercept traffic.

Client → DHCP Discover (Broadcast)

Client sends a broadcast to find available DHCP servers.

Server → DHCP Offer

One or more servers respond with an available IP and configuration options.

Client → DHCP Request

Client requests the offered IP from a specific DHCP server (based on the offer it chooses).

Server → DHCP Acknowledgment (ACK)

Server confirms lease of the IP and provides full network configuration.

Additional DHCP Messages

Beyond the basic DORA process, DHCP includes several other important message types:

• DHCP NAK: If the server denies the request (e.g., IP conflict), it sends a negative acknowledgment.
• Lease Renewal: Client will later re-request lease before it expires (handled via unicast, not broadcast).
• DHCP Decline: If the client detects that the offered IP is already in use, it may send a decline message.

Key Functions

• Inform source IP when delivery of packets fails (e.g., destination unreachable, time exceeded).
• Helps identify problems like routing failures or unreachable hosts.
• Operates at Network Layer alongside IP.

ICMP and Traceroute

Traceroute uses ICMP to map the path packets take across an IP network. Each router that decrements the TTL to 0 sends back a "Time Exceeded" ICMP message, revealing its IP address along the route.