Skip to content

📘 03-03: IPv6 Addressing

Network Systems

Module 03: Addressing

Home All Notes Practice Quiz

💡 Why IPv6?

IPv4's 32-bit space (~4.3 billion addresses) is exhausted. IPv6 uses 128 bits, providing a virtually unlimited address space.


📌 Format

Eight groups (hextets) of four hex digits, separated by colons:

2001:0db8:85a3:0000:0000:8a2e:0370:7334

🔹 Compression rules

  1. Drop leading zeros in each hextet: 0db8 → db8, 0000 → 0.
  2. Replace one run of consecutive all-zero hextets with :: (only once).
2001:0db8:85a3:0000:0000:8a2e:0370:7334
2001:db8:85a3::8a2e:370:7334        (compressed)

📌 Address Types

Prefix Type Notes
2000::/3 Global unicast Internet-routable
fe80::/10 Link-local Valid only on the local link
::1/128 Loopback Equivalent to 127.0.0.1
ff00::/8 Multicast No broadcast in IPv6
fc00::/7 Unique local Private-like (ULA)

📌 Interface ID & Autoconfiguration

  • A LAN subnet is almost always /64: first 64 bits = network prefix, last 64 = interface ID.
  • EUI-64 derives the interface ID from the 48-bit MAC (insert FFFE, flip U/L bit).
  • SLAAC lets a host build its own address from the Router Advertisement prefix + interface ID — no DHCP needed.
  • NDP (ICMPv6) replaces ARP for neighbor/router discovery.

Tip

Key idea — IPv6 = 128 bits, eight hextets, compressed with :: (once) and dropped leading zeros. LANs use /64; hosts can self-configure via SLAAC/EUI-64.

See also: IPv4 Addressing, Subnets in IPv6


Previous: IPv4 Addressing  |  All Notes  |  Module 03 Exercise  |  Next: DHCP