Learning paths › CCNA › 1. Network Fundamentals
IPv6 Addressing & Representation
Written reference: Explore the 128-bit IPv6 address architecture, hexadecimal notation, address abbreviation rules (leading zeroes, double colon), and IPv6 address types.
About 11 minutes
Why it matters
IPv6 addresses are long and easy to mistype. Knowing the shortening rules lets you read and write them accurately.
The Next Generation 128-Bit Internet Protocol
IPv4 provides 4.3 billion addresses (32-bit); IPv6 provides 340 undecillion addresses (128-bit), permanently solving address exhaustion.
IPv6 addresses are written in 8 groups of 4 hexadecimal digits (quartets/hextets) separated by colons.
Rule 1 of abbreviation: Omit leading zeros in any quartet.
Rule 2 of abbreviation: Replace contiguous groups of all-zero quartets with a double colon '::' (ONLY ONCE per address).
Address Compression Rules with Examples
Original Uncompressed Address: 2001:0db8:0000:0000:0000:ff00:0042:8329
Step 1: Remove Leading Zeros (Rule 1): 2001:db8:0:0:0:ff00:42:8329
Step 2: Compress Consecutive Zeros with '::' (Rule 2): 2001:db8::ff00:42:8329
Loopback Address Example: 0000:0000:0000:0000:0000:0000:0000:0001 -> Compressed: ::1
Unspecified / Default Route Address Example: 0000:0000:0000:0000:0000:0000:0000:0000 -> Compressed: ::
The words you need
IPv6 — Internet Protocol version 6. Uses 128-bit addresses (16 bytes) to replace IPv4.
Quartet / Hextet — A group of 4 hexadecimal digits (16 bits) in an IPv6 address (e.g. 2001, 0db8, fe80).
Leading Zero Compression — Dropping leading zeros inside any quartet (e.g. 0042 becomes 42, 0000 becomes 0).
Zero Compression (::) — Compressing one consecutive run of all-zero quartets into a double colon (e.g. 2001:db8:0:0:0:0:0:1 becomes 2001:db8::1).
The address types the exam expects you to recognise on sight
An IPv6 address is 128 bits, written as eight groups of four hex digits. Learn the prefixes that identify the type: global unicast starts 2000::/3 (in practice addresses beginning with 2 or 3), link-local is fe80::/10, unique local is fc00::/7 (in use as fd00::/8), multicast is ff00::/8, loopback is ::1/128 and the unspecified address is ::/128. There is no broadcast in IPv6; its jobs were taken over by multicast groups such as ff02::1 (all nodes) and ff02::2 (all routers). Anycast is not a separate range: it is a unicast address assigned to several devices, and the network delivers to the nearest.
Link-local addresses are mandatory and automatic on every IPv6-enabled interface; they only work on the local link, which is why a router can use one as the next hop for a static route or as the gateway a host learns from a Router Advertisement. The documentation prefix 2001:db8::/32 is reserved for examples, which is why you will see it in every lab and exam item.
Compression rules, and the mistake that appears in every set of options
Two rules only. First, leading zeros inside a group can be dropped (0db8 becomes db8, 0000 becomes 0). Second, one run of consecutive all-zero groups may be replaced by a double colon, and only one run, because two double colons would make the address ambiguous. If there are two equal runs, compress the longer one; if they tie, compress the first.
Worked example: 2001:0db8:0000:0001:0000:0000:0000:0005. Dropping leading zeros gives 2001:db8:0:1:0:0:0:5. There are two zero runs, a single group and a run of three, so the longer one is compressed: 2001:db8:0:1::5. The option 2001:db8::1::5 is invalid (two double colons), and 2001:db8:0:1:0:0:0:5 is valid but not fully compressed. To expand an address, count the groups that are present, then insert enough 0000 groups at the double colon to reach eight.
Prefix length and why /64 is everywhere
The prefix length works like IPv4 CIDR: the leading bits identify the network. Convention fixes the subnet for ordinary LANs at /64, leaving a 64-bit interface identifier. That is not just tidiness: SLAAC and EUI-64 assume a 64-bit interface ID, so a LAN that is not a /64 breaks stateless autoconfiguration. A site normally receives a /48 or /56, and the bits between that and /64 are the subnet field (a /48 gives 16 bits, so 65,536 /64 subnets). Point-to-point router links are sometimes /127 or /126 for neighbour-cache hygiene, but that is a design choice, not an exam default.
Exam framing: "Which of these is a valid IPv6 address?", "Which address type is fe80::1?", and "What is the shortest form of ...?". Verify real addressing on a router with show ipv6 interface brief, which lists the link-local and global addresses per interface; an interface with only a link-local address either has no global address configured or has not received one.
Check: address type
An interface shows the address fe80::2a0:c9ff:fe12:3456. What type of address is it?
- Link-local, valid only on the directly attached link
- Global unicast, routable on the internet
- Unique local, routable inside a site
- Multicast, delivered to a group
Takeaways
- Describe the 128-bit structure of an IPv6 address.
- Shorten an address using the leading-zero and double-colon rules.
- Name the main IPv6 address types.
Next: Network Diagnostics
More free lessons in 1. Network Fundamentals
- Network Fundamentals
- Network Devices
- Cables & Interfaces
- TCP/IP & OSI Models
- Switch Interfaces
- IPv4 Header
- Wireless RF & Frequencies
- IEEE 802.11 Standards & Wi-Fi Generations
- Cisco IOS CLI
- Ethernet Switching 1
- Ethernet Switching 2
- IPv4 Addressing
- IPv4 Configuration
- Subnetting Basics
- Finding Subnets Fast
- VLSM Subnetting
- Network Diagnostics
- IPv6 Address Scopes & Types
- IPv6 Configuration & EUI-64