IPv6 Address Fundamentals

What you will be able to do

• Describe what an IPv6 address identifies and state its size.
• Distinguish unicast addresses from anycast addresses by the destination set they represent.
• Explain how prefixes and subnet prefixes organize IPv6 addressing.
• Recognize the purpose and required presence of Link-Local unicast addresses.

01

1. What an IPv6 address represents

An IPv6 address is a 128-bit identifier connected to an interface or to a group of interfaces.

Its basic job is to provide an identity that IPv6 communication can use when referring to network attachment points.

For example, one interface might hold a configured address that identifies that interface alone. Another address can represent several interfaces as one destination choice.

The phrase “address” does not always mean one physical device. IPv6 also supports addresses whose meaning spans multiple interfaces.

02

2. Unicast: one interface

An IPv6 unicast address identifies one interface. Communication sent toward that address has a single interface as its addressing target.

This makes unicast useful when a sender needs to name one particular interface rather than a collection of possible receivers.

Imagine a laptop interface configured with one unicast address. That address names the laptop’s specific interface, not every interface on its local link.

Unicast is different from anycast because unicast identifies one interface, while anycast identifies a set of interfaces.

03

3. Anycast: a set of interfaces

An IPv6 anycast address identifies a set of interfaces. Those interfaces are commonly located on different nodes.

Anycast lets one address represent several eligible interfaces. The address therefore describes a destination set instead of one uniquely named interface.

Consider three nodes configured as members of one anycast set. The shared address identifies the set, even though each member has its own interface.

Anycast addresses do not use a visibly special address shape. They come from the unicast address space and use unicast formats.

04

4. How an anycast set relates to topology

An assigned anycast address has a longest prefix that describes the topological region containing every interface in its set.

This prefix connects the shared address with the region where its members are located. It gives the set a topological boundary in the addressing plan.

For example, an anycast set could have members within one defined region. The set’s longest relevant prefix identifies that common region.

If the members do not share a topological locality, the set can use the null prefix. In that case, no narrower common region is available.

05

5. Prefixes organize addresses

IPv6 unicast addressing permits prefixes with any bit length. A prefix selects the leading portion of an address.

Prefixes can also be grouped with CIDR-style notation. This supports aggregation, where related address ranges are represented together.

A subnet prefix has a more specific role. It is associated with one link in the IPv6 addressing model.

A general address prefix and a subnet prefix are related, but they are not identical ideas. One describes address grouping, while the other associates an address range with a link.

06

6. Common IPv6 unicast types

IPv6 includes Global Unicast, Link-Local unicast, and site-local unicast address types. Site-local unicast is deprecated.

Link-Local unicast addresses are intended for communication on one link. Their scope is therefore limited to that single link.

For example, a Link-Local address can represent an interface for link-scoped IPv6 activity. It is not described as a general address for every network location.

Global Unicast and Link-Local unicast are distinct types. Site-local unicast is another named type, but it should be treated as deprecated.

07

7. One interface can hold several addresses

Every IPv6 interface must have at least one Link-Local unicast address.

An interface can also have multiple IPv6 addresses. These may include unicast, anycast, or multicast addresses with different scopes.

For example, one interface could have its required Link-Local address plus another unicast address. It could also participate in an anycast arrangement.

This means an interface is not limited to one address or one scope. Address count and address role can vary together.

08

8. Solicited-Node multicast membership

IPv6 computes a Solicited-Node multicast address from each configured unicast or anycast address.

This creates a multicast group associated with the configured address. The relationship is derived rather than chosen as an unrelated value.

For every unicast or anycast address configured on an interface, the device joins the corresponding derived multicast group.

If an interface has several configured unicast and anycast addresses, the node performs this process for each one.

09

9. Lesson recap: ready for Practice and Prove

IPv6 addresses are 128-bit identifiers for interfaces or sets of interfaces. Unicast names one interface, while anycast names a set.

Anycast uses unicast address formats, so its written form does not reveal the anycast role. Its topological prefix describes the shared region when one exists.

IPv6 prefixes can have arbitrary lengths and support aggregation. A subnet prefix associates one link with its addressing plan.

Link-Local unicast addresses serve a single link, and every IPv6 interface must have at least one. Site-local unicast is deprecated.

Interfaces may hold several addresses and scopes. For every configured unicast or anycast address, the node computes and joins the matching Solicited-Node multicast group.

In Practice and Prove, classify address roles, compare destination sets, interpret prefix relationships, and identify the required Link-Local and Solicited-Node behavior.