UEU-co logo


Previous Page Next Page

IPv6 Addressing

IP Version 6 (IPv6) was developed under the auspices of the Internet Engineering Task Force (IETF) and was first defined in RFC (Request for Comments) 2460. In terms of functionality, IPv6 is kind of like an extension of IPv4, and the application and transport protocols in the TCP/IP stack don’t really need a lot of tweaking to work with IPv6. (FTP, believe it or not, is actually an exception.) IPv6 uses a 128-bit addressing system that provides 40,282,366,920,938,463,463,374,607,431, 768,211,456 addresses. This is certainly a very large number of IP addresses, and it should be a little more difficult to exhaust this pool of addresses when you compare it to the smaller pool of addresses that was provided by IPv4.

By the Way

You can take a look at the RFCs related to IPv6 (and IPv4) at the IETF Documents website. Go to http://tools.ietf.org/html/. You can also use any web search engine to locate RFCs on the Web.

The IPv6 address is a hexadecimal address that consists of eight 16-bit parts or blocks. Each 16-bit block is delineated by a colon (:). A sample address is


In terms of comparing an IPv6 address to how we view an IPv4 address (in terms of what part of the address provides the network ID and what part provides the host ID), IPv6 addresses basically split the available bits in half; the first 64 bits provide the network prefix and the last 64 bits provide the interface ID as shown in Figure 7.2. (IPv6 identifies interfaces rather than hosts.)

Figure 7.2. IPv6 addresses divide the bits in an address equally to identify the network prefix and the interface ID.

By the Way

IPv6 is a more secure addressing protocol than IPv4 in that IPSec (discussed in Hour 21, “Working with the Windows Firewall and IPSec”) is built into the protocol rather than an add-on, as it was with IPv4.

The network prefix is really a combination of information. The first 48 bits of the area designated as the network prefix (see Figure 7.2) is the site prefix, which would be the equivalent of the network ID in an IPv4 address. So, in the address


the FE80:BA98:7654, meaning the first three 16-bit parts (for a total of 48 bits), would be the public portion of the IPv6 address in that it identifies your network. The next 16-bit part in the address (the fourth 16-bit part) would then be used to identify the subnet ID; so it is the fourth 16-bit part of the IPv6 address that is used to create subnets for networks.

IPv4 addressing is approached by typically assigning a single IP address to each node on the network (although some servers obviously will have multiple IPv4 addresses for load balancing and special services that require more than one network interface, such as routing or Network Address Translation). IPv6 addressing is approached differently and each interface on a node (a node being a computer or pretty much any type of IP-enabled device) is assigned both a global and a link-local address.

The global or global unicast address is the equivalent of a public IPv4 address and is used to route data to other networks (or links, as they are referred to in IPv6 lingo). The link-local address is used by nodes to communicate with other nodes the local network or link.

By the Way

IPv6 really assigns three IP addresses to an interface: link-local, global, and loopback. The loopback address for IPv6 is ::1.

So, the question then becomes, How are IPv6 addresses assigned to nodes (and their interfaces) on the network? You have three alternatives: the new IPv6 self-addressing strategy, termed stateless auto-configuration; a DHCP (DHCPv6) server; or manually assigned, static IPv6 addresses.

Stateless auto-configuration enables an interface to dynamically assign itself an IPv6 link-local address. The address is generated from the interface’s MAC hardware address (which is 48 bits; an additional 16 bits are added to make the 64-bit interface ID portion of the IPv6 address). Figure 7.3 shows the auto-configuration link-local address generated by an IPv6-enabled server.

Figure 7.3. The IPv6 link-local address is auto-generated by the IP client.

[View full size image]

In terms of DHCP, you can deploy a DHCP server running Windows Server 2008 to provide IPv6 addresses to network clients. DHCP is discussed in Hour 16, “Using the Dynamic Host Configuration Protocol.”

As already mentioned, you can also configure a network client with a static IPv6 address. When you specify a static IPv6 address for an interface (configuring TCP/IP settings is discussed in the next section), the auto-configuration is still enabled and so an auto-configured address as well as the static address (you assign) is assigned to the interface.

In terms of subnetting, you already saw that the fourth 16-bit portion of the IPv6 address contains the subnet ID. This provides you with 16 possible bits for your IPv6 subnets (which can then be configured as scopes on subnet DHCP servers). These “subnet prefixes” can actually be advertised by network routers and supplied to IPv6 clients in a specific subnet.

Working with IPv6 addressing is obviously a little more intimidating when compared to working with IPv4 addressing, mainly because IPv6 addresses are larger; after you get used to working in hexadecimal rather than dotted decimal, you will find that IPv6 addressing really is easier to work with and you no longer have to deal with subnet masks.

By the Way

IPv6 does not use broadcasts the way that IPv4 does. This helps cut down on network broadcast traffic on subnets. IPv6 uses multicast addressing and multicasts, negating the need for broadcasts to all nodes.

Previous Page Next Page

Leave a Reply

Time limit is exhausted. Please reload the CAPTCHA.


apply_now Pepperstone Group Limited