Please Whitelist This Site? I know everyone hates ads. But please understand that I am providing premium content for free that takes hundreds of hours of time to research and write. I don't want to go to a pay-only model like some sites, but when more and more people block ads, I end up working for free. And I have a family to support, just like you. :) If you like The TCP/IP Guide, please consider the download version. It's priced very economically and you can read all of it in a convenient format without ads. If you want to use this site for free, I'd be grateful if you could add the site to the whitelist for Adblock. To do so, just open the Adblock menu and select "Disable on tcpipguide.com". Or go to the Tools menu and select "Adblock Plus Preferences...". Then click "Add Filter..." at the bottom, and add this string: "@@||tcpipguide.com^$document". Then just click OK. Thanks for your understanding! Sincerely, Charles Kozierok Author and Publisher, The TCP/IP Guide

NOTE: Using software to mass-download the site degrades the server and is prohibited. If you want to read The TCP/IP Guide offline, please consider licensing it. Thank you.

The TCP/IP Guide  9  TCP/IP Lower-Layer (Interface, Internet and Transport) Protocols (OSI Layers 2, 3 and 4)       9  TCP/IP Internet Layer (OSI Network Layer) Protocols            9  Internet Protocol (IP/IPv4, IPng/IPv6) and IP-Related Protocols (IP NAT, IPSec, Mobile IP)                 9  Internet Protocol Version 6 (IPv6) / IP Next Generation (IPng)                      9  IPv6 Addressing

IPv6 Address Space Allocation 1 2 3 4 5 IPv6 Interface Identifiers and Physical Address Mapping

Generic Division of the Unicast Address Space

The most generic way of dividing up the 128 bits of the unicast address space is into three sections, as shown in Table 61.

The global routing prefix and subnet identifier represent the two basic levels at which addresses need to be hierarchically-constructed: global and site-specific. The routing prefix consists of a number of bits that can be further subdivided according to the needs of Internet Registries and Internet Service Providers, to reflect the topography of the Internet as a whole (we'll discuss this soon). The subnet ID gives a number of bits to site administrators for creating their own internal network structure.

In theory, any size for “n” and “m” could be used. The implementation chosen for IPv6, however, assigns 48 bits to the routing prefix and 16 bits to the subnet identifier. This means 64 bits are available for interface identifiers, which are constructed based on the IEEE “EUI-64” format, as described in the next topic. Thus, the overall IPv6 unicast address format is constructed as shown in Table 62 (illustrated in Figure 96.)

Due to this structure, most end sites (regular companies and organizations, as opposed to Internet service providers) will be assigned IPv6 networks with a 48-bit prefix. In common parlance, these network identifiers have now come to be called 48s or /48s.

The 16 bits of subnet ID allow each site considerable flexibility in creating subnets that reflect the site's network structure. For example:

• A smaller organization can just set all the bits in the Subnet ID to zero and have a “flat” internal structure.
  
  
• A medium-sized organization could use all the bits in the Subnet ID to perform the equivalent of “straight” subnetting under IPv4, assigning a different Subnet ID to each subnet. There are 16 bits here, so this allows a whopping 65,536 subnets!
  
  
• A larger organization can use the bits to create a multiple-level hierarchy of subnets, exactly like IPv4's Variable Length Subnet Masking (VLSM). For example, the company could use two bits to create four subnets. It could then take the next three bits to create eight sub-subnets in some or all of the four subnets. There would still be 11 more bits to create sub-sub-subnets and so forth.

IPv6 Address Space Allocation 1 2 3 4 5 IPv6 Interface Identifiers and Physical Address Mapping