IPv6

Abstract

Due to the overwhelming global response to the release of the internet there has been a diminished capacity of available network addresses. Although the internet community was aware of this as early as the late 1980s it was not until the mid 1990s that they developed a solution known as IPv6 or Internet Protocol version 6. This updated version boasts many advantages but the transition process has been slow and there are many obstacles to full deployment. However, IPv6 must be deployed soon since there will be no address space left within IPv4. As access to the internet increases and a greater number of networks are created there is a need to create more address space on the web. Forecasters have predicted there will be a total depletion of IPv4 addresses sometime between 2010 and 2012 (Bradner, n.d.). However, there is relief in sight for the address exhausted IPv4 system, Internet Protocol version 6. This newer version has vastly expanded address space and new features to simplify the address assignment process. Although it was designed in the 1990s this system is still considered “new” and in 2008 was used less than 1% in internet enabled hosts in any country (IETF Tools, n.d.). In the late 1980s it became evident that there was a crunch on the number of addresses the IPv4 could assign. No one could have predicted that the internet would have gained the widespread popularity and accessibility that it has over the past three decades. However, in the 1990s IPv6 was introduced and in 1994 adopted by the Internet Engineering Task Force. At this time it was referred to as the IPng or Internet Protocol Next Generation. It was hoped that IPv6 would catch on quickly and spread across the globe but progress has been slow; especially with the introduction of network address translation (Grossetete, 2006, 18). NAT has partially alleviated the problem of address exhaustion but also makes it difficult to use some peer-to-peer applications such as Voice over Internet Protocol and multi-user games. With the full adoption and implementation of IPv6 internet users will find many improved internet services. The IPv6 packet is composed of three main parts: the fixed header, optional extension headers, and the payload. The fixed header makes up the first 320 bits of an IPv6 data packet and contains the source and destination address as well as the traffic classification options, a hop counter, and an indication of the next header. The next header field can either lead to a chain of zero or more extension headers. Extension headers carry the directions for a packet along the way or at its destination, including special instruction for routing, fragmenting, and security. The last next header field points to the upper-layer protocol that is carried in the packet’s payload. A payload can have a size of up to 64 KB or larger with a jumbo payload option is chosen in the Hop-by-Hop extension header. Internet service around the globe will improve from the implementation of the IPv6 system. Most notably, cell phones and mobile devices will each receive a unique and permanent address. Currently, the IPv4 system has the capability to support just over 4 billion addresses. Unfortunately, even with a number this large, it is not enough to assign every person in the world with an address, much less support all of the phones, PDAs, cars, etc. that request internet access. Luckily, the IPv6 will support 340 undecillion addresses (History of the IPng Effort, n.d.). However, it was not the intent to provide everyone with an address its just one of the many advantages of the IPv6 system. The most important feature of the Ipv6 is the capability to have an address 128 bits long. While this does support an increase in the number of addresses that can be assigned it also changes the length of the address. A longer address will allow for a better, systematic, hierarchical allocation of addresses and efficient route aggregation. IPv4 renumbering an existing network was a major effort but with IPv6 the process is simplified. By changing the prefix announced by a few routers an entire network can be renumbered. This is due to the fact that the host identifiers can be independently self-configured by a host. When an IPv6 host is connected to a routed IPv6 network it will automatically configure itself using the Internet Control Message Protocol. The ICMP is an integral part of IPv6 that performs error reporting, neighbor discovery, and diagnostic functions. Using the ICMP a host, when first connected to a network, will send a link-local multi-cast router solicitation request for its configuration (Davies, 2008, 58). If the configuration is appropriate router will respond with a router advertisement packet. This packet will contain network-layer configuration parameters. If the configuration is unsuitable hosts can be configured in one of two ways, through stateful configuration with the Dynamic Host Configuration Protocol or statically. Another advantage in IPv6 is the ability to send a single packet to multiple destinations, or multicast. This is in contrast to IPv4 which uses a broadcast system to send a packet to all hosts on the attached link. Instead, the IPv6 sends a packet to the link-local all hosts multicast group. This revision is because the broadcast address for a specific subnet in Ipv4 is considered a normal address in IPv6. Unfortunately, at this time most environments do not have their network infrastructures configured to route multicast packets. However, you can still multicast on a single subnet but multicasting globally will not work. Another advantage to the IPv6 multicast is that even the smallest IPv6 global routing prefix assigned to an organization also allows the organization the use of globally routable source specific IPv6 multicast groups to assign for inner-domain or cross-domain multicast applications. IPv6 also includes Embedded Rendezvous Point which simplifies the deployment of cross domain solutions. Simplicity is one of the main advantages to using IPv6. The packet headers and process of packet forwarding has been made more efficient by simplifying the packet processing by routers. First, many of the rarely used IPv4 fields have been moved to separate options creating IPv6 addresses that are four times larger but only have a header that is twice the size of the IPv4. Also, the IPv6 header is not protected by a checksum. This is due to the fact that a double check is assumed to be performed at the link layer and a higher layer. Therefore, routers do not need to recompute a checksum when header fields change. Finally, routers are no longer required to compute the time that a packet has spent in a queue or perform fragmentation. Lastly, another advantage of the IPv6 network is its mobility. Mobile IPv4 uses triangular routing but Mobile IPv6 avoids this and therefore is just as efficient as the normal IPv6. If an IPv6 router supports Network Mobility or NEMO then entire subnets will be allowed to move to a new router connection point without renumbering. Despite its many advantages a complete deployment of IPv6 is not possible. In order to make the change from IPv4 to IPv6 as smooth as possible a transition system will have to be put into place meaning that IPv4 and IPv6 will have to coexist until all systems are able to support IPv6. One of the fundamental transition technologies involves the presence of two Internet Protocol software implementations in an operating system, one for IPv4 and the other for IPv6. This is known as a dual stack. These dual stack IP hosts have the availability to run IPv4 and IPv6 independently of one another or through hybrid implementation. This form is more commonly implemented in modern operating systems on server and end-user computers. Hybrid dual stack implementations allow programmers to write networking code that works transparently on IPv4 or IPv6. The software allows for IPv4 or IPv6 packets to be accepted but the address is represented in a special IPv6 address which is actually an IPv4 mapped address (Grossetete, 2006, 126). Unfortunately, due to the significant internal differences between IPv4 and IPv6 some of the lower level functionality in IPv6 is not available to programmers. There are also common IPv6 stacks that do not support the IPv4 mapped address feature; the two stacks are different implementations such as XP and Server 2003 or because of security concerns. Tunneling is the technique used to get IPv6 packets to the IPv6 Internet through the IPv4 infrastructure. Generally, the IPv6 datagrams are encapsulated in IPv4 packets indicated with a specific protocol number. However, the IPv6 can also be encapsulated within the UDP packets or through other ways as used in AYIYA or GRE (Davies, 2008, 178). In configured tunneling the tunnel endpoints are configured explicitly either by an administrator or by the operating system’s configuration mechanisms. Closely related to this is automated tunneling when the tunnel endpoints are configured by a tunnel broker or automatic service. However, in large, well-administered networks it is recommended that configured tunneling is used because it is more deterministic and easier to debug. The advantage to automated tunneling is the compromise between the ease of use and the deterministic behavior of configured tunneling. Other barriers to a complete deployment of IPv6 include problems with hardware and equipment as well as consumer buy-in. Existing hardware and equipment poses a problem for several reasons. First, some of the companies that originally manufactured the hardware no longer exist or the manufacturer refuses to support the upgrade because it is too expensive. Second, some of the equipment is so dated that upgrades are impossible or the device has insufficient resources to implement the IPv6 stack. Finally, in some devices IPv6 is supported but it has poor performance results. Consumer buy-in is also a block to full deployment of IPv6. Often times the consumer is not looking ahead at what they will need, they are looking to the most cost and time efficient solution. Plus, at the moment the consumer doesn’t recognize or see any of the existing problems with address exhaustion. Publicity and education are expensive therefore many people are unaware that an address exhaustion problem even exists. Small consumers are the ones most likely to suffer because they often view networking devices in the same manner as small appliances; they rarely need repairs and never have to be configured or updated. IPv6 is definitely not something that drives consumer purchases. But, with only a few years before a universal upgrade to IPv6 there are many devices that will not function. It is mainly a software issue that prevents devices from working. Therefore smart equipment that contains software will need explicit IPv6 support and manufacturers have not shown an interest in updating equipment so that it will support IPv6 when it becomes necessary. Manufacturers, at this time do not want to spend money on development costs for hardware that has already been sold when they can make more money from the “IPv6” ready sales of equipment. Mainly, there is just a hope that somehow IPv4 and IPv6 will be able to coexist and streamline the transition (IETF Tools, n.d.). This leaves the internet community divided on the issue of whether it should be a quick switch or longer transition and whether or not they should require all internet servers to be prepared to serve IPv6 clients only in the near future. That near future is closing in quickly as the demand for internet address space on IPv4 increases and the amount of addresses available diminishes. Even with its many advantages manufacturers and consumers are slow to adopt IPv6. Hopefully, as small consumers are made more aware of the issue it will gain publicity and manufactures and the internet community as a whole will being to make moves and decisions that lead to the complete deployment of IPv6. If the Y2K disaster taught internet users anything, hopefully, it is that the best prevention is early intervention.
References
Bradner, S., & Mankin, A. (n.d.). RFC 1752 - The Recommendation for the IP Next Generation Protocol. IETF Tools. Retrieved June 5, 2010, from http://tools.ietf.org/html/rfc1752
Davies, J. (2008). Understanding IPv6, Second Edition (2 ed.). Washington D.C.: Microsoft Press.
Grossetete, P., Levy-Abegnoli, E., & Popoviciu, C. (2006). Deploying IPv6 Networks (Networking Technology) (1 ed.). Indianapolis: Cisco Press.
Hagen, S. (2006). IPv6 Essentials (2 ed.). Sebastopol: O'Reilly Media, Inc..
History of the IPng Effort. (n.d.). Playground.Sun.COM. Retrieved June 2, 2010, from http://playground.sun.com/ipv6/doc/history.html
RFC 5214 - Intra-Site Automatic Tunnel Addressing Protocol (ISATAP). (n.d.). IETF Tools. Retrieved June 15, 2010, from http://tools.ietf.org/html/rfc5214