Chapter 1. Why IPv6?

The IP version currently used in networks and the Internet is IP version 4 (IPv4). IPv4 was developed in the early ’70s to facilitate communication and information sharing between government researchers and academics in the United States. At the time, the system was closed with a limited number of access points, and consequently the developers didn’t envision requirements such as security or quality of service. To its credit, IPv4 has survived for over 30 years and has been an integral part of the Internet revolution. But even the most cleverly designed systems age and eventually become obsolete. This is certainly the case for IPv4. Today’s networking requirements extend far beyond support for web pages and email. Explosive growth in network device diversity and mobile communications, along with global adoption of networking technologies, new services, and social networks, are overwhelming IPv4 and have driven the development of a next-generation Internet Protocol.

IPv6 has been developed based on the rich experience we have from developing and using IPv4. Proven and established mechanisms have been retained, known limitations have been discarded, and scalability and flexibility have been extended. IPv6 is a protocol designed to handle the growth rate of the Internet and to cope with the demanding requirements on services, mobility, and end-to-end security.

When the Internet was switched from using Network Control Protocol (NCP) to Internet Protocol (IP) in one day in 1983, IP was not the mature protocol that we know today. Many of the well-known and commonly used extensions were developed in subsequent years to meet the growing requirements of the Internet. In comparison, hardware vendors and operating system providers have been supporting IPv6 since 1995 when it became a Draft Standard. In the decade since then, those implementations have matured, and IPv6 support has spread beyond the basic network infrastructure and will continue to be extended.

It is very important for organizations to pay attention to the introduction of IPv6 as early as possible because its use is inevitable in the long term. If IPv6 is included in strategic planning; if organizations think about possible integration scenarios ahead of time; and if its introduction is considered when investing in IT capital expenditures, organizations can save considerable cost and can enable IPv6 more efficiently when it is needed.

An interesting and humorous overview of the history of the Internet can be found in RFC 2235, “Hobbes’ Internet Timeline.” The account starts in 1957 with the launch of Sputnik in Russia and the formation of the Advanced Research Projects Agency (ARPA) by the Department of Defense (DoD) in the United States. The RFC contains a list of yearly growth rate of hosts, networks, and domain registrations in the Internet.

Some excerpts from the RFC:

  • 1969: Steve Crocker makes the first Request for Comment (RFC 1): “Host Software.”
  • 1970: ARPANET hosts start using Network Control Protocol (NCP).
  • 1971: 23 hosts connect with ARPANET (UCLA, SRI, UCSB, University of Utah, BBN, MIT, RAND, SDC, Harvard, Lincoln Lab, Stanford, UIU©, CWRU, CMU, NASA/Ames).
  • 1972: InterNetworking Working Group (INWG) is created with Vinton Cerf as Chairman to address the need for establishing agreed-upon protocols. Telnet specification (RFC 318) is published.
  • 1973: First international connections to the ARPANET are made at the University College of London (England) and Royal Radar Establishment (Norway). Bob Metcalfe’s Harvard PhD thesis outlines the idea for Ethernet. File transfer specification (RFC 454) is published.
  • 1976: Queen Elizabeth II sends an email.
  • 1981: Minitel (Teletel) is deployed across France by France Telecom.
  • 1983: The cutover from NCP to TCP/IP happens on January 1.
  • 1984: The number of hosts breaks 1,000.
  • 1987: An email link is established between Germany and China using CSNET protocols, with the first message from China sent on September 20. The thousandth RFC is published. The number of hosts breaks 10,000.
  • 1988: An Internet worm burrows through the Net, affecting 10 percent of the 60,000 hosts on the Internet.
  • 1989: The number of hosts breaks 100,000. Clifford Stoll writes Cuckoo’s Egg, which tells the real-life tale of a German cracker group that infiltrated numerous U.S. facilities.
  • 1991: The World Wide Web (WWW) is developed by Tim Berners-Lee and released by CERN.
  • 1992: The number of hosts breaks 1,000,000. The World Bank comes online.
  • 1993: The White House comes online during President Bill Clinton’s time in office. Worms of a new kind find their way around the Net—WWW Worms (W4) are joined by Spiders, Wanderers, Crawlers, and Snakes.
  • 1994: Internet shopping is introduced; the first spam mail is sent; Pizza Hut comes online.
  • 1995: The Vatican comes online. Registration of domain names is no longer free.
  • 1996: 9,272 organizations find themselves unlisted after the InterNIC drops their name service as a result of their not having paid their domain name fees.
  • 1997: The 2,000th RFC is published.

This is how far the RFC goes. But history goes on. According to, the worldwide online population reached 361 million users in 2000 (a penetration rate of 5.8%) and 587 million users in 2002. In 2003, the U.S. Department of Defense announced that they would be migrating the DoD network to IPv6 by 2008, and the Moonv6 project was started (now concluded). In 2005, Google registered a /32 IPv6 prefix, and Vint Cerf, known as “Father of the Internet,” joined Google. By that time the number of Internet users had reached 1.08 billion. Today, at the time of writing in 2014, we are at approximately 2.4 billion Internet users, which corresponds to a penetration rate of 34%.

So while these numbers reflect all Internet users, independent of the IP protocol version, now we are starting to watch the growth of the IPv6 Internet. It is in its early days, but according to the growth numbers of the last two years, we expect growth to be exponential, and probably much faster than even the enthusiasts among us expect. The growth of the IPv6 Internet can be seen on the Google IPv6 Adoption statistics and the stats as of spring 2014 are shown in Figure 1-1.

The stats show that in early 2011 (when the IANA IPv4 pool ran out), the percentage of native IPv6 Internet users was at approximately 0.2%. The stats also show that the percentage of users that were not native IPv6 (e.g., 6to4 or Teredo, red line) dropped to almost zero and are since then insignificant. Within one year the number of IPv6 Internet users doubled to 0.4%—a small number but still growth. In January 2013, the IPv6 Internet had crossed the 1% mark, and we entered 2014 with almost 3% IPv6 Internet users, which corresponds to approximately 72 million users. At the time of delivering this chapter, in April 2014, we were at 3.5%. The number of IPv6 Internet users currently doubles approximately every nine months.

These are just a few selected events and milestones of the Internet’s history. Keep watching as more history unfolds. We are all creating it together.

Google’s global IPv6 adoption statistics as of spring 2014
Figure 1-1. Google’s global IPv6 adoption statistics as of spring 2014

The History of IPv6

The Internet Engineering Task Force (IETF) began the effort to develop a successor protocol to IPv4 in the early 1990s. Several parallel efforts to solve the foreseen address space limitation and to provide additional functionality began simultaneously. The IETF started the Internet Protocol Next Generation (or IPng) area in 1993 to investigate the different proposals and to make recommendations for further procedures.

The IPng area directors of the IETF recommended the creation of IPv6 at the Toronto IETF meeting in 1994. Their recommendation is specified in RFC 1752, “The Recommendation for the IP Next Generation Protocol.” The Directors formed an Address Lifetime Expectation (ALE) working group to determine whether the expected lifetime for IPv4 would allow the development of a protocol with new functionality, or if the remaining time would allow only the development of an address space solution. In 1994, the ALE working group projected that the IPv4 address exhaustion would occur sometime between 2005 and 2011 based on the available statistics.

For those of you who are interested in the different proposals, here’s some more information about the process (from RFC 1752). There were four main proposals: CNAT, IP Encaps, Nimrod, and Simple CLNP. Three more proposals followed: the P Internet Protocol (PIP), the Simple Internet Protocol (SIP), and TP/IX. After the March 1992 San Diego IETF meeting, Simple CLNP evolved into TCP and UDP with Bigger Addresses (TUBA), and IP Encaps became IP Address Encapsulation (IPAE). IPAE merged with PIP and SIP and called itself Simple Internet Protocol Plus (SIPP). The TP/IX working group changed its name to Common Architecture for the Internet (CATNIP). The main proposals were now CATNIP, TUBA, and SIPP. For a short discussion of the proposals, refer to RFC 1752.


CATNIP is specified in RFC 1707; TUBA in RFCs 1347, 1526, and 1561; and SIPP in RFC 1710.

The Internet Engineering Steering Group approved the IPv6 recommendation and drafted a Proposed Standard on November 17, 1994. RFC 1883, “Internet Protocol, Version 6 (IPv6) Specification,” was published in 1995. The core set of IPv6 protocols became an IETF Draft Standard on August 10, 1998. This included RFC 2460, which obsoleted RFC 1883.


Why isn’t the new protocol called IPv5? The version number 5 could not be used, because it had been allocated to the experimental stream protocol.

One of the big challenges but also one of the main opportunities of IPv6 is the fact that we can redesign our networks for the future. This is what enterprises should focus on most when planning their IPv6 integration in order to make sure they don’t just copy old concepts onto a new protocol. We have to rethink our architectures. This once-in-a-lifetime opportunity can be used to get rid of a lot of legacy. An interesting RFC that helps in the process of seeing the big picture is RFC 6250, “Evolution of the IP Model.” It shows how much this model has changed in the many years of operating our networks. So it helps to free our minds for thinking in new ways. One funny little quote that demonstrates what I am talking about is included below.

In this RFC there is mention of the first IP model and addressing architecture, and it quotes RFC 791, which defined IPv4 and the IPv4 address:

Addresses are fixed length of four octets (32 bits). An address begins with a one-octet network number, followed by a three-octet local address. This three-octet field is called the “rest” field.

This is how far we have come. Now project this into the future with the vast IPv6 address space in mind. Making meaningful use of the new address architecture and the enormous space will write the next chapter of the evolution of the IP model.


Here’s a quote from Vint Cerf:

The vast IPv6 address space opens up serious opportunities for the re-examination of the notion of address. The IETF has only allocated 1/8th of the IPv6 address space for current use. The remaining 7/8ths of the address space is still to be allocated. In consequence we may be able to interpret new segments of the IP address space in ways that are different from topological end points. This is precisely the reason that a focus on the future of IPv6 is so important at this point in the evolution of the Internet.

What’s New in IPv6?

IPv6 is an evolution of IPv4. The protocol is installed as a software upgrade in most devices and operating systems. If you buy up-to-date hardware and operating systems, IPv6 is usually supported and needs only activation or configuration. In many cases it is activated by default. Currently available transition mechanisms allow the step-by-step introduction of IPv6 without putting the current IPv4 infrastructure at risk.

Here is an overview of the main changes:

Extended address space
The address format is extended from 32 bits to 128 bits. This provides multiple IP addresses for every grain of sand on the planet. In addition, it also allows for hierarchical structuring of the address space in favor of optimized global routing.
Perhaps the most intriguing new feature of IPv6 is its Stateless Address Autoconfiguration (SLAAC) mechanism. When a booting device in the IPv6 world comes up and asks for its network prefix, it can get one or more network prefixes from an IPv6 router on its link. Using this prefix information, it can autoconfigure for one or more valid global IP addresses by using either its MAC identifier or a private random number to build a unique IP address. In the IPv4 world, we have to assign a unique IP address to every device, either by manual configuration or by using DHCP. SLAAC should make the lives of network managers easier and save substantial cost in maintaining IP networks. Furthermore, if we imagine the number of devices we may have in our homes that will need an IP address in the future, this feature becomes indispensable. Imagine reconfiguring your DHCP server at home when you buy a new television! Stateless Address Autoconfiguration also allows for easy connection of mobile devices, such as a smartphone, when moving to foreign networks.
Simplification of header format
The IPv6 header is much simpler than the IPv4 header and has a fixed length of 40 bytes. This allows for faster processing. It basically accommodates two times 16 bytes for the Source and Destination address and only 8 bytes for general header information.
Improved support for options and extensions
IPv4 integrates options in the base header, whereas IPv6 carries options in so-called Extension headers, which are inserted only if they are needed. Again, this allows for faster processing of packets. The base specification describes a set of six Extension headers, including headers for routing, Quality of Service, and security.

Why Do We Need IPv6?

For historic reasons, organizations and government agencies in the United States used the largest part of the allocatable IPv4 address space. The rest of the world had to share what was left over. Some organizations used to have more IPv4 address space than the whole of Asia (where more than 50% of the world’s population live). This is one explanation of why the deployment of IPv6 in Asia is much more common than in Europe and the United States.


An interesting resource site for statistics can be found at the following link:

The IPv4 address space has a theoretical limit of 4.3 billion addresses. However, early distribution methods allocated addresses inefficiently. Consequently, some organizations obtained address blocks much larger than they needed, and addresses that could be used elsewhere are now unavailable. If it were possible to reallocate the IPv4 address space, it could be used much more effectively, but this process is not possible, and a global reallocation and renumbering is simply not practical. In addition to that it would not buy much, as even 4.3 billion addresses would not suffice for long at the current growth rate. We have to take into account that in the future we will need IP addresses for billions of devices. Vendors in all industries are developing monitoring, control, and management systems based on IP.

As the previous section shows, the IPv6 working group has done more than just extend the address space. For many complex networks of today and tomorrow, and for the number of IP devices of all types, the Autoconfiguration capability of IPv6 will be a necessity. The management of such services can’t be accomplished with traditional addressing methods, and Stateless Address Autoconfiguration will also help to reduce administrative costs for organizations.

The extended address space and the restoration of the original end-to-end model of the Internet allows for the elimination of Network Address Translation (NAT), in which a single or a few public IPv4 address(es) are used to connect a high number of users with private addresses to the Internet by mapping the internal addresses to the public address(es). NATs were introduced as a short-term fix for solving the address space limitations with IPv4, since IPv6 was not ready yet (refer to RFC 1631; the original NAT specification was obsoleted by RFC 3022 in 2001). NATs have become pretty common in IPv4 networks, but they create serious disadvantages in management and operation: in order to do the address mapping, NATs modify end node addresses in the IP header. Very often, Application Level Gateways (ALG) are used in conjunction with NAT to provide application-level transparency. There is a long list of protocols and applications that create problems when used in a NAT environment. IPsec and peer-to-peer applications are two well-known examples. Another known issue with NAT is the overlapping of private address space when merging networks, which requires either the renumbering of one of the networks or the creation of a complex address-mapping scheme. The amplification of limited address space, the primary benefit of NAT, is not needed with IPv6 and therefore is not supported by design.

By introducing a more flexible header structure (Extension headers), the protocol has been designed to be open and extensible. In the future, new extensions can easily be defined and integrated in the protocol set. Based on the fact that IPv4 has been in use for almost 30 years, the development of IPv6 was based on the experience with IPv4 and focused on creating an extensible foundation; you can expect it to last a long time.

Broadband penetration rates in many countries continue to accelerate and, in some cases, have reached 65% or more. This level of always-on connectivity with substantial bandwidth capacity means that there is greater opportunity for devices to be connected. And many consumer electronic manufacturers have taken advantage of this. Online gaming is no longer the sole purview of games on PCs. Gaming stations, such as Sony’s PlayStation 4, Xbox One, or Nintendo Wii U, have added capabilities to take them online. Many telecommunication carriers are providing television-type services (movies, audio content, etc.) over their IP networks. Even appliances, such as refrigerators, stoves, water heaters, and bathtubs, are getting connected. While it may seem rather silly to network-enable a bathtub, many of these devices are being connected to facilitate things such as power management, remote control, and troubleshooting, and for telemetry/monitoring purposes. We are entering the age of smart buildings and smart cities. The end result of this network-enablement process is a greater number of devices that need addressing, many of which will not have standard user interfaces. In these cases, the IPv6 address space, coupled with features such as Neighbor Discovery, Stateless Autoconfiguration, and Mobile IPv6, will help to usher in a new era of computerization in the home, but hopefully without the enormous deployment headache that it would cause if it were attempted with the current protocol.

The growth of the wireless industry (both cellular and wireless networks) has been nothing short of phenomenal. In more and more countries the number of cell phones actually exceeds the number of people. In this world of continuous reachability and reliance on the ability to access information at any time, the mobility requirements for end users have become exceptionally important. From the carriers’ perspective, especially those supporting multiple media access types (e.g., 3G, WiMax, LTE), leveraging IP as the method of transporting and routing packets makes sense. Smartphones access the Internet, play games with other users, make phone calls, and even stream video content. Instead of supporting all of these functions using different transport protocols and creating intermediary applications to facilitate communications, it is far more efficient to leverage the existing network infrastructure of the Internet and a company’s network. We will see later that from a technical perspective, Mobile IPv6 is very elegant in its design, supporting mobile users in a highly efficient manner and providing the overlay mechanisms for users to maintain their connections when moving between networks, even if those networks do not use the same type of media access.

There still remain some questions about the value of IPv6 to the enterprise, and it is worth conceding that each organization needs to evaluate the benefits and best timing of IPv6 for their own internal use. In many instances, organizations can find clever ways to use IPv6 to solve “pain” issues without migrating their entire network. Adoption can occur in an incremental fashion with a plan that minimizes integration pain but also ensures that everything is ready when the time comes to “flip the switch.” As many case studies show, well-planned introduction costs substantially less than you would expect; the main cost-saving aspect is the fact that the advance planning lets you use all your refresh cycles, which minimizes cost. The step-by-step introduction allows you to learn as you go, thereby saving a lot of money and headaches, and you can do it without putting the current IPv4 infrastructure at risk.

But with all these thoughts and considerations, let’s not forget the most essential advantage of IPv6. With its new structure and extensions, IPv6 provides the foundation for a new generation of services. There will be devices and services on the market in the near future that cannot be developed with IPv4. This opens up new markets and business opportunities for vendors and service providers alike. The first-mover opportunities are substantial, as are the opportunities to extend current product life cycles by refreshing their technology with IPv6. On the other hand, it means that organizations and users will require such services in the mid-term. It is therefore advisable to integrate the new protocol carefully and in a nondisruptive manner, by taking one step at a time to prepare the infrastructure for these new services. This protects you from having to introduce a business-critical application based on IPv6 at unreasonably high cost with no time for thorough planning.

Common Misconceptions

When considering all these advantages, maybe the question should be: “Why not IPv6?” When talking to customers, we often find that they share a similar set of misconceptions preventing them from considering IPv6. Here are the most common ones:

“The introduction of IPv6 puts our current IP infrastructure—our networks and services—at risk.”
This concern is unsubstantiated. A major focus in IPv6’s development was to create integration mechanisms that allow both protocols to coexist peacefully. You can use IPv6 both in tandem with and independently of IPv4. It is possible to introduce IPv6 and use it for access to new services while retaining IPv4 to access legacy services. This not only ensures undisrupted access to IPv4 services, but it also allows a step-by-step introduction of IPv6. I discuss these mechanisms in Chapter 7. Your biggest risk is to not take advantage of all the opportunities IPv6 offers. You can only use these opportunities if you plan while there is time.
“The IPv6 protocol is immature and hasn’t proven that it stands the test of time or whether it is capable of handling the requirements.”
This was a concern of many people back in 2006 when we published the second edition of this book. Now in 2014 this is not true anymore. Many ISPs and organizations are deploying IPv6, vendors are getting up to speed, and the working groups have developed and optimized mechanisms that help with the integration. There is no technical reason not to do IPv6.
“The costs of introducing IPv6 are too high.”

There will certainly be costs associated with adopting IPv6. In many cases, newer networks will find that the level of IPv6 support in their current infrastructure is actually high. Regardless, the transition will necessitate some hardware and software costs. Organizations will need to create new designs, review current concepts, train their IT staff, and may need to seek outside expertise in order to take full advantage of IPv6.

However, the cost savings associated with IPv6 are becoming easier to define. Networks based on IPv4 are becoming increasingly more complex. New IT services such as VoIP, instant messaging, video teleconferencing, IPTV, and unified communications are adding layers of middleware and complexity. Merging organizations or those conducting B2B transactions are implementing NAT overlap solutions that have high management costs and are difficult to troubleshoot. And a growing market of mobile devices and network appliances requires robust access models that are expensive and difficult to implement in an IPv4 world. In all of these cases, IPv6 presents a cleaner and more cost-effective model in the long run than IPv4 can provide. And the fact is that an investment in IPv4 is an investment in an end-of-life technology, while an investment in IPv6 is an investment in the future technology.

“With Stateless Address Autoconfiguration, we will not be able to control or monitor network access.”
While this statement may generally be true for networks that widely utilize Stateless Address Autoconfiguration, administrators will have a choice about their level of control. DHCPv6 as defined in RFC 3315 has been extended to support two general modes of operation, Stateful and Stateless. Stateful mode is what those who currently utilize DHCP (for IPv4) are familiar with, in which a node (DHCP client) requests an IP address and configuration options dynamically from a DHCP server. DHCPv6 also offers a Stateless mode in which DHCPv6 clients simply request configuration options from a DHCPv6 server and use other means, such as Stateless Address Autoconfiguration, to obtain an IPv6 address.
“Our Internet Service Provider (ISP) does not offer IPv6 services, so we can’t use it.”
You do not have to wait for your ISP to use IPv6 in your corporate or private network. If you want to connect to the global IPv6 Internet, you can use one of the transition mechanisms and tunnel your IPv6 packets over the IPv4 infrastructure of your ISP. This may be doable for smaller organizations. On the other hand, at the time of writing in 2014, you could expect a large ISP targeting enterprise customers to support IPv6. And this should be your standard requirement in any renewal of contract and SLAs (Service Level Agreements). If your ISP does not provide IPv6 services, consider finding a new provider.
“It would be too expensive and complex to upgrade our backbone.”
The transition mechanisms make it possible to use IPv6 where appropriate without dictating an order of upgrade. Usually for the backbone it is advisable to wait for the regular life cycle, when hardware needs to be exchanged anyway. Make sure to choose hardware that supports performance IPv6 routing. In the meantime, you can tunnel your IPv6 packets over the IPv4 backbone. Networks that use MPLS have an easy way to tunnel IPv6 packets over their IPv4 MPLS backbone. Read more about it in Chapter 7. More and more organizations are considering migrating their backbone and data centers to IPv6 only with the next refresh or redesign cycle, because it substantially reduces operational cost. In this scenario we will start to tunnel IPv4 packets over IPv6 backbones. IPv4 as a service is the new keyword.
“It would be too complex and expensive to port all of our applications to IPv6.”
The effort necessary to port applications to run over IPv6 is often much lower than expected. If an application is well written, it may simply run over IPv6 without modification. Instead of assuming that it won’t work, test it to find out. For applications that need modifications that are not yet available, or for applications in which porting does not make sense, there are mechanisms available that support IPv4 applications in IPv6 networks and IPv6 applications in IPv4 networks. Alternatively, you can run a dual-stack network, in which you use IPv4 to access IPv4 applications and IPv6 to access IPv6 applications. In any case it is recommendable for enterprise customers to start the planning process early and provide good labs for the application teams to test their applications before there is time pressure.
“We have enough IPv4 addresses; we don’t need IPv6.”
True—if you have enough IPv4 addresses, there may be no immediate need to integrate IPv6 today. But ignoring IPv6 for this reason is a perspective that assumes that your network stands completely isolated from the rest of the world, including your vendors, partners, and customers. IPv6 adoption is further along in Asia and Europe than in the United States, so even though you may have adequate address space for your operations in Denver, interconnecting with a partner organization in Tokyo may eventually become complicated if you do not support IPv6. Plus, the assumption that IPv6 is about address space only doesn’t account for the advanced features that IPv6 brings to the table.

When Is It Time for IPv6?

The answer in 2014 is now! If the rest of the world moves to IPv6 while you insist on continuing to use IPv4, you will exclude yourself from global communication and reachability. The risks if you wait too long include losing potential customers and access to new markets and the inability to use new IPv6-based business applications.

There is a golden rule in IT: “Never touch a running system.” As long as your IPv4 infrastructure runs well and fulfills your needs, there is no reason to change anything. But from now on, whenever you invest in your infrastructure, you should consider IPv6. An investment in the new technology gives it a much longer lifetime and keeps your network state-of-the-art.

These are the main indicators that it may be time for you to consider switching to or integrating IPv6:

  • You need to extend or fix your IPv4 network or NAT implementation.
  • You are running out of address space.
  • You want to prepare your network for applications that are based on advanced features of IPv6.
  • You need end-to-end security for a large number of users and you do not have the address space, or you struggle with a NAT implementation.
  • You need to replace your hardware or applications that are at the end of their life cycles. Make sure you buy products that support IPv6 adequately, even if you don’t enable it right away.
  • You want to introduce IPv6 while there is no time pressure.

The following provisions can be taken in order to prepare for IPv6 adequately:

  • Build internal knowledge, educate IT staff, and create a test network.
  • Include IPv6 in your IT strategy.
  • Design future-proof network, security, and service concepts while you have time.
  • Create integration scenarios based on your network and requirements.
  • Put IPv6 support on all of your hardware and software purchasing guidelines. Be specific about which features (RFCs) must be supported. Don’t forget to add IPv6 requirements to outsourcing and service contracts, as well as SLAs.
  • Compel your vendors to add IPv6 support to their products.

If you do this, you can determine the right moment for the introduction of IPv6 in your network. You can also assess whether a further investment in your IPv4 infrastructure makes sense or whether introducing IPv6 would be a better way to go.

There will be no “flag day” for IPv6 like there was for the 1983 move from NCP to IPv4. Probably there will be no killer application either, so don’t wait for one. Or as some people like to say, the killer application for IPv6 is the Internet. IPv6 will slowly and gradually grow into our networks and the Internet. Taking a step-by-step approach to IPv6 may be the most cost-efficient way to integrate it, depending on your requirements. This method does not put your current infrastructure at risk or force you to exchange hardware or software before you are ready, and it allows you to become familiar with the protocol, to experiment, to learn, and to integrate what you’ve learned into your strategy.


You may want to enable IPv6 in your public services first. Due to the lack of IPv4 addresses, ISPs that want to grow their customer base (and who does not want to do that?) make use of NAT-type mechanisms to extend their IPv4 address space. This includes CGN (Carrier Grade NAT), which means multiple customers share one single public IPv4 address and sit behind multiple layers of NAT.

These users may have a bad user experience accessing your IPv4 website, and for e-commerce or other more complex services it may even fail. The users will not know that it is the provider’s CGN causing the issue and will blame your website for their problems. If you provide your website dual-stack, these users can access it over IPv6 and bypass the IPv4 NATs.


Find more information on CGN in the “NAT to extend IPv4 address space” section in Chapter 7.

IPv6 Status and Vendor Support

As previously mentioned, IPv6 is implemented in most up-to-date versions of routing and operating systems. For standard applications, assume that IPv6 support has already been added or will be added with their next major release at the latest. For creating an IPv6 integration plan for your corporate network, you will need to assess the status and degree of IPv6 support with each vendor individually. Many vendors have an information site that can often be found at http://www.<vendor>.com/ipv6.

It can be said that IPv6 support up to the network layer is mature, tested, and optimized. This includes routing, transition mechanisms, DNS, and DHCPv6.

Development is most active in the security, transition mechanism, IPv4/IPv6 MIB integration, and Mobile IPv6 areas. More work needs to be done in the areas of network management and firewalls. Vendors such as Cisco, Checkpoint, Juniper, and many others are working on these areas. The application area is continuously developing, and new applications will appear on the market that will make use of the advanced features of IPv6. Thanks to the transition mechanisms, you can still use IPv4 applications in IPv6 networks.


Find more information on the planning process in Chapter 9.

Now you know why you should care about IPv6. The following chapters in this book aim to make learning about IPv6 a joy. So please read on.


Here’s a list of the most important RFCs mentioned in this chapter. Sometimes I include additional subject-related RFCs for your further personal study.


  • RFC 1, “Host Software,” 1969
  • RFC 791, “Internet Protocol,” 1981
  • RFC 1347, “TCP and UDP with Bigger Addresses (TUBA),” 1992
  • RFC 1526, “Assignment of System Identifiers for TUBA/CLNP Hosts,” 1993
  • RFC 1561, “Use of ISO CLNP in TUBA Environments,"1993
  • RFC 1631, “The IP Network Address Translator (NAT),” 1994
  • RFC 1707, “CATNIP: Common Architecture for the Internet,” 1994
  • RFC 1710, “Simple Internet Protocol Plus White Paper,” 1994
  • RFC 1752, “The Recommendation for the IP Next Generation Protocol,” 1995
  • RFC 1883, “Internet Protocol, Version 6 (IPv6) Specification,” 1995
  • RFC 2235, “Hobbes’ Internet Timeline,” 1997
  • RFC 2324, “Hyper Text Coffee Pot Control Protocol (HTCPCP/1.0),” April 1, 1998
  • RFC 2460, “Internet Protocol, Version 6 (IPv6) Specification,” 1998
  • RFC 2555, “30 Years of RFCs,” 1999
  • RFC 2663, “IP Network Address Translator (NAT) Terminology and Considerations,” 1999
  • RFC 3022, “Traditional IP Network Address Translator (Traditional NAT),” 2001
  • RFC 3027, “Protocol Complications with the IP Network Address Translator,” 2001
  • RFC 4677, “The Tao of IETF: A Novice’s Guide to the Internet Engineering Task Force,” 2006
  • RFC 5902, “IAB Thoughts on IPv6 Network Address Translation,” 2010
  • RFC 6250, “Evolution of the IP Model,” 2011
  • RFC 6269, “Issues with IP address sharing,” 2011
  • RFC 6921, “Design Considerations for Faster-Than-Light (FTL) Communication,” April 1, 2013
  • RFC 7168, “The Hyper Text Coffee Pot Control Protocol for Tea Efflux Appliances (HTCPCP-TEA),” April 1, 2014

Get IPv6 Essentials, 3rd Edition now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.