 |
Cybertelecom Federal Internet Law & Policy An Educational Project |
IPv6 Transition
|
IPv6 Transition |
|
|
Internet Addresses - DNS - History - NTIA & Fed Activity - ICANN - Root Servers - ccTLDs - - .us - - -.kids.us - gTLDs - - .gov - - .edu - - .mil - - .xxx - WHOIS - WGIG - ENUM - IP Numbers - - IPv6 - BGP - NATs - Ports - Security - Trademark - AntiCybersquatter Consumer Protection Act - Gripe Sites - Truth in Domain Names Telephone Addresses |
Derived From: GAO, Internet Protocol version 6, Federal Agencies Need to Plan for Transition and Manage Security Risks, p. 21-22 (May 2005)
Transition of the Internet from IPv4 to IPv6 has been compared to the Y2K effort, demanding significant investment of resources and time in order to achieve. [Cook] Heretofore, the transition has been slow and controversial. [ICANN 10-07] Much like Y2K, companies do not necessarily recognize IPv4 number exhaustion as an immediate problem, and therefore are putting off the investment required into the future - with the risk that in the future (just like Y2K), there may be insufficient time for remediation and remediation may be more expensive.
IPv4 and IPv6 are different numbering protocols. IPv6 is not backward compatible with IPv4. This means that during the transition, the network will utilize two fully distinct addressing schemes. [Bush slide 7] It is likely that IPv4 numbering were remain a legacy in the network for a good time into the future. [ICANN 10-07]
Partial list of equipment that will need to be transitioned include
- Routers
- Backbones
- Exchange Points
- Regional ISPs
- Local Networks
- Authentication and session set up
- Servers
- Server farms
- Local Networks
- Authentication and session set up
- DNS
- Root Servers
- Back office equipment
- Home CPE
- DSL Modems
- Cable Modems - DOCSIS
- Fiber Modems
- Mobile hand units (mobile phones)
- Applications
- Telephony
- Firewalls & Security
- Databases
Various transition methods exist to ensure that a computer running IPv6 can communicate with a computer running IPv4. These transition methods or techniques include the following:
In a dual stack network, hosts and routers implement both IPv4 and IPv6. Figure 6 depicts how dual stack networks can support both IPv4 and IPv6 services and applications during the transition period. Currently, dual stack networks are the preferred mechanism for transitioning to IPv6. [The Sprint to IPv6, Internet News 12.19.07 (Sprint will utlize Dual Stack solution)]
Tunneling allows separate IPv6 networks to communicate via an IPv4 network. For example, for one type of tunneling method, IPv6 packets are encapsulated by a border router, sent across an IPv4 network, and decoded by a border router on the receiving IPv6 network. Figure 7 depicts the tunneling process of IPv6 data inside an IPv4 network. [Sprint Paper] B. Carpenter, K. Moore, IETF RFC 3056, Connection of IPv6 Domains via IPv4 Clouds (Feb. 2001) (tunnels)
Translation allows networks using only IPv4 and networks using only IPv6 to communicate with each other by translating IPv6 packets to IPv4 packets. The use of a translator allows new systems to be deployed as IPv6 only, while older systems remain IPv4 only. While this method may result in bottlenecks while packets are being translated, it can provide a high level of interoperability.
These transition methods represent a few of the common approaches for ensuring interoperability between IPv6 and IPv4 communications. They can be used alone or in concert to enable communication among IPv4 and IPv6 networks. However, while such techniques mitigate interoperability challenges, in some instances, they may result in increased security risks if not analyzed and managed.
NATs: This project would be a lot further along if the Washington Nationals were a part of the solution; however, until the Washington Nats get their new stadium built, this solution refers to Network Address Translation devices. Basically this is a black box, typically at the border of a network, that translates one set of addresses into another (much like the translation example above). On the outside the NAT is typically public IP numbers; on the inside on the private network are used a private set of numbers not visible to the Internet. There is set-aside a specific block of IP numbers for internal private use that are never suppose to be seen in the public routing table. This is a nice conservation trick. The NAT can be assigned one IP number; behind the NAT on the private network can be hundreds or thousands of devices with private IP numbers that can be reached through the NAT. NATs, however, break the end-to-end design of the Internet and thereby do not work as well with certain applications such as P2P and VoIP. [ICANN 10-2007]
One of the benefits of IPv6 is that it increases the supply of IP numbers, removes the need for NATS, and thereby restores end-to-end connectivity. Applications can talk directly to each other without a NAT in the middle which can break that connectivity. Ironically, many network operators like NATs exactly for this reason, it breaks end-to-end connectivity. They would prefer to have one or a few addresses interacting with the Internet as a whole, as opposed to 1000s of addresses inside their network interacting with the Internet. NATs create single doors into networks at which a great deal of security can be implemented. SysOps have more control over what is going on and over their network. Even if IPv6 is implemented, a number of Sysops have indicated that they will continue to use NATs.
Overall US Transition
Matt Hines, What the US is missing by ignoring IPv6 , InfoWorld (Mar. 12, 2008) (" In the U.S., the relatively high number of IPv4 addresses still available and the widespread use of NAT has made IPv6 a "why bother?" project for most companies, Gomi said, even though most computers, operating systems, and network hardware built in the last four years support IPv6. ") Cost of Transition
Derived From: Potential Impacts on Communications from IPv4 Exhaustion & IPv6 Transition Robert Cannon, OSP, December 2010 Paper: Word | Acrobat |
The cost of transitioning to IPv6 could be problematic. Costs involved in the IPv6 transition include renumbering networks, running two separate networks (IPv4 and IPv6) simultaneously, upgrading relevant software and hardware, training staff, and testing implementations. 1The cost of IPv6 will involve capital investment and ongoing operational costs that will have to be diverted from other business goals and which can be difficult to bear in today's economic climate. 2 Some networks may be averse to expending financial resources to make the transition until absolutely required. 3 According to an IEEE White Paper,
A report generated for the National Institute of Standards and Technology (NIST) in 2005 stated that it would take 25 years to have a total transition to IPv6 at a cost of $25B, in 2003 dollars. However, a scholarly report on the adoption of IPv6 indicates that we will run out of IPv4 addresses well before the 25 years is up. Note that the same NIST report indicates the $25B would be less than 1% on network infrastructure spending, and they estimate the benefits of migrating to IPv6 are $10B per year .Also take into consideration that 25 years is still relatively fast for technology adoption. The introduction of digital switching to analog switching took more than 35 years. Moreover, there are still analog switches used in the public switched network. Likewise, we are twelve years into a 25-year migration from switched voice and video services to predominantly IP-based, end-to-end, voice and video services. What is different is the old technologies coexisted fairly well with the new technology, and it was hard for the average user to notice they were communicating with older technology (except for some features or quality).
The NIST report also mentioned the cost to ISP's for migrating to IPv6 would be $136M (2003 dollars). Again, this cost is a fraction of annual ISP network equipment spending, and thus should not be a major impediment. However, without a clear return-on-investment to the ISP, other than being able to offer IPv6 connectivity, it is hard to get them to make the investment. 4
Geoff Huston notes that ISPs will bear an additional cost as the result of the transition without an improvement of service to customers. Indeed, Huston notes, since many of the transition methods deteriorate end-to-end connectivity and quality-of-service, ISPs who deploy transition solutions will incur increased costs while offering inferior service - and thus will be at a potential competitive disadvantage. 5
Conversely, officials from the Defense Research and Engineering Network (DREN) have been sharing information from their IPv6 transition experience. DREN was an early IPv6 mover and was able to incorporate IPv6 into the regular lifecycle of their networks. As a result, they indicate that they were able to migrate their networks to IPv6 with little additional money set aside for the IPv6 transition. 6 The DREN experience suggests that, with planning, anticipated expenses could be mitigated.1 Next Generation Internet: IPv4 Address Exhaustion, Mitigation Strategies and Implications for the US, IEEE-USA White Paper, p. 17 (2009); Factsheet: IPv6 - The Internet's Vital Expansion, ICANN (Oct. 2007); Briefing Paper: IPv6 Deployment: State of Play and the Way Forward, Internet Society (June 18, 2009); IPv6 Economic Impact Assessment, RTI International for NIST (Oct. 2005) ; OECD Study: Economic considerations in the management of IPv4 and in the deployment of IPv6, p. 36 (May 2008) (" The cost of IPv6 deployment cannot be evaluated generically, as such costs vary on a case-by-case basis according to network needs and business "); Marco Hogewoning, IPv6 at XS4ALL RIPE 59 Lisboa (Slides).
2 Briefing Paper: IPv6 Deployment: State of Play and the Way Forward, Internet Society (June 18, 2009).
3 The financial cost of the transition was likewise an issue during the NCP-to-IPv4 transition. See Janet Abbate, Inventing the Internet, p. 141 (MIT Press 2000) ("Most host system managers had no compelling interest in converting to the Internet protocols, and the transition required a number of steps that would cost the host sites time and money").
4 Next Generation Internet: IPv4 Address Exhaustion, Mitigation Strategies and Implications for the US, IEEE-USA White Paper, p. 19-20 (2009).
5 Geoff Huston, Is the Transition to IPv6 a "Market Failure," The ISP Column (Sept. 2009) (The Transition Process). See also J. Curran, RFC 1669, Market Viability as a IPng Criteria, IETF (Aug. 1994) (" No internetworking vendor (whether host, router, or service vendor) can afford to deploy and support products and services which are not desired in the marketplace. ").
6 DREN Helps Make the Transition to Internet Protocol 6 (IPv6), Department of Defense DREN (accessed Dec. 4, 2010); John M. Baird, Defense Research and Engineering Network (DREN) IPv6 Deployment Joint Engineering Team (Sept. 21, 2010).
Major Networks IPv6 Status (This is not up to date)
| Network | Comment |
| Above.net | Received IPv6 Allocation from ARIN (per RIPE). |
| Allstream | |
| AOL | Received IPv6 Allocation from ARIN (per RIPE). |
| AT&T |
|
| British Telecom | Global Networked IT Services Key Features. |
| Cable & Wireless | "Since October 2000 Cable & Wireless offers native IPv6 peering at the INXS" INXS Native IPv6 Peering |
| Cablevision | |
| CenturyTel | |
| Charter | Charter Hints at DOCSIS 3.0 , Light Reading 's Cable Digital News (Feb. 27, 2008) (""We plan to test next-generation Docsis 3.0 in the second half of this year," said CEO Neil Smit during today's earnings call with reporters and analysts. He later noted that deployments likely won't follow until 2009.") |
| Cincinnati Bell | |
| Clearwire | |
| Cogent | Received IPv6 Allocation from ARIN (per RIPE). |
| Comcast |
|
| Cox | Mike Robuck, Migrating Your Network to DOCSIS 3.0, Communications Technology (Feb. 5, 2007) ("In contrast to Comcast, Cox's Ben Bekele, senior DOCSIS engineer, said IPv6 wasn't as big a priority for his company because of how Cox has its systems architected.") |
| Covad | |
| Deutsche Telekom | |
| Defense Research and Engineering Network DREN | Dual Stack. IPv6 Network Pilot |
| Earthlink |
|
| Embarq | Received IPv6 Allocation from ARIN (per RIPE). |
| France Telecom |
|
| First Communications | |
| Global Crossing |
|
| Hughes Direcway | |
| Insight | |
| Internet2 |
|
| KDDI |
|
| Level3 | |
| Localnet | |
| MediaCom |
|
| NTT |
|
| Mediacom | |
| Moonv6 | |
| Qwest | Received IPv6 Allocation from ARIN (per RIPE); IPv6 : A carrier's perspective, Qwest 2003 ARIN XI ("Qwest today operates a native [OC3] IPv6 network separate from IPv4 network") |
| TW Road Runner | |
| SAVVIS |
|
| Speakeasy |
|
| Sprint |
|
| Teleglobe | The peer 2 peer Internet: Crossing the IPv6 chasm to new territories of revenue opportunities , Teleglobe Dec 2004 US IPv6 Summit (Slide 36 "Teleglobe facilitates the world 's first intercontinental native IPv6
connection ; becomes a founding member of the IPv6 forum.
- Teleglobe presents its original IPv6 plans at the Telluride March
2000 IPv6 Forum
- 2003: Teleglobe starts an IPv6 pilot and is the first Canadian
carrier to announce commercial IPv6 introduction.") |
| Time Warner Telecom | |
| United Online | |
| Verizon / MCI / WCOM / UUNET |
|
| Windstream | |
| XO | ARIN IPv6 Allocation. |
© Cybertelecom ::Note: This is similar to the situation of Y2K disclosure statements. Survey of networks is based on "readily apparent information" found on websites. This means we went to the website and made a good faith effort to try to find the information. Information is not guaranteed to be complete or accurate.