The next best thing to IPv6?
(IDG) -- IPv6 may have an understudy waiting in the wings.
The Internet engineering community, concerned about the snail-paced deployment of IPv6, has begun revving up development of a protocol that could serve as either an interim measure or an alternative to the next major version of the Internet communications protocol.
Currently in draft form, the Realm Specific Internet Protocol (RSIP) solves some of the addressing constraints of today's IPv4. RSIP also supports end-to-end security and allows for a smooth transition to emerging Internet applications, such as teleconferencing, instant messaging, streaming media and telephony. RSIP would reside in client and router gear.
For enterprise network users, RSIP offers a replacement for Network Address Translation (NAT), a common IPv4 workaround used by organizations to support multiple private addresses through a single public Internet address. RSIP also could be used to aid in the transition from IPv4 to IPv6, which will be a daunting endeavor for network managers.
The transition to IPv6 is a huge task that will require users to update IP on every device on the network. This requirement is widely recognized as one of the huge drawbacks for IPv6.
The Internet Engineering Task Force (IETF), which has completed work on IPv6, revealed its decision to develop RSIP as a backup plan for IPv6 in a draft document posted on its Web site last month. The document details an invitation-only meeting that the IETF's Internet Architecture Board held in July to consider a variety of scenarios, from full migration to IPv6 to complete failure of IPv6.
"This represents the first time the IETF has seriously and formally examined the possibility of IPv6 not succeeding," says Noel Chiappa, a participant in the July meeting and a former area director for the IETF. "Before we had Plan A - IPv6 - and there was no Plan B. Now we're starting to work on one."
As part of its contingency planning, the IETF made several recommendations to allow the network layer of the Internet to support more users and applications until IPv6 is deployed. One key recommendation was to speed up RSIP development.
"The IETF is proceeding with cautious optimism on RSIP," says Mike Borella, a senior engineer with 3Com's advanced technologies group and one of the authors of the protocol. "They've given us their blessing to continue the work."
At the heart of the debate over IPv6, RSIP and NAT is the fact that Internet address space is running out. How fast space will run out has been the topic of much debate over the years. Within the IETF, projections vary from two years to 20 years.
IPv6 solves the addressing problem by replacing the 32-bit addressing scheme in IPv4 with 128-bit addressing. The larger address space allows for an exponential number of devices to be connected directly to the Internet.
After six years of development work, the IETF recently finished defining IPv6 standards. However, few compliant products are available and only a handful of end users have adopted IPv6.
Because IPv6 products were not widely available to solve the Internet address shortage, many network managers deployed NAT devices as a quick fix. IETF leaders say NAT devices are so prevalent that they could cause serious problems for the growth of the Internet over the long term.
The Net was designed so every device would have a unique address; with NAT this is no longer true. NAT also prevents end-to-end security and makes it harder to deploy new Internet applications.
"Ultimately, translation-based scenarios lead us to a problem," says Brian Carpenter, chair of the IETF's Internet Architecture Board and program director for Internet standards and technology at IBM. "It's not a crisis this week or next week . . . but long term it will create difficulties." RSIP was introduced to the IETF 18 months ago by 3Com, but it was just last month that the protocol was given the blessing of the IETF's top leaders as a potential replacement for NAT and a step toward IPv6. NAT does not currently have a transition path to IPv6; RSIP allows for one and makes it so NAT and IPv6 aren't mutually exclusive.
"We were assuming a transition between IPv4 and pure IPv6," Carpenter says. "The good news is that there is also a transition path from NAT and RSIP to IPv6."
Enterprise users also are interested in RSIP as a replacement for NAT. Robert Cooper, a project leader with Indianapolis Power and Light, uses a 3Com OfficeConnect ISDN LAN modem with built-in NAT to dial out over the Internet from any of the machines on his 12-client network at home.
From his network, Cooper can access any of the computers in the company's power plants to check system health or conduct remote programming. "I would be very interested in RSIP because I wouldn't have to run through hoops to get encryption," he says.
Arlington Industries, a distributor of imaging supplies based in Libertyville, Ill., is using the NAT features of Secure Computing's Sidewinder firewall to hide the private Internet addresses of its 100 employees behind a single public Internet address.
Network administrator Chris Kozlov says he would be interested in RSIP because it would let him encrypt all Internet traffic, including Web surfing. "We'd like to be encrypting users regardless of the type of content," he says.
Until RSIP is finalized, which will likely be another year, IETF leaders are encouraging end users to think twice before deploying NAT.
"If you have NAT in your enterprise and your business partner has NAT, there may be some applications that you develop or buy that will not work [because of interoperability problems with NAT]," says Matt Holdrege, chair of the IETF's NAT working group and manager of technology development at Lucent Internetworking Systems.
For example, NAT doesn't allow for end-to-end security applications, nor does it support any Internet protocol that puts addressing information in the body of the packet rather than the header.
Meanwhile, the IETF is evaluating protocol enhancements that would make NAT devices work better.
"I think RSIP, while interesting, is less likely to be the key part of the eventual path forward than NAT simply because using it means you have to modify the end hosts," Chiappa says. "That is the great charm and benefit of NAT, and it's the reason it's most likely to succeed."
"Original" hypertext project delivers -- 30 years later
RELATED IDG.net STORIES:
Microsoft stalls IPv6 progress
Realm Specific IP: Protocol Specification (IETF draft)
|Back to the top||
© 2001 Cable News Network. All Rights Reserved.|
Terms under which this service is provided to you.
Read our privacy guidelines.