IPv6 Prep Warnings Get More Urgent

Service providers are banging the drum for the coming shift to IPv6, hoping to convince business customers that they need to prepare well in advance of the exhaust of IPv4 addresses, expected to happen in 2011.

Verizon Enterprise Solutions issued the latest such warning today, putting out a detailed set of instructions for how businesses should be planning for this crucial transition, well ahead of being forced to do so. (See Verizon Offers IPv6 Transition Aid.)

Qwest Communications International Inc. (NYSE: Q) had issued its own statement in mid-summer, and both companies say they are providing professional services assistance to businesses. NTT America Inc. has been very active on the IPv6 front as well, encouraging companies to make the transition sooner rather than later. (See NTT America Charts Future Plans.)

The effort began earlier this year but has been picking up steam since spring, says William Schmidlapp, senior consultant in product marketing for Verizon Business. Verizon's warning encourages businesses to look in detail at their own situation now because every business has its own issues to address, based on the way the business is operated and how it uses Internet connections. (See Too Many Not Ready for IPv6, Lab Warns and Is IPv6 Finally on the Verge?.)

"We want our customers to not only look at their Internet-facing capabilities but their network capability overall," Schmidlapp says. "One of the things we try to communicate is that businesses need to be doing a deep dive within their own internal infrastructure to know how that infrastructure interacts with devices that hit the Internet, such as smartphones, or getting content from the Internet. And how much of the consumer-based market do they want to reach for content delivery or e-commerce?"

The proliferation of smartphones and the coming of 4G wireless networks are both drivers for IPv6 deployment, and businesses may be unpleasantly surprised to realize there is a segment of the consumer population they can no longer reach, if they haven't prepared, Schmidlapp warns. Some of the most vulnerable applications may be homegrown systems, with embedded IPv4 addresses but no obvious need for a IPv6 upgrade.

One reason businesses have been slow to embrace the IPv6 transition is that many of them see it as a technology transition with no revenue payoff, admits Qwest CTO Pieter Poll. But a major reason to plan for the transition now, ahead of IPv4 number exhaust, is to use capital resources in the best possible way, upgrading only where necessary.

"That's why we are encouraging them to do an in-depth analysis of how this affects them," Poll says. "They need to look beyond the obvious -- 'Does my carrier support IPv6? Does my router support IPv6?' to their own back-office systems. That way they can determine what has to be done now, and what they can afford to put off until later."

Doug Junkins, CTO of NTT America, says businesses may find the IPv6 transition isn't really about forklift upgrades to major systems, but about fine-tuning and adapting what you have.

"The fear about having to spend money on this problem is larger than the reality," he says. "Most of the major equipment vendors have been shipping IPv6-enabled hardware and software for some time now, so it's really more of a process of testing and tuning systems. It does take time and money from a resource perspective, but I don't think it is the huge capital outlay that some enterprises are fearing."

This is a commercial opportunity for service providers, to provide consulting and professional services, Schmidlapp admits, but there are many things that enterprises can and should be doing on their own, since they are most familiar with their own operations.

"From a professional services perspective, we are working with customers in order to help educate them on V6 and guide them on some of the things they need to look for -- how their apps are set up, how their infrastructure is set up," he says.

Alliance for Telecommunications Industry Solutions (ATIS) , the US technology standards group that includes all major telecom service providers, has published a set of best practices for IPv6 readiness, Poll points out. And this is one area on which competing service providers agree.

The real danger, Junkins says, is that enterprises that don't plan ahead will find themselves scrambling at the last minute, probably spending more than necessary, to keep critical applications up and running.

"Companies that put their heads in the sand on this may find it's more cost prohibitive in the end, and they may wind up with equipment that isn't exactly right for their business," agrees Verizon's Schmidlapp.

— Carol Wilson, Chief Editor, Events, Light Reading

Page 1 / 4   >   >>
shygye75 12/5/2012 | 4:24:49 PM
re: IPv6 Prep Warnings Get More Urgent

The IPv4 death knell began sounding somewhere around the time that Bill Clinton moved into the White House. At some point, the ring has to be true.

Pete Baldwin 12/5/2012 | 4:24:49 PM
re: IPv6 Prep Warnings Get More Urgent

Anyone else get a Y2K feeling out of IPv6?

Jeff Baumgartner 12/5/2012 | 4:24:48 PM
re: IPv6 Prep Warnings Get More Urgent

Y2K or Chicken Little... not sure which... a little bit of both? JB

jdbower 12/5/2012 | 4:24:46 PM
re: IPv6 Prep Warnings Get More Urgent

It's a simple question of economics.  Mid next year we should "run out" of IPv4 addresses.  The Internet won't crash and things will work fine, but unquestionably the way we acquire IP addresses will change.  Rather than getting them on the cheap, we can now have bidding wars from companies and institutions with class As who don't actually need them.  Rather than a single NAT we can go through many - breaking things along the way.  And we'll see people not wanting to deal with all that and realizing that IPv6 isn't that scary.  

I've got 2^48 IPv6 addresses for my own personal use from SixXS - and they were free.  That's enough addresses to model every device on the Internet today thousands of times over.  And, as a hobbyist it took me perhaps a day to set up an IPv6 network including IPv6-enabled router, web server, DNS server (that listens on IPv6 instead of cheating and only handing out AAAA records), and other goodies.

So the net result will be figuring out if $100 each for IP addresses is the breaking point or $1000 each and determining whether the cost of deploying and supporting tiered NATs is more expensive than an IPv6 address space.  Either way, now's a good time for IT professionals to lose their fear of IPv6 because, like COBOL programmers circa 1998, they can command some decent money by getting people onto v6 before people realize how easy it is.

Pete Baldwin 12/5/2012 | 4:24:45 PM
re: IPv6 Prep Warnings Get More Urgent

Thanks JD.  In a way, you've tapped into what I was thinking about Y2K.

Y2K was real -- meaning, the year 2000 really happened, and some computers really did misinterpret the date as 1900. What didn't happen was the massive collapse of society... I don't recall water systems or food supply lines getting cut off, at least not here in Silicon Valley.

That's partly because people did the work to avoid catastrophe. But it's also because the problem was deeply overstated.

IPv6 lacks the same shrillness, but I think the analogy holds.  It's not the end of the world.

IPv4 address exhaustion is a real problem, and in some cases, it's going to cost money. I'm having a hard time seeing major catastrophe coming out of this, though. Companies will eventually migrate to IPv6, and many of them will do it beyond the last minute. We've already known that for years.

guerin 12/5/2012 | 4:24:44 PM
re: IPv6 Prep Warnings Get More Urgent

While IPv6 and Y2k may share the fact that an end-of-the-world analogy was (over) used for both, there are fundamental differences.

Y2k was a punctual event,  Things broke or they did not, and if they broke you could then rush to fix them.  And yes, if the break was bad enough, it could kill you (figuratively speaking)

IPv6 wont happen in one instant and wont kill you on the spot.  It can, however, bleed you (the Internet) to death if the migration takes forever to happen.  The issue is that the IPv6 decision is not yours only.  It has Internet-scale applications and costs.

Let me clarify what I mean by that.  If today you get an IPv6 address, and only n IPv6 address, i.e., no IPv4 address for you, what you can reach is a minute fraction of the Internet (less than 0.2% by our latest, arguably partial measurements -see http://mnlab-ipv6.seas.upenn.e... for that data).  So unless the IPv4 Internet jumps on the IPv6 bandwagon, i.e., IPv4 devices get IPv6 addressess as well, we are stuck with a need for ever bigger NAT-live devices (e.g., a la DS-Lite).  They can break things, but more importantly will end-up costing a lot of money if you have to keep them around for a long time.

So the real challenge is not to force new users to adopt IPv6.  They eventually will, when IPv6 addresses are free and IPv4 addresses cost money.  It is to convince the current IPv4 Internet that it should become IPv6 accessible.  This was the original plan (everyone had both).  Unfortunately, like many plans, things did not exactly addording to it.


My 2c.

guerin 12/5/2012 | 4:24:43 PM
re: IPv6 Prep Warnings Get More Urgent

I did not say there was an obvious reasos for current IPv4 users (companies) to become IPv6 accessible.  As a matter of fact, I think I pointed to the exact opposite, which indeed translates into hardly anyone having budgeted anything for migration.

The main reason is that there are no real incentives to do so.

Existing companies already have IPv4 addresses and often enough to sustain some period of future growth.  The ISPs are the ones who will be the first to need new IP addresses for new customers (some even switched to using IPv6 simply because their internal management infrastructure -that used private addresses- was getting too big).  The issue they then face is that if they give someone an IPv6 address (because they don't have any IPv4 addresses left), they *have to* ensure that this IPv6 address gives that customer the same functionality as an IPV4 address.  And in today's world where with only an IPv6 address you can reach about 0.2% of the Internet, it means that they will have to start deploying IPv6<->IPv4 gateways to allow those IPv6-only customers to access the (IPv4) Internet. 

Those boxes cost money, but the worst part is that once they are there, they further decrease any incentive for the existing IPv4 companies to become IPv6 accessible.  Why would they?  IPv6 users can now access them on their IPv4 address...  This means that this transition period is likely to drag on for a long time, and as a result be much more expensive than it really needs to.

This was the main point I was trying to make, i.e., unless Internet content becomes widely IPv6 accessible, we'll be stuck in a transition period for a very long time. 

The other point is that if you want to break this cycle, you (ISPs) need to identify a good enough incentive for current IPv4 users to budget for the cost of becoming IPv6 accessible.  My own pet suggestion is to ensure that you get better service through native IPv6 access than if you have to go through gateways (that's unfortunately not always so today), and preferably service that is at least as good as what IPv4 customers get.  If that is the case, at least as the number of IPv6-only users start growing, current IPv4 companies will see an increasing benefit to becoming themselves IPv6 accessible.

paolo.franzoi 12/5/2012 | 4:24:43 PM
re: IPv6 Prep Warnings Get More Urgent


How many companies (do you think) have budgeted to migrate to IPv6?

The big thing about Y2K was that it caused a bajillion dollars to be spent on the IT infrastructure in modernization.



davekresse 12/5/2012 | 4:24:43 PM
re: IPv6 Prep Warnings Get More Urgent

There is a real challenge underlying the transition to IPv6.  Many of the dominant services today including voice and video contain embedded IP addresses within their underlying protocols.  These applications and services, as well as the underlying infrastructure, need to be modified in order to work as end users expect them to when the migration occurs. 

The article is right in pointing out that this is fundamentally a testing challenge, and the good news is that solutions exist today to ovecome this challenge.  The path to certifying IPv6 readiness is out there with respected labs like the University of New Hampshire.  In addition, there are commercially available test solutions that can help operators and vendors ensure that their solutions will work in the dual-stack environment we can expect to see for many years.  The key is starting now and thinking through the right testing strategy with the right solutions in order to get to this new world quickly with minimal disruption, and with the confidence that end users will continue to get the applications and services they need. 

Pete Baldwin 12/5/2012 | 4:24:42 PM
re: IPv6 Prep Warnings Get More Urgent

I like your points, Guerin.  To me, that's the more interesting issue here -- not "OMG the numbers will run out when we expected them to run out."  I like your prediction of a long, lingering transition (maybe because I've been following the death of Sonet for about a decade).  It'll be interesting to see who benefits from that.

Page 1 / 4   >   >>
Sign In