Thursday Feb 3rd, 2011, as the most of the United States unburied itself from one of the largest blizzards in decades, a few news agencies like CNN and the AP put out a story that made the bottom of any list of headlines.  To most people it’s not a big deal.  To those of us that were around for the real birth of the Internet it’s a HUGE deal.

On Thursday, the primary issuing agency for IP addresses gave out the last unallocated block of IP’s to APNIC (Asia) for assignment.  This means that essentially there are no major blocks of IPs left in the original pool of IPv4 space.

Yes, there are still unallocated IPs in the v4 space.  Major carriers and providers still have IPs to give if you order services on Monday.  This exhaustion is at the core – so once AT&T and all of the others run out of addresses there will simply be none left.  Although as Asia and other countries adopt IPv6, and old IPs are placed back into service, it still doesn’t change the fact that we’re out.  While this is a temporary exhaustion of the pool, it still should be a major wake-up call to North America.

Back in 1999, IPv6 appeared and promised to be the silver bullet for our upcoming IPv4 woes.  In the original IPv4 space there were approximately 4.3 billion IPs.  Seeing as how there are six billion people on earth, and with smartphones and mobile devices taking more and more IPs, it’s conceivable that each person may need three or four IPs just to exist in today’s digital age.

IPv6 Moved us from a 32-bit IP space to a 128-bit space that provided for 340 trillion blocks of IPs, with each block allowing 1 trillion IPs each.  For those that can’t picture a number that big, one trillion-trillion is called a undecillion – that’s a 1, followed by 36 zeros – and with IPv6 we have 340 of those blocks!  That’s more IP addresses than there are estimated sun-like stars in the Milky Way Galaxy, and enough IPs that each person in the world could have 100,000 IPs each…  And we’d still have trillions left.

After it was announced that IPv6 was ready, there it sat.  It was the best untold story in IP engineering, and unless you worked for AT&T you didn’t even know it was there.  So why hasn’t anyone converted?  Actuall, they have, and they’ve been converting for quite some time.

The issue is that while most of Asia was an early adopter of IPv6 due to limited supply of IPs, most of North America has not jumped on the bandwagon.  Large carriers and Asia have already built much of their backbone on IPv6.  Network equipment manufacturers have supported it in most, if not all, of their hardware for years now.  Windows, DNS, DHCP, and other systems and services are fully aware.  The only thing holding back IPv6′s deployment in the enterprise?  Enterprises.

IPv6 has not been mandated for adoption.  End users still need to buy space from ARIN and it can be expensive.  In North America and especially in the US, we’ve had this “I’ll worry about it when it happens” stance to the issue – much like we did in the late 90′s with the Y2K issues.

Lets face it, the undertaking to convert is simply mind-blowing to most engineers.  Most engineers supporting networks today don’t have experience with it and haven’t been trained.  IPv6 is a different animal, and understanding hex is the name of the game.  NAT is no longer needed, and bridging between v4 and v6 space seems more like witchcraft than network engineering.  Engineers that grew up with v4 and its ways now have a lot of learning to do.  Calculating subnet masks is no longer something the seasoned engineer can do in his head – at least easily.

In October of 2010, I used an online survey to take a poll about IPv6 adoption.  I posted the survey to several forums and sent it to customers asking about planned IPv6 deployment.  Over 75% of respondents said they had no plans to implement IPv6 primarily because IPv4 was working just fine in their eyes, and they weren’t being forced to convert.  Basically the thought is “it’s not broken…”  Also, over 50% of respondents said they had no training in IPv6, weren’t planning to get any, and had never even chatted with their engineers about what it would take to convert!  Nearly 300 businesses, and a majority of them weren’t doing anything about IPv6.  And on Thursday we ran out of IPs.

Each business is different, and NAT, along with reclamation of IPv4 space as larger enterprises and carriers convert, will keep IPv4 around for a long time – I predict at least  6 to 8 years.  Small businesses that rely on RFC-1918 addresses shouldn’t immediately start converting to V6.  However large enterprises, with footprints on multiple continents and thousands of users, should begin to train and prepare their engineering staff for what most certainly will come.  As carriers run out of IPs engineers shouldn’t be surprised when the next MPLS link goes in with IPv6 required to route it.

As the pool of IPv4 space runs out completely, businesses are going to find themselves rushing to plan for what should have been done years ago – adopt and implement IPv6.  Those businesses that in the next two years refuse to start planing and looking into IPv6 are already behind the ball, and management has swung and missed.  So do yourself a favor…  If you haven’t had that talk with your VAR and favorite engineer, maybe it’s time to grab a beverage and start down the path.  Tomorrow will be here sooner than we think, and the last thing you want to hear when expanding your business is “sorry, we’re out.”