Why are we here!

The internet – particularly the North American part of the internet – is out of the most important building block required for its foundation….the IP address.  The internet is supposed to be this infinitely expandable utopia where everyone can build whatever interconnections they want and we can just keep expanding forever without any central bottleneck or resource depletion, but limited availability of IPv4 space has really popped that balloon.
The original IPv4 pool of addresses contains approximately 4.3 billion addresses.  On February 3, 2011 the last /8 address blocks were allocated to the Regional Internet Registries (also known as RIRs, these groups include ARIN, APNIC, RIPENIC, etc. and are the bodies that actually allocate IP addresses to end user networks within their geographic region) to allocate to their end users.  With that allocation, IPv4 addresses were officially exhausted.
The question remains – why are we here? How could 4.3 billion addresses run out so quickly?  How could this problem of exhaustion not gain any visibility until it was too late?  Why weren’t there more alarm bells?  What do we do now? For all of these questions there are two sets of answers.  The first set comes from the RIRs and ARIN (the American Registry of Internet Numbers) in particular.  These answers are the byproduct of ARINs amazingly complex bureaucracy and its policy makers who are more interested in deflection and nursing that bureaucracy than building a functional organization.  The second set of answers is the truth as seen by the experts here at Accuro.
The “official” reasons for IPv4 Depletion
Failure of the market to adopt widespread use of Network Address Translation (NAT) for server-side applications.
Advent of multiple devices per Internet user.
Inefficient use of allocations by end users
Increased demand by the end users of other RIRs (particularly users in China, Brazil, and India.
High density data centers requiring more than one address per machine (i.e. cloud computing and virtualization).
Before we get on to our answers, we’ll take on each of these “official” reasons briefly.  The reader will notice that each of the “official” reasons for IPv4 exhaustion has a grain of truth in it, but does not excuse our current situation.  Similarly a farmer who planted corn in the Sahara desert with borrowed money would not be excused by his creditors after his crop failed if he simply stated the obvious….”It didn’t rain enough.”  The farmer remains at fault for making bad choices from the start.
NAT in the datacenter environment has, and always will be, a pipe dream.  If you owned a company that sold TVs, you could very easily publish several official documents advising your clients to smash their TVs before going to bed every night, however, it would have to be some really compelling writing to entice your customers to waste that much money based on a white paper.  Server side NAT finds itself in the same predicament.  It certainly would help ARIN and all the RIRs to a great extend, however, no rationale datacenter operator, host, or corporate user is going to give up the freedom required to implement it just be a good corporate citizen.  Using the failure of the market to adopt a hideously ineffective conservation measure as an excuse for the shortage IP space is not viable.
The other reasons are equally as weak.  Anyone with a pulse in the late 90’s new that every tech company wanted to sell the average consumer as many internet gadgets as possible.  It was only a matter of time before the technologies caught up with demand.   Similarly, anyone with a pulse in the late 90’s should have known that India, China, and Brazil were rapidly modernizing and where going to need a lot of infrastructure in the coming decades.  ARIN and the other RIRs approved every single allocation, so there is no blame to be laid upon end users.  Very few have broken the rules.
So what are the real reasons?
If you’ve read this far, you probably have a pretty good idea of what comes next.  The real reason for IPv4 depletion rests solely on the shoulders of the RIRs and their practices.  The RIRs management of the IPv4 pool has been the greatest – and most lasting – failure of any internet related resource.  This is not to say those poor souls working at the RIRs – and carrying out the allocation policies – are incompetent.   Instead, we are saying that the legacy the RIRs inherited was doomed to fail the marketplace and, furthermore, the system for allocating IPv4 addresses was fatally flawed.
Prior to the creation of the RIRs, large allocations (normally a /8 or an incredible 1/256th of the entire internet) were handed out much like candy to children, primarily to companies and universities involved with the creation of the internet.  Did MIT expect to one day have 1/256th of 4.3 billion computers on its campus?  Why did these organizations accept and why do they refuse to give back these allocations that have crippled the rest of the internet?  The answer is simple – there’s no incentive to take exactly what you need and there’s no incentive to give back any extra space in your inventory.  If anything, there’s a disincentive to do so as it may cripple your unexpected growth in the future.
Let’s say a new uninhabited island the size of Texas was discovered just off the coast of the mainland US and the US government decided to give away all the land to current citizens.  Let’s also say the method for giving away that land was to have a citizen submit paperwork showing, without any real verification, how much land they’ve used historically and how much they plan to use in the future.  How many people are going to say, “No No!  I only need 50 acres.”?  Very few.  The vast majority are not only going to get all they can get but they are going to try to hang on to all that land – even land they aren’t using.  There’s no incentive to ask for more and there’s no incentive to give back unused land.
Once the RIRs were created, the same land-grab system was maintained.  A provider asked for a certain amount of space and, with minimal verification, these requests were granted if the provider met fairly simple steps outlined by each RIR.  This system encourages providers to ask for – and keep – as much space as is even remotely plausible.  The result is huge amounts of space sitting idle in the virtual “IP Bank Accounts” of end users worldwide.
The solution is simple
Indeed it is.  Monetizing IP space is the only way to maintain the IPv4 pool until IPv6 is a reality.  IPv4 doesn’t make something that is finite suddenly become infinite but it does buy the internet some time.  This isn’t an academic argument anymore.   A solution must be found to keep our internet functioning and to keep all the end users connected to the internet growing.
Check back for upcoming blog posts.  In our next post, we’ll explore the mechanics, plusses, and minuses of monetizing IPv4 further.  Also look for upcoming posts on the reality of IPv6 and other exciting subjects.
Your feedback, questions, and topic ideas are always welcome.  Please email us at blog [at] ipaddressbroker [dot] net

This entry was posted in Accuro Consulting. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *