In the previous blog, I discussed the near term event were we will see ARIN distribute the remaining public IPv4 address space to applicants. This event is coming and there is nothing we can do to change that fact (there are a finite number of IPv4 addresses), so our best course of action is to respond to this in some sort of organized fashion. As I mentioned, there are 3 scenarios by which more addresses will become available and we are likely to see a combination of all 3 as we move forward. As with all change there are risks and rewards and I’ll try and describe some of the bad news and good news about this particular change.
The bad news…
The risk of Internet routing get (more) “disorganized” as a result of non-ARIN officiated transfers will increase if the secondary market is unregulated. The association between issued IPv4 address space allocations and the organization to which it has been issued and/or uses it in North America is managed by ARIN (also known as the Whois database). Due to certain ARIN policies that control the transfers of IP addresses by organizations (NRPM 8.3 or NRPM 8.4), there will assumedly be a desire to circumvent the official process. This IPv4 address space black market will cause greater amounts of discrepancies between ARIN records and what is actually in place on the Internet. Consumers of this black market IPv4 address space will also be gambling with their company’s Internet stability by running services using unofficially transferred IPv4 addresses.
There can be technical implications in the scenario (#3 ) where organizations re-organize their external address space to be more efficient, enable growth, or try to profit through sale of owned addresses. The use of complex, and sometimes unstable, solutions such as Carrier-grade Network Address Translations (CGN) can cause issues with Internet applications. We have already seen reports of CGN causing issues with popular gaming console, unified communications applications, and specific VPN technologies where the application is unable to establish a full connection between end points. A more detailed report of specific issues observed while using CGN has been published in RFC 7021 by the IETF.
The good news…
This event was seen coming a long time ago (to the point discussed in my previous post where Microsoft spent what it did four years ago to buffer itself from the impending exhaustion) and those that are immediately affected have (hopefully) planned for it by ensuring they have sufficient existing IPv4 address space allocated and/or the free cash to purchase more on the open market. Much of the world has also been working very diligently on the end game solution, IP version 6 (IPv6).
Without getting too deep into the technical details, IPv6 provides 2^96 times more address space than made available by IPv4. One of the best comparisons I have read depicts the IPv4 address space as a postage stamp to which the IPv6 address space would comparatively be depicted by the Solar System. By transitioning from IPv4 to IPv6, the Internet will be allowed to grow for the foreseeable future. Forward thinking companies have already begun, if not completed, implementation of IPv6 throughout their network in order to support this new protocol and address space.
For the time being…
All Internet sites for the immediate future will remain accessible via IPv4 and ever greater amounts will add support for IPv6 as well. But despite not being immediately affected, the time is coming for the rest of us where we will have to take action. This will happen when the cost of open market IPv4 address space exceeds that which the market can bare and some services are ONLY accessible via IPv6 addresses. When this happens and you cannot reach a web site because your network does not support IPv6, the impact will be felt.
One solution that you should be aware of is when ISPs start peddling an IPv4 Internet connection supported by CGN. While many consumers will never realize an impact from the use of CGN, this can have very real impact on businesses by limiting the functionality of their Internet connection. Using CGN should, in the worst case, be a temporary fix and since the technical debt it creates directly impacts a network’s performance, stability, and functionality.
In my next and final post on the topic of ARIN IPv4 exhaustion I will present some information about what organizations can be doing to prepare for the impending requirement of IPv6 support including a short and simple assessment to evaluation your current state of readiness.