Despite its generous capacity, IPv6 is not ready to satisfy the increasing demand for Internet connections because very few network operators and end users are actually moving to IPv6, due to a number of challenges. Replicating the IPv4 infrastructure hardware and software, and developing IPv6-specific applications and other content — while still operating the IPv4 stack — is expensive without any meaningful immediate return on the investment.
By the end of April 2015, IPv6 traffic represented approximately 6% of total Internet traffic, according to Google's IPv6 adoption statistics. Migration to IPv6 will not occur fast enough to support continued Internet growth before all of the RIR IPv4 free pools (with the exception of Africa) are effectively depleted in mid-2015.
Life after exhaustion: The workarounds
IPv4 free-pool depletion workarounds are filling some of the gap. Dual stacking is a network infrastructure architecture that requires each end point, and every router and switch in the "dual-stacked" network, to run both IPv4 and IPv6 protocols in anticipation of receiving data in either format.
Tunneling is a network configuration, with supporting systems, used to encapsulate IPv6 packets into IPv4 packets to facilitate communication between IPv6-enabled end points over an IPv4-based network.
Dual stacking and tunneling ease the IPv4 to IPv6 incompatibility problems for organizations that want to switch their networks, content and applications to IPv6 while continuing to communicate with their end users, peers and upstream network operators who largely remain on the IPv4 Internet. But these technological "fixes" do not generally improve the business case for investing in a migration to IPv6 — not when most of the content and customers remain on IPv4.
Large scale/carrier grade Network Address Translation is a third technical workaround. CGNs allow many end points served by a single carrier to share a smaller number of unique IPv4 addresses. By using CGNs, network operators can extract more value from their existing IPv4 infrastructure investment and reduce their need for additional IP numbers, but there are costs.
Designing, procuring, implementing and operating CGNs give rise to capital and operating expenses. There are also other problems. Endpoints sharing IP addresses behind CGNs are not separately discoverable from the public Internet. Because more than one subscriber may be using the same IP address, CGNs can frustrate law enforcement seeking to identify bad actors, and impede proper functioning of Internet security and other Web-enabled applications that depend on unique endpoint IP address mapping.
Extensive use of CGNs adds complexity to the networks that use them, and that complexity can compromise network availability, reliability and scalability. There is also a concern within the Internet community that end-point obfuscation caused by CGNs will stunt the development of important Internet innovations and adversely affect its architecture.
Sign up for Computerworld eNewsletters.