Five years ago, I wrote a column questioning the need for IPv6 because I felt IPv4 was more than adequate. Two years later, I wrote another column basically saying the same thing - which did nothing to endear me to the IPv6 contingent. Then last year, my interest in IPv6 heightened, and I wrote a column recommending that users consider IPv6 for any new implementations.
Now, one year later, I gaze across the network horizon and ask, So where is all the IPv6? Even with all the hype about IPv6, I know of very few implementations outside of educational institutions, the federal government, telecom companies or research organisations. In the corporate world, IPv4 remains king. And from what I can see, it will remain king through this decade.
That doesn't mean IPv6 is dead. Eventually, most companies will be forced to switch to IPv6 because of vendors eliminating IPv4 support. While IPv6 has definite technical advantages, the business value and functionality doesn't justify the conversion costs.
The increased availability of WAN technologies such as Border Gateway Protocol-based Multi-protocol Label Switching (MPLS) VPNs facilitates the continued use of both IPv4 and private IP address spaces. The highly touted increased address space of IPv6 still is not critical to most corporations. IPv4-based technologies such as IPSec, SSL, MPLS, Differentiated Services and router-based queuing mechanisms adequately match the security and quality of IPv6 service enhancements.
From an application perspective, IPv6 actually can create problems. While operating system vendors are shipping IPv6 stacks and IPv6 versions of HTTP and FTP, most off-the-shelf applications are not IPv6-based and might require the use of tunnelling or other modifications to work properly in an IPv6 environment. This only adds to the cost, complexity and service issues of an IPv6 migration.
As most corporate IT organisations are cost centres constantly under pressure to reduce expenditures while increasing service, IPv6 still is seen as a solution looking for a problem. A migration to IPv6 will neither reduce near-term operational costs nor enhance functionality. On the contrary, migrating to IPv6 will create short-term cost increases and might affect service because of training, support and implementation issues. Budgetary-constrained IT managers are opting to invest their dollars in other initiatives that will solve immediate problems.
While IPv6 might never be "required," I think it will replace IPv4 when the economics are right. Most new equipment supports IPv6. IT managers are laying the foundation for the eventual migration to IPv6 as they replace their fully depreciated legacy routers, switches, desktops, servers and operating systems through normal end-of-life processes.
Within five years, most corporations will have an infrastructure capable of supporting IPv6. At that time, the transition to IPv6 will become a "no-capital" tactical initiative that will be more financially palatable to corporate management. Until then, IPv4 will continue to reign supreme in the corporate world, providing cost-effective networking services and functionality.
Chuck Yoke is director of business solutions engineering for a corporate network in Denver.