Filed under: Network Performance Management
Way back in 2007, our team here at AppNeta wrote about the panic in IT about the coming IPv6 transition. It’s entertaining now to read about this particular panic, and about the more general technology-specific panic that has occasionally gripped anyone in IT thinking about worst-case scenarios.
Back then, the big worry was that the pool of IPv4 addresses had been depleted. But users have gotten around that with the paid pools of addresses that are available. Nine years later, the internet hasn’t collapsed under the weight of IPv4 addresses. We’re still gradually moving toward IPv6—though it feels less urgent or noteworthy than it did then. The IPv6 migration plan for many enterprises is hardly a plan, but rather an evolution along with the market, hardware and needs. At the moment, enterprises often have both IPv4 and IPv6 in use. This dual-stack approach means that companies can support both protocols as needed, without a dramatic shift from one to the other and the associated costs and change management.
Google estimated that IPv6 can claim about 10% global adoption (that was about a year ago). They keep an updated graph of IPv6 adoption among their users, reflecting the long path it’s taking. This adoption seems slow compared to earlier network technology shifts like WiFi. That may be because migrating from IPv4 to IPv6 doesn’t just require one configuration change or hardware upgrade. It requires that every piece of the network chain gets the IPv6 upgrade, including routers, servers and firewalls. Our old blog post tagged the migration process per organization as taking about two years. That’s probably an overestimate now.
Other estimates think it’ll be close to a decade before IPv6 is totally implemented, even though it’s been around close to 20 years. Here at AppNeta, we added IPv6 support this year to support customer needs.
IPv6 Adoption Crawls Along
Why the lag time in adopting IPv6? Maybe it’s not as urgent as it once seemed, or maybe it’s made more complex by cloud hosting providers and other off-site concerns. There are other possible reasons, too, like network address translators (NAT) prolonging the IPv4 lifespan and the lack of backward compatibility from v6 to v4.
So while we’re entertained by our 2007 counterparts’ IPv6 worries, it’s not a huge concern for anyone in IT these days. Maybe that’s because we have bigger problems, like figuring out cloud network metrics, migrating tons of data to the cloud or solving user problems. Which of IT’s current worries will have completely fizzled in seven years from now? Only time will tell.