Broadband

IPv6 Day post-mortem: Smooth sailing or hidden gotchas?

Mark Underwood provides the after-action report for the recent IPv6 day. Find out how the trial-run actually went for some of those who participated and what overall participation was like.

June 8, 2011 was World IPv6 Day. So how did it go? As with any assessment, it depends on whom you ask and how you phrase the question.  I decided to ask one of the firms who advertised its participation in the exercise. Meebo is a mature startup (2005) that offers a Web-based communication and discovery service. A user of the chat service for several years, I noticed their announcement posted the day before IPv6 Day. What had they learned from the test? How much preparation had they needed?

In speaking with Meebo for this story, I got the impression that the firm needed no special incentives to participate, but it's also true that Meebo's operations meet one of AT&T's early IPv6 adopter criteria: "A business that relies on marketing and sales communications with Asia should place the highest priority on adopting IPv6."

Meebo's Report from the Front Lines

Following are the highlights of my discussion with Mark Doliner, Lead Developer at Meebo.

Q: What did Meebo do in-house in advance of Test Day to prepare? Sandbox, etc.

A: Planning meetings, configuration changes, a few code changes, some testing.

Q: Did you encounter any DNS issues?

A: We did encounter a bug with one of our DNS providers where one of their servers returned an empty response rather than a CNAME record when the client queried for a AAAA record. We notified them of the issue and they were able to fix it before the end of World IPv6 Day.

Q: Did you invest in any specialized staff training or consultants?

A: No. Meebo's internal development and operational teams handled the changes.

Q: Did you measure how many IPv6 connections were received during Test Day?

A: Yep! It was somewhere in the ballpark of 3 percent of our users. [Ed: This number closely matches the results found elsewhere.]

Q: I saw that you notified users by Meebo note. Did you make other help desk preparations?

A: We briefed our support team on how to assist users who may have experienced problems, and we created an email alias for IPv6-specific questions.

Q: Do you anticipate increases in WAN traffic?

A: We don't anticipate any increase in WAN traffic from the use of IPv6.

Q: Going forward, what will you be measuring to help plan for IPv6 transition?

A: We would like to add a permanent AAAA record for our domains in the future, at which point we'll measure how many users connect over IPv6 vs. IPv4.

Q: Do you see any indirect competitive advantages or is this just a technology compliance initiative?

A: A little of both. We believe that IPv6 is important to the future of the Internet and we want to make sure Meebo remains accessible by everyone as the remaining number of unused IPv4 addresses continues to decline.

Q: Any lessons learned not mentioned?

A: Hurricane Electric's Tunnel Broker service at http://tunnelbroker.net/ is fantastic.

Lessons learned by others

Other participants noticed issues with IPv6 VPN's - for at least one organization, it was a problem that is still being investigated as of 16 June. Also, because much of the IPv6 traffic was non-native (i.e., it was tunneled or encapsulated within IPv4 to ensure compatibility), IPv6 packets were showing slightly slower latency. Some security analysts expressed concerns that the protocol's implementations are relatively immature, and thus more vulnerable to threats.

Despite these concerns, Cisco reported that there was no dramatic uptick in call center call volume.

Good Test? How many sites participated in World IPv6 Day? According to Australia's WatchMouse, only 17% of the top 500 sites supported IPv6 on their main "www" host using a dual stack approach, and 11% supported it on separate IPv6-only sites. Not overwhelming. Is your organization on the list of procrastinators?

This is one case where the U.S. Department of Defense is in the lead. Not surprising since that's where the Net was born. For instance, the Navy's SPAWAR has had years of operating IPv6 on its network, and has worked with vendors to identify problems with specific features. SPAWAR's Ron Broersma, enterprise network security manager and chief information technology division engineer, warned last year that testing of products can be challenging because quality assurance suites were not ready for full IPv6 testing. On the plus side, the Navy was able to transition to IPv6 without additional staff, and was able to deploy IPv6 hardware as part of its routine technology refresh process.

If you want to check current network readiness, try the Internet Society's IPv6 interactive connectivity assessment. My Verizon FIOS readiness for native IPv6 scored an unimpressive 0/10.

About

Mark Underwood ("knowlengr") works for a small, agile R&D firm. He thinly spreads interests (network manageability, AI, BI, psychoacoustics, poetry, cognition, software quality, literary fiction, transparency) and activations (www.knowlengr.com) from...

10 comments
jeff.allen
jeff.allen

It will be swamped by the nay-sayers and "leave it as it is" sayers... No thought for what inventions and utilizations are down the track, just waiting for more address space. Every motor vehicle should have an IP address and GPS. Why? Impending collision avoidance? Stolen car retrieval or disabling? "Panic" alarms that notify authorities where the car is? Yes, it's all available, but with more address space ALL vehicles could be so fitted. That's only one aspect.

marie.kennedy
marie.kennedy

Don't know about you guys, but we're busy enough keeping things working. I'm investigating and learning about it but no time for implementation at this point. That will come in time, no doubt.

jhoward
jhoward

IPV6 lags in maturity and availability because there is no business case to invest in it. Updating core switches and routers to be able to handle both IPV4 and IPV6 is costly, time consuming and has little to no foreseeable ROI. Without some monetary gain or threat the IPV6 initiative will remain stagnant. The subject does not need to be analyzed any further than the money. How many of you can say your IT department gets to do stuff that is a good idea but will eat up time and resources without any actual benefit within the next 5 years? So to reiterate - while this was interesting information - until the cost of change is outweighed by the urgency of change, the topic of IPV6 will remain stagnant.

pgit
pgit

Roadrunner broadband. Some of their techs have told me "we'll get to it when we get to it." (IPv6 implementation)

pgit
pgit

One man's "collision avoidance" is another man's "travel prevention." Every car with an IP will make possible requiring a "flight plan" for every trip you take in a car. So when this technology arrives you may have to justify your every move, beg for permission to drive for mere pleasure, and be punished for any deviation from the 'approved' movement. All technology has a good and a bad use. With history as a guide I'll put my money on the bad use being the primary use, even the impetus for deploying the technology in the first place, every time.

gechurch
gechurch

Those inventions sound great, but they don't exist now. So where's the payoff for investing a large amount of time and money into IPv6 now? There is none. Companies will continue to invest money in things that make them more money back, not in black holes with no business case. If killer apps like collision avoidance become a reality people will start moving to IPv6 then.

PeterM42
PeterM42

JHoward is absolutely right - most people with do nothing until it is ABSOLUTELY NECESSARY, then the IT department, etc. will need to implement it YESTERDAY!!! Ho hum, motivations in IT have only got worse in 40-odd years..

pgit
pgit

"a company that didn't fold in it's first year of operation?" How about "...been making it over the hurdles since 2005?" :) Would the opposite of a "mature startup" be an "immature dinosaur?"

xangpow
xangpow

I think most technicians will agree when i say this. When it comes to IT in the work place, no matter how cool, how secure, or how urgent the "new tech" is. The decision makers will always go back to the same question to decide on whether or not to implement it. How much will it cost? Then we have to explain that while yes it will be costly in the beginning it will save the company money later on down the road.

Editor's Picks