Nicolas Fischbach is Director Strategy, Architecture and Innovation. Nicolas has been with Colt for over 14 years starting out as an IP engineer. Before moving into this role he was in charge of Security Engineering and Operations. Nicolas holds an Engineer degree in Networking and Distributed Computing and is a recognised authority on Service Provider security as well as on next-generation architectures.
So what you might say? Well, it’s some sort of an achievement as it required a combination of a number of small and not so small things: my laptop was using its IPv6 stack, my web browser behaved properly, the local wireless LAN didn’t break it and handed it over to Colt’s network, it went across the Internet which didn’t drop it mysteriously until its final destination which was one of Facebook’s datacenter and the Facebook application which is also natively supporting IPv6.
Where did you do this from might be your next question. The origin isn’t my DSL line at home - as my broadband service provider doesn’t seem to be anywhere close to offer this as a standard feature – which is likely to be related to subpar IPv6 support in many Customer Premises Equipment (CPEs) and the difficulties related to upgrades or new roll-outs. And for many CIOs, IPv6 for internal business systems isn’t a key priority at the moment, as there are many other projects competing for attention and in reality, such an initiative isn’t at the top of the list. While there is no reason to be alarmed if IPv6 internal projects have not commenced, IPv6 shouldn’t be ignored, especially on externally facing platforms. Here’s the truth: I posted this from the v6 World Congress in Paris where Colt has been providing business Internet access allowing attendees to enjoy native, full spectrum, IPv4 and IPv6 connectivity for the last 3 years.
Your next question might be why did it take you so long? It’s true that another year has passed: first there was an April Fool’s, then we had the IPv6 World Day in 2011 followed by the IPv6 Launch Day in 2012 first to raise awareness and then to get as many interested parties to enable IPv6 going forward. While many people expect the Internet to “just work”, if you try to visualize the end-to-end path from my fingers tips typing “This is my first FB post over IPv6 from 2001:920:7000:101:119c:7faa:6c83:a3fa” to the database write operation on Facebook’s infrastructure when I hit “Enter”, there are many pieces of hardware and software involved along the chain that required upgrades and changes to support IPv6!
Finally, it’s clearly not the end of the journey and there’s still much more work to do to tidy things up. For example, two quite complex software components which affect the user experience are the operating system (including smartphone OSes) as well as the web browser. Most use a different approach to select between IPv6 and IPv4 to reach a remote system, which is likely to make the user experience non satisfactory and quite often non-transparent. This will clearly cause pain to many helpdesks trying to sort issues, which might not be easily repeatable. It’s also a way for me to call on our customers (and the IT industry at large) to be more engaged, try to “break” it and help this quite complex “beast” who is still in its infancy to become as predictable, reliable and dependable as the IPv4 Internet. We are getting there, so let’s keep the momentum going!
Please choose your country
3 comments on 'My first Facebook post... over IPv6' Add a new comment
Saurabh Bhardwaj says
25/03/2013 8:00:12 AM
Like this |
2 likes
Congrats Nico. Definitely another achievement/milestone for Colt towards IPv6 journey :-)
Reply to this comment
Simon Leinen says
09/04/2013 10:12:00 AM
Like this |
3 likes
Nice post. Here's how I look at this: Deploy IPv6. It will not be perfect, people will need to learn how to operate it, just like they had to for IPv4 in the day. You will also encounter specific new security risks, just like with IPv4 as people use new applications. You'll learn how to address these. As people use it (thanks to Google, Facebook etc.), your IPv6 deployment will get better over time. Meanwhile, IPv4 connectivity will continue to become less "predictable, reliable and dependable", as more and more work-arounds are needed because of the address shortage. So at some point, the lines will cross. The strategic question is when and how much we invest in getting our IPv6 deployments to a usable baseline state from where we can continually improve them, versus how long we continue to invest in these workarounds that keep IPv4 somewhat usable.
Reply to this comment
Nicolas Fischbach says
20/04/2013 9:17:55 AM
Like this |
2 likes
Replying to...
Thanks Simon. You raise some good points (e.g. I can read “avoid Carrier Grade NAT” between the lines) and there’s a follow-up blog in the queue that will cover those. Stay tuned!"Nice post. Here`s how I look at this: Deploy IPv6. It will not be perfect, people will need to learn"