That is it, Cisco Live Milan is over! The “before” of anticipation, seemingly a moment ago, is replaced by the “after” of takeaways and accomplishments for the week. Some passed their CCIE certification, some met a new business partner, but it’s likely that everyone learned something new. I am not an exception.
During the week, I presented at two different technical breakout sessions (BRKRST-2304 – Hitchhiker’s Guide to Troubleshooting IPv6 and BRKEWN-2666 – IPv6 on WiFi: You talk too much! NOT anymore) and spent my remaining time working with everyone in the Network Operations Center (NOC) to ensure that IPv6 is a smooth ride for all of the 9,000+ devices on the network. Not only did I learn a lot, but this year at Cisco Live Milan was a year of “firsts” for me.
- For starters, it was the first time I shared details about my experience with large-scale IPv6 WiFi setups with Cisco Live attendees in the form of a breakout session. After talking with attendees, my main takeaway – this blend of the technologies is indeed an uncertain area for many.
- It was also the first time an IPv6-only SSID was advertised venue-wide. I was very conservative about publicizing the access information for it, so the peak number of simultaneous users is not a big number to brag about: 60. However, what I consider to be a more important milestone was that this SSID was treated like a grown-up; allowing attendees to experiment with completely switching off IPv4 without sacrificing the coverage. I hope that the other Cisco Live events continue the tradition and allow for a parallel IPv6-only WiFi experience – we’ve proved it works well and does not affect the mainstream SSID.
- Another “first” was using the “not on-link” prefix to allow the host-to-host communication while controlling the number of ND cache table entries. It worked perfectly! As a result, I am now recommending it as a default for this kind of event. Interested in what it is? Download the slides from BRKEWN-2666.
- The last and probably the most pleasant “first” was the percentage of the dual stack hosts. It was the first time that the curve did not dip with the beginning of the main event. See for yourself here and compare it with the 2013 stats!
IPv6 is becoming more and more widespread. For more IPv6 insight at Cisco Live Milan, read Steve’s latest blog about IPv6 enabled demos in the World of Solutions. So when is the right time to start turning off IPv4? Can’t we try it already?
Coincidentally, that’s exactly what our friends did at the FOSDEM conference in Brussels that took place right after Cisco Live Milan. But these details are a separate topic for another blog post – stay tuned!
Hello Andrew,
do you have IPv4/IPv6 gateway traffic stats as last year? I see only total traffic on north and south gateway.
Matej
Matej,
alas – I did not poll the v4/v6 stats myself – since the guys were running NetFlow which gave much better granularity…
If I remember well from their stats, the v6 traffic was in the ballpark of 5%. But I must admit to being lazy and not doing IPv6 DNS server assignment via DHCPv6 this year (the “interesting” part was the wifi scalability optimizations stuff) – so I think some clients might have decided to prefer IPv4.
–a
Some very interesting deployment tips. Thanks for sharing..
Hi Andrew,
thanks for sharing.
I was searching for your IPv6 in WiFi Slides directly on ciscolive.com but the search did not come up with any results (searching by “IPv6” and BRKEWN-2666). The direct link in your blog post works however.
I dont know if this is the right “spot” for this info, but i thought I let you know.
Cheers,
Chris
Hi Chris,
thanks a lot! I did notice myself that 2666 is missing on the ciscolive365.com. Normally it takes some time before the slides/talks appear – maybe later.
Meantime, I’ve also submitted an IETF draft around IPv6+WiFi, for the upcoming meeting in London. You might find it useful (and I would love to hear any comments on it btw!).
The link is:
http://tools.ietf.org/html/draft-yourtchenko-colitti-nd-reduce-multicast-00
Note that the last part proposing small changes in the protocols is not something is doable (yet 🙂 ) in the real-world networks.
–a