On 6/20/2015 11:32 PM, Randy Bush wrote:
My understanding is that the most recent NANOG had issues with clients
picking channels sequentially vs by signal strength. There may have
been other issues but when all devices use 149 because that's the
first they can and they get link that's not good.
If people know of tricks to solve this when there are 600-1000 devices
per room i am certain the NANOG eng team would love to know about it.
not really; they're in denial. why did san antonio work; the only nanog
in 4 or more which did? why does ietf work?
wireless is ugly. few know how to deploy at scale. it's just not easy.
randy
If people are curious what Cisco does for their 3x a year Cisco Live
events (last week in San Diego there was 35TB of data transferred over
that network), there's a panel discussion about how they deploy things
and what tools they use for it.
https://www.ciscolive.com/online/connect/sessionDetail.ww?SESSION_ID=76483&backBtn=true
That's the session from Milan 2014, may require a free account to view
the slides and video.
The session from San Diego is at
https://www.ciscolive.com/online/connect/sessionDetail.ww?SESSION_ID=83806&backBtn=true
Doesn't look like they've finalized the slides and video for that
session yet though.
In Milan they deployed 325 APs across 6 controllers (3 HA pairs). From
experience at the US Live events, there's 10-15K people in the main hall
during keynotes, there's probably close to 100 APs in that room alone
with the stadium antennas for the density needed. There's a LOT of
people trying to tweet during and this year periscope the keynote speeches.
If people are interested, I know a couple of the Cisco folks tend to
lurk on this and other lists and can probably provide more details if
asked nicely.
Jeremy "TheBrez" Bresley
b...@brezworks.com
