Next NANOG

> That brings up an interesting question. Is it desirable to have a
> larger group?
> I don't have a simple answer to this, but it's harder and harder to
> get good interaction with a large group.

It's already hard-- anyone nervous of mikes (yeah there are a few in
this crowd) or just sitting far enough away from the aisles doesn't
speak up. More people makes both problems worse.

Well, we run into a problem there... NANOG meetings have, as much as
informative, been kind of a social event for those attending and I
think that this has become a very large reason many people come. With
a large group, that somewhat tends to go away.

Actually, what happens in large groups is that they splinter. You
still see your friends, but it's harder to meet new people outside
your company or your special area of interest. This is certainly an
issue for IETF for example, and seemed to be a factor at the one or
two big USENIX events I've been to.

The next answer for space is an arm of a convention
center. And frankly, thats just not what I want to see happen to this
thing.

Probably not, although small convention centers aren't necessarily
more hassle or less friendly than big hotels. Lack of a bar would be a
big problem though :).

IETF in Pittsburgh last summer and other similar venues-- convention
center for meeting space, attached hotel for bars and wireless-enabled
common areas-- seemed to work OK.

I would be in favor of Vegas.. its a reasonably convenient place with
good airport access. Of course, I think this is as much up to the
sponsor (if any has been found) as anyone.

How about New Orleans? No, wait, not in the summertime....

It's an oft-overlooked attribute of the location problem that a
willing sponsor is absolutely required and can be hard to find. The
expense and the hassle factor are not trivial, although to some degree
they can be traded off against each other, and the logistical
requirements (transit bandwidth, wireless coverage, a/v) have been
growing faster than the head count. Anyone who wants to dump on either
Merit or local hosts about logistics should probably start their
comments by volunteering to host themselves....or even just show up
early and provide help. Kids, do *not* try this at home, at least not
without plenty of bodies, patience, gear, and Tums.

I am all for San Diego. The same hotel as the IETF would work fine for me.

Bora

I am all for San Diego. The same hotel as the IETF would work fine for me.

this was an offer to host?

connectivity was "interesting" there. you need some ds-3 microwave gear
and roof access on a tall building downtown (not sure which one qualcom
used)

joelja

I wish I could Randy.

Ask me in two more years.

Bora

this was an offer to host?

I wish I could Randy.

my comment was a hint. i can be less subtle if necessary.

If I could get a straight answer on out-of-pocket costs involved, NAC may
be interested in hosting one in NYC/NJ.

Alex -

We did the one in Eugene for under 10k, but we're a University
with lots of personnel in place - and student bodies to throw
at the thing.

Our big costs were US West curcuits into the hotel (6xT1 - we could have
gotten away with 4xT1) and cable - we had lenghts cut to fit the table
layout in the ballroom. We had switches & such on hand, and we "borrowed"
terminal room machines from one of the student labs -

Everyones starting resource base is different, so take this with
a grain of salt -

My notes from our event can be found here:

http://pythia.uoregon.edu/~llynch/hosting.html
http://pythia.uoregon.edu/~llynch/nanog16.html

This was a great event for the UO - and a LOT of work, as I'm
sure any of the other recent hosts could tell you.

Lucy E. Lynch Academic User Services
Computing Center University of Oregon
llynch@darkwing.uoregon.edu (541) 346-1774
Cell: (541) 912-7998 5419127998@mobile.att.net

If these costs were negligible, how much would it have cost?

(assuming, for example, an 802.11b shot from a hotel to an already-
connected nearby building, donated transit, doing wireless-only in the
conference and having some friendly vendor loan the machines for the
terminal room).

Joe

> Our big costs were US West curcuits into the hotel (6xT1 - we could have
> gotten away with 4xT1) and cable - we had lenghts cut to fit the table
> layout in the ballroom. We had switches & such on hand, and we "borrowed"
> terminal room machines from one of the student labs -

If these costs were negligible, how much would it have cost?

(assuming, for example, an 802.11b shot from a hotel to an already-
connected nearby building, donated transit,

we actually tried to do a wireless run as a backup plan, but couldn't find
an open conduit, and the hotel balked at the thought of our wiring guys
coring 11 floors in order to get to the roof...

I don't think nanog is quite ready to go wireless only in the meeting room
although cutting down on the wired infrastructure deployed is something
that's been worked on...

Well - we paid for stuff like shipping (gear), power strips, gaffer tape,
video tapes, a rental truck, volunteer munchies, and a lot of misc.

Verio picked up a big chunk of our costs (Thanks Randy!) -

If you've got gear, local loop, & transit you might be able to
do a meeting for under $3,500. (modulo a blues band - which
SHOULD be a requirement)

Lucy E. Lynch Academic User Services
Computing Center University of Oregon
llynch@darkwing.uoregon.edu (541) 346-1774
Cell: (541) 912-7998 5419127998@mobile.att.net

I'd rather see NANOG/MERIT invest in like 15 or 20 base stations, rather
than wasting the time/money on the cat5 cable and switches -- etc.

A wireless card costs all of $200 these days... everyone could just get
one.

Merit actually has enough access points (it really only takes 4-6 of
them... it's more of a user mindshare issue... hence the continuing
availability of laptop drops... note however that there are fewer at the
recent meeting (ie. we had almost 300 for one room at nanog 17).

the cisco folks used about 24 access points(overkill) for the ietf in
sandiego that was enough to blanket the conference center and the hotel
below the 3rd floor...

joelja

As long as they keep providing power sockets in the rooms for those of us who have laptops that can't survive 4 hours straight on
battery.

Thomas

the more access points, the more aggregate bandwidth you will get on the
802.11 cloud...

well... yeah but if your link is a couple multiplexed t-1's the difference
between say 4 and 10 access points may not be all that great...

joelja

Also, it would be nice if the hotel had ethernet in the rooms
(in-house DSL or whatnot) or wireless sufficent to cover the hotel.

  If the conference moves to larger and larger hotels the wireless
covering the whole hotel isn't as possible.

  The IETF-49 density of access points was quite nice

  - Jared

At this last NANOG, it seemed that the geometry, rather than just the size,
of the hotel made covering even a small portion of rooms extremely
difficult. The bar coverage was certainly nice. Interestingly, the Merit
folks had far more wireless access points than they actually used at this
last NANOG. The biggest problem was getting wired connectivity to all the
places that access points were needed. The Sheraton was a very old hotel
with primitive wiring. This may be easier at hotels of more recent vintage.
Which may be a lesson learned for future hosts.

- Daniel Golding