perhaps as an educational exercise in network troubleshooting whoever is
operating the meeting network could explain what the frack is wrong with
the meeting network, how it is being debugged, and what they have
learned about the cause of the suckage.
randy
if it's wifi that's causing the trouble, the usual causes are:
- insufficient density of APs for the number of clients
- APs configured with TX too high (should be set as low as possible)
- APs configured to accept dot11b <= 9 megs
- APs configured to use auto channel selection
- stupid broken clients screaming at high volume across the room to APs
which are impossibly far away
There is a more fundamental problem, though: wifi was not designed with
crazyass density in mind.
Bring back UTP?
Nick
I'm not seeing the problem, but have heard one other person say they are having trouble.
Perhaps some details of the problem you are seeing would help diagnose the troubles as there are many of us who are not seeing it. I am using the 'NANOG-a' network without trouble.
- Jared
Don't forget RFI and various forms of spoofing used for MITM.
;>
is the complaint the hotel ROOM wireless? or the meeting-room? I
noticed the nanog-a-secure bounce me 2x, so I moved back to
ipsec-tunnel on nanog-a.. in the past nanog (plain) has been more
'stable' for me in general (and all you mac users can happily fight
over -a!)
As to the hotel room wifi... apparently when you have 490 rooms in the
hotel (full) and only provision your internal NAT space as a /23 ...
things work 'fine' most days. When a networking conference comes to
visit with 3+ devices requiring IP in each room... the whole hotel
network stops 
Last night the display systems in the lobby and the
hotel registration machines were all broken The hotel's network
people (in NYC) are supposedly 'on a fix', who knows... (is expanding
the nat subnet THAT hard?)
-chris
if it's wifi that's causing the trouble, the usual causes are:
is the complaint the hotel ROOM wireless? or the meeting-room?
meeting net, a-secure and a. really bad during the night, but still
bouncing up until 08:30 when i turned laptop off to participate in
breakfast.
and conjecturbation as to what the problem was is amusing at best. i
asked for actual diagnosis from whover is running the net.
randy
Problem for me at least has not been the MAC layer (either hotel room
or meeting room), it was that the DHCP server was not responding.
Ironically, I could still see everyone's Bonjour and SMB service
advertisements.
--Richard
Sigh, if only there were people somewhere near the hotel who knew how to
configure this sort of stuff...
Nick
It would be wise for NANOG to approach future venues and specifically discuss these things with the hotel IT departments in question ahead of time so that they have some remote chance of being prepared.
Owen
On the hotel network, I have also seen some issues beyond getting an address. I can usually trace just fine, but applications, specifically web is extremely slow, or non responsive. The hotel appears to be shoving all traffic through a squid proxy, which does not appear to be big enough to handle the traffic. I have gotten various error messages from squid.
I would think that the contract with the hotel for the conference would include the specific requirements for the network. Is that not the case?
-Randy
I would think that the contract with the hotel for the conference
would include the specific requirements for the network. Is that not
the case?
underlying problems
o no hotel believe that we'll actually be significantly high use.
they simply can not conceive of it. ietf, apricot, ... have
seen this time and time again
o the hotel does not manage the network, so you have two comms hops
to anyone who can do anything. and anyway, they are not going to
provision more bandwidth
but the problems of which i spoke were the meeting network. which we
do supposedly control.
randy
I am noticing far worse performance and reliability with IPv6 as opposed to IPv4. For a good 10 minutes until just now there was *no* IPv6 routing even to the first hop, but IPv4 was still "working." Now things are merely slow, not broken.
I also got bounced a few times from -secure and -a and when I got back I couldn't get an address (IPv4 or IPv6) for some time.
Conjecturbation: They're using an old proteon router and someone is exhausting its arp and ndp caches.
michael
Conjecturbation: They're using an old proteon router and someone is
exhausting its arp and ndp caches.
i wanna picture!
VPN traffic was also slow / bursty. So I guess there's some capacity issues
as well as layer 7 cruft.
On the hotel network, I have also seen some issues beyond getting an
address. I can usually trace just fine, but applications, specifically web
is extremely slow, or non responsive. The hotel appears to be shoving all
traffic through a squid proxy, which does not appear to be big enough to
handle the traffic. I have gotten various error messages from squid.
I would think that the contract with the hotel for the conference would
include the specific requirements for the network. Is that not the case?
-Randy
o no hotel believe that we'll actually be significantly high use.
they simply can not conceive of it. ietf, apricot, ... have
seen this time and time again
WEG] this is a problem that is quite solvable via the careful application of real data from past events
I assume most of these conferences can track number of unique devices seen (by MAC address) peak and total, show peak and average network usage BW graphs, etc.
Wes George
This E-mail and any of its attachments may contain Time Warner Cable proprietary information, which is privileged, confidential, or subject to copyright belonging to Time Warner Cable. This E-mail is intended solely for the use of the individual or entity to which it is addressed. If you are not the intended recipient of this E-mail, you are hereby notified that any dissemination, distribution, copying, or action taken in relation to the contents of and attachments to this E-mail is strictly prohibited and may be unlawful. If you have received this E-mail in error, please notify the sender immediately and permanently delete the original and any copy of this E-mail and any printout.
To be fair, that's not a hotel-only problem. We've seen that problem within
the IT industry. Actual discussion with a vendor who wanted to analyze our
logs so they could size a solution:
"Send us a day's worth of logs" "OK"
...
"We said a *day's* worth, not a *week*". "That *was* a day"
"Wow, that logfile was huge, we didn't think anybody actually did that much traffic a day..."
The sad part was that the vendor in question *really* should have known better,
we're pretty sure they targeted their solution at many sites bigger than us.
Or maybe the other sites are bigger, but we pound the bejeebers out of stuff. I dunno.
My experiences planning and operating ARIN meeting networks taught me that it is difficult if not impossible to get a hotel make any changes to in room based wireless or wired networking. Often times they have contracts with third parties and don't actually have any control over how the network operates or issues such as capacity. This can also include an inability to disable access points or otherwise limit the ability of the contracted service provider to provide access to rooms.
In room access is always something that we looked at when judging potential venues but the poor state of most in room Internet access infrastructure and the realities of existing business relationships made this a nice to have rather than a hard requirement.
Then the RFP for the meeting needs to be more specific with some basic SLAs
that result in a smaller bill if not met.
Frank
In my historical knowledge of this: there are only so many venues that can have 500-650 people and fit.
Jared Mauch
I have been at other conference that have triple or more participants, and it has never been anything close to the issues we are having at this hotel. Slightly slower performance is expected. Completely not working is not.
-Randy