Re: [WIRELESS-LAN] Wireless in Residence Halls

2012-12-19 Thread Karl Reuss
	We had the APs for our highrise dorms located in the hallways which was 
convenient for installation and maintenance, but it made for a poor RF 
design.  Lots of co-channel interference and dynamic power management 
problems.  Moving them into student rooms solved this, although 
coordinating the installation with the residents was a hassle.  We have 
not seen an increase in loss, damage or troubles.  Every once in a great 
while we get a user who is concerned with health aspects of having an AP 
in their room.


-Karl Reuss
 University of Maryland College Park

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.


Re: [WIRELESS-LAN] eduroam question(s)

2012-11-12 Thread Karl Reuss

On 11/12/2012 6:39 PM, Lee H Badman wrote:

Does anyone keep stats on how much your Eduroam efforts get used?
Like, other than just being in the club, is it really providing
benefits that an easy-to-use guest network wouldn't? Not being
snarky, but genuinely wondering.


We don't have any officially generated stats but a quick check of the 
numbers for this month shows we've had about 2000 traditional guests and 
500 eduroam guests.  The advantage eduroam guests had is that they were 
pre-approved before coming to campus and their devices were already 
setup.  Our guest system is a little clunky and could use some cleanup, 
but it will never just work like eduroam does for it's users.


We also get good feedback from our faculty and staff who visit other 
institutions, and that is hard to quantify with stats.  So far this 
month about 150 of our folks have authenticated at other eduroam sites.


-Karl

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.


Re: [WIRELESS-LAN] 4-channels in 2.4 GHz

2012-05-08 Thread Karl Reuss
	Cisco has a paper on this with some graphs showing energy overlap on 
4 channel deployments for both 802.11b and 802.11g.  The 802.11g OFDM 
signal seemed more prone to interference in a 4 channel setup so we 
stuck with 1,6,11.



http://www.cisco.com/en/US/docs/wireless/technology/channel/deployment/guide/Channel.html


-Karl Reuss
 University of Maryland, College Park

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.


Re: [WIRELESS-LAN] NAT in large scale wireless networks

2008-07-01 Thread Karl Reuss

Last academic year we ran NAT on our main wireless network.  We had
about 13,000 unique users per day and about 8,000 simultaneous
connections at peak times, roughly 95% student traffic. It worked,
but there were a couple of issues for us:

1) Picking the correct NAT box.   Catalysts 6500s do wirespeed
NAT, but they can't keep up with the number of new connections
per second.  A single ASA5550 handled the job well, now we have
a pair.

2) The NAT logs are enormous.  Finding space to keep them is
fun, going through them to find incidents is painful.


We did NAT because we added wireless to our dorms last year
and we weren't sure what the pace of our rollout would be, or how
fast the users would migrate over.  We didn't want to be shuffling
IP ranges all year.  We'll be going back to fixed IP addresses
next year for most wireless use.

-Karl Reuss
 University of Maryland, College Park




Michael Dickson wrote:
Though we currently have enough available routed IP space for our 
wireless clients we are looking toward the future and wondering if 
NAT-ing the wireless network makes sense.


Does anyone have any experiences, good or bad, using NAT for the 
wireless client pool in a large scale environment? What features go 
away (i.e. RFID or user tracking, etc.) Are there any gotchas?


We're an Aruba shop and expect about 3000+ wireless clients this 
semester and have been adding more APs by the week.


Thanks,
 Mike

***
Michael Dickson Phone: 413-545-9639
Network Analyst [EMAIL PROTECTED]
University of Massachusetts
Network Systems and Services
***

**
Participation and subscription information for this EDUCAUSE Constituent 
Group discussion list can be found at http://www.educause.edu/groups/.


**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.


Re: [WIRELESS-LAN] Monitor lightweight APs through NMS

2008-03-04 Thread Karl Reuss

Hector J Rios wrote:
The issue is that we want to see the APs in NMS with their names 
associated to an IP and a MAC address.


If your NMS supports ICMP you can just have it ping the APs.
That way the NMS will get basic up/down info for them.

Beyond that you'll need to do as Mike suggested and have your
NMS and controllers talk via SNMP.  The controllers have a fairly
rich MIB; I believe everything the WCS knows is obtained via SNMP
from the controllers.  Of course programming your NMS to obtain
all of this info is essentially recreating the WCS product.  It's
a shame the WCS can't generate some higher level traps from the
info obtains from the controllers.

-Karl

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.


Re: [WIRELESS-LAN] Authentication method comparison

2007-10-24 Thread Karl Reuss

Jonn Martell wrote:

Hi Donald,

You don't need to have AD to support PEAP.  Your RADIUS/LDAP
infrastructure does need to support MSCHAPv2 (aka native NT users and
domains).

Look how RADIATOR does it for a good off-the-shelf solution to
supporting PEAP on a non-Microsoft backend.



One thing to watch out for, however, is that most (all?)
non-Microsoft implementations of MSCHAPv2 require the radius
server to have access to the plain text user password.  Some
sites may have this, others not.  It depends on whether you
are running Kerberos, LDAP, etc.  Some LDAP implementations
will save it plain text, others (like ours!) do not.

-Karl

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.


Highrise dorm RF design

2007-03-27 Thread Karl Reuss

We're getting ready to expand our campus wireless
coverage into the dorms; full coverage for 12,000 students
over the next year.  The recent dorm discussions here have
been very helpful.

I'm wondering if anyone has experience with dense
AP deployments in traditional high-rise dorms.  About half
of our students live in these monsters.  8 floors, 250' straight
hallway down the middle of each, rooms on either side, block
walls, 70 users per floor.  Sort of like prison cells:)  Our
field guys and residential facilities folks would rather not
put the APs in student rooms, which basically just leaves the
hallways.  I'm worried about co-channel interference on the b/g
side.  6 or 7 APs down a hallway in clear sight of each other
will surely step on each other.  Loss through the floors only
seems to be 10db, which means we need to watch the vertical as
well.  Dropping power would only help a little, and at the
expense of room penetration.  External patch antennas are
one idea were looking at.  If anyone has any experience or
advice in this area they could share, I would be grateful!

Thanks,
-Karl Reuss
 University of Maryland, College Park

**
Participation and subscription information for this EDUCAUSE Constituent Group 
discussion list can be found at http://www.educause.edu/groups/.