Thank you Matt, I appreciate the feedback and may want to get more of your Meru experiences offline. A 5GHz RSSI (PHY) survey seems to be the common denominator for legacy and .11n clients. Its likely this provides adequate coverage for 2.4GHz clients. In fact it may be overkill for 2.4GHz, given the better penetration. Assuming equitable power levels (some vendors are more strict than others when it comes to 5GHz max power levels with non-captured antennas) equal cell sizing can be approximated. Do you happen to know if Meru has any power limits in 5GHz for their APs? Bruce T. Johnson | Network Engineer | Partners Healthcare Network Engineering | 617.726.9662 | Pager: 31633 | bjohns...@partners.org
________________________________ From: The EDUCAUSE Wireless Issues Constituent Group Listserv on behalf of Barber, Matt Sent: Thu 1/29/2009 9:20 AM To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU Subject: Re: [WIRELESS-LAN] 802.11n testplans Hi Bruce, We didn't have a formal test plan, but have had many experiences I am more than willing to share. Surveying was pretty interesting, as we deployed before there were any 11n capable tools available. Back in the summer of 2007, we pretty much just had to make some assumptions and then survey with what we had. Our goal was for full 5 GHz coverage, but without knowing exactly how the 5 GHz 11n coverage was going to look, we surveyed and deployed for 11a. We made the incredibly safe assumption that 11n coverage would be equal to or greater than 11a. The end result was a pretty dense environment all around. Since we deployed Meru single-channel, the overlapping AP coverage helps as opposed to hinders our deployment. This may not be the case with other vendors, but I don't have any personal experience with anything else. This approach left legacy clients covered just fine. In the summer of 2008 we had a chance to use the new version of Ekahau to do some testing of 3x3 vs 2x2 antenna configurations. We have been running on 2x2 with normal 802.3af power since we deployed in October 2007. We found that bumping up to 3x3 significantly improved the data rates for clients at further distances. The difference was enough that we went ahead and got 802.3at (assuming the standard gets all wrapped up) injectors. In terms of considering legacy clients for deployments, it may be useful to see how legacy clients behave with an 11n AP at 3x3. If you survey and deploy for full coverage at 5GHz with 3x3, 11g clients may end up fully covered anyways. If I were to do a new deployment today, that is how I would survey. Depending on your client mix, you may be able to even deal with only "decent" 11g coverage as the number of 11n clients grows. I hope this helps. I would love to hear how 11n deployments and surveying are going for the group at large. Is everyone still surveying based on legacy clients, or do 11g clients end up working fine if you target 5 GHz 11n? Matt Barber Network Analyst Morrisville State College 315-684-6053 -----Original Message----- From: The EDUCAUSE Wireless Issues Constituent Group Listserv [mailto:wireless-...@listserv.educause.edu] On Behalf Of Johnson, Bruce T Sent: Wednesday, January 28, 2009 11:36 AM To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU Subject: [WIRELESS-LAN] 802.11n testplans Toivo et al, Great comments. Does anyone have any 802.11n testplans they are willing to share? 802.11n Survey experiences? Has it turned the traditional survey methodology on its head, or do we still have to consider legacy and so the "n" simply stands for "Nice (if you have it)." Anyone with experience with the Ixia WLAN Test suite? Does it have 802.11n capability? Thanks all, Bruce T. Johnson | Network Engineer | Partners Healthcare Network Engineering | 617.726.9662 | Pager: 31633 | bjohns...@partners.org ________________________________ From: The EDUCAUSE Wireless Issues Constituent Group Listserv on behalf of Toivo Voll Sent: Wed 1/28/2009 9:48 AM To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU Subject: Re: [WIRELESS-LAN] Comments about Aruba and Cisco???? Some tests we found worthwhile: -Check to see if multicast works like you expect. -Related to multicast and in general, check to see if fragmentation also leads to reordering of fragments and if your applications can live with this. -Test client throughput in various scenarios (Single client, multiple clients, multiple clients some of which are legacy, bonded N channels vs. unbonded, as many client cards as possible) and with varying number of TCP streams per client. In particular with 802.11n the throughput behavior between Aruba and Cisco was quite different depending on the number of concurrent streams a client was sending / receiving. -Test WPA2 authentication with whatever authentication backend you wish to use, including roaming between APs. Unless you get several controllers, you may not be able to see whether the hand-off between APs on different controllers introduces longer delays. -Run some customer support scenarios trying to find out whether a client is working right, seeing what might be the cause for bad performance, and look at logging of information within the various systems. -You didn't mention the scale of your deployment, but see what additional pieces you might need to go full-scale, such as how many APs/Controllers one WCS box can handle before you need several and Navigator. I'm not sure what the equivalent in Aruba parlance is. -You mentioned you're looking at the 1200 series (our new Ciscos are 1142s) but also look at mounting and physical security options as well as harmonious life with your Friendly Fire Marshall on your gear in regards to plenum issues. -If you are planning to use PoE gear in a mixed-vendor environment, test the behavior of that as well. You'd think this would be easy-peasy but we didn't find this to necessarily be the case. -If you're using rogue detection features, see whether the alerts are valid, and in a case of multiple rogues you'd like to contain whether you can correctly un-contain some or add new rogues to the containment list. -Test for controller failures and AP behavior -- also make sure to see what happens when the downed controller is brought back. -- Toivo Voll Network Administrator Information Technology Communications University of South Florida On Tue, Jan 27, 2009 at 8:59 PM, Johnson, Ken <ken.john...@med.fsu.edu> wrote: > All, > > I am a member of an evaluation team at Florida State University considering > Cisco and Aruba wireless products. We are focusing on LWAPs and controllers. > For evaluation configuration and pricing purposes, we have requested from > the companies information and pricing relating to configurations with 128 > and 1200 APs. The Aruba LWAP is the AP125 while Cisco LWAP is the recently > release 1142. The Aruba controller is the M3 while the Cisco product is the > WiSM. There are other aspects, too. I know many of you have experience with > Cisco and Aruba and have gone through similar experiences. I am interested > in learning about any observations and experiences you have that we should > consider in our efforts. Please send me your thoughts. > > Thanks. > > Ken > > ~~~~~~~~~~~~~~~~~~~~~~ > > Ken Johnson > > Director, Information Technology > > FSU College of Medicine > > 1115 Call Street > > Tallahassee, FL 32306-4300 > > e-mail: ken.john...@med.fsu.edu > > phone: 850.644.9396 > > cell: 850.443.7300 > > fax: 850.644.5584 > > > > "Please note: Florida has very broad public records laws. > > Most written communications to or from state/university > > employees and students are public records and available > > to the public and media upon request. Your e-mail > > communications may therefore be subject to public disclosure." > > > > ********** 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/. The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail. ********** 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/. ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.
