One curious note I saw today between two Macs- one was definitely using short guard interval as configured on the AP, along with wide-channels and no legacy mojo to get to 300 Mbps stated data rate. But- the other would top put at 270- would not use SGI. As far as I can tell, there's no difference between the client machines, and there is nothing to set on the Mac... Going against an Aruba test environment.
Curious. -----Original Message----- From: The EDUCAUSE Wireless Issues Constituent Group Listserv on behalf of Jeffrey Sessler Sent: Wed 3/4/2009 4:59 PM To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU Subject: Re: [WIRELESS-LAN] Mac OSX and 5Ghz Concerning the channel 161 issue... While not specific to channel 161, there is an issue with the broadcom chipset as installed in Apple and other products. The Cisco unified AP's broadcast a "world mode" information item that the client should use to determine power level. In the case of the broadcom chips/driver, when it sees this information item in the beacon, it causes the driver to set the client power levels incorrectly (like at zero or bouncing). Lower channels seem to do better than higher, thus why channel 161 seems to have issues. There is currently no way to disable the world mode IE in unified, but cicso is working on it. I have new AP code that disables it, and it does fix the broadcom issues in my Macs. Broadcom is also working on a driver update, but who knows how long it's going to take before it shows up and clients update. best, Jeff >>> James Nesbitt <n...@duke.edu> 3/4/2009 12:23 PM >>> David, In your output, the channel reading does not indicate bonding (channel number followed by ,1 for above or ,-1 for below). Also, the SNR listed in this output is excellent, this client should have an MCS data rate of 14 or 15. Try changing the AP channel to anything but 161. I have been seeing some strange issues with Mac clients and at this point the only common thread is channel 161. I don't know if Apple is secretly doing something with channel 161 or what. Maybe to enhance the speed for Apple to Apple ad-hoc. In the couple of instances that I have seen this the issue cleared up when I changed the channel. James Nesbitt Duke University On Mar 4, 2009, at 1:10 PM, David Wang wrote: > Thanks James. Here is my output: > > ccs-nss-macbook:~ nsteam$ /System/Library/PrivateFrameworks/ > Apple80211.framework/Versions/Current/Resources/airport -I > agrCtlRSSI: -55 > agrExtRSSI: 0 > agrCtlNoise: -95 > agrExtNoise: 0 > state: running > op mode: station > lastTxRate: 54 > maxRate: 54 > lastAssocStatus: 0 > 802.11 auth: open > link auth: wpa2 > BSSID: 0:22:90:92:9c:be > SSID: uog-wifi-secure > MCS: -1 > channel: 161 > > Any idea? And as Jeff suggested, WMM is enabled, all WCS rates are > enabled, channel is indeed binding. And the windows client is > definitely showing 240M on 802.11n with same AP. > > David Wang Networking Services, CCS > www.uoguelph.ca 519-824-4120 ext 52046 > > On 4-Mar-09, at 11:00 AM, James Nesbitt wrote: > >> Confirm what MCS rate you are connecting at, signal strength, and >> noise. Is your Mac client really bonding (should have a channel >> number with + or -1)? Make sure all of your data rates are enabled >> for the A band as well. >> >> user-111-123-111-1:~ $ /System/Library/PrivateFrameworks/ >> Apple80211.framework/Versions/Current/Resources/airport -I >> agrCtlRSSI: -59 >> agrExtRSSI: -66 >> agrCtlNoise: -92 >> agrExtNoise: -92 >> state: running >> op mode: station >> lastTxRate: 300 >> maxRate: 270 >> lastAssocStatus: 0 >> 802.11 auth: open >> link auth: unknown >> BSSID: 0:1f:9e:8d:70:fe >> SSID: test >> MCS: 15 >> channel: 44,1 >> >> James Nesbitt >> Duke University >> >> On Mar 4, 2009, at 10:23 AM, David Wang wrote: >> >>> I am testing Mac OS X 10.5.6 with cisco 1140 a/b/g/n APs, but only >>> see the AirPort's speed up to 78M with 802.11n once, most time it >>> stuck at 54M of 802.11a. Another Linksys 802.11n card can easily >>> up to 240M with 40M channel on 802.11a/n. Do I miss something? >>> >>> David Wang Networking Services, CCS >>> www.uoguelph.ca 519-824-4120 ext 52046 >>> >>> On 4-Mar-09, at 9:00 AM, David Wang wrote: >>> >>>> Looks like Apple is listening: >>>> http://support.apple.com/downloads/AirPort_Client_Update_2009_001 >>>> >>>> About AirPort Client Update 2009-001 >>>> This update is recommended for all Intel-based Macintosh >>>> computers running Mac OS X 10.5.6. >>>> >>>> It addresses issues with roaming and network selection in dual- >>>> band environments. >>>> >>>> >>>> David Wang Networking Services, CCS >>>> www.uoguelph.ca 519-824-4120 ext 52046 >>>> >>>> On 23-Feb-09, at 8:53 PM, Lee H Badman wrote: >>>> >>>>> William, >>>>> >>>>> I have no answers, but would encourage you to vigorously engage >>>>> your >>>>> local rep and ask for an ear in Apple's development. It took us >>>>> a while, >>>>> but feels like they are starting to get that their products are >>>>> actually >>>>> less than perfect on the WLAN. >>>>> >>>>> -Lee >>>>> >>>>> Lee H. Badman >>>>> Wireless/Network Engineer >>>>> Information Technology and Services >>>>> Syracuse University >>>>> 315 443-3003 >>>>> -----Original Message----- >>>>> From: The EDUCAUSE Wireless Issues Constituent Group Listserv >>>>> [mailto:wireless-...@listserv.educause.edu] On Behalf Of William >>>>> Green >>>>> Sent: Monday, February 23, 2009 7:56 PM >>>>> To: WIRELESS-LAN@LISTSERV.EDUCAUSE.EDU >>>>> Subject: [WIRELESS-LAN] Mac OSX and 5Ghz >>>>> >>>>> We continue to experience problems with Mac OSX devices preferring >>>>> 2.4GHz (g) networks instead of 5GHz (a/n). We want devices at >>>>> 5GHz >>>>> whenever possible. From the little we can tell the pllst >>>>> (/Library/Preferences/SystemConfiguration/ >>>>> com.apple.airport.preferences. >>>>> plist) >>>>> appears to be remembering channels seen across our campus-wide >>>>> SSID, >>>>> and then chooses 2.4GHz for whatever reasons. If you delete the >>>>> slower channels in the plist and don't change APs, it stays on the >>>>> 5GHz channel. Unfortunately our wireless infrastructure is mixed >>>>> (lots of g only) so a user will quickly get a 2.4GHz channel in >>>>> their >>>>> plist which sticks them in 2.4GHz again. >>>>> >>>>> Does anyone have additional information about this problem? >>>>> Work-arounds (on a campus scale)? Dates for fixes? >>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> William C. Green e-mail: gr...@mail.utexas.edu >>>>> Director, Networking phone: +1 512-475-9295 >>>>> ITS (Information Technology Services) fax: +1 512-471-2449 >>>>> University of Texas >>>>> 1 University Station Stop C3800 >>>>> Austin, TX 78712 >>>>> >>>>> ********** >>>>> 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/ >>>> . >>>> >>> >>> ********** 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/ > . > ********** 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/.
