Greetings, That makes for an interesting read especially the client hand-off measurement. just two observations... The low dis/re-association rate associated with the apple clients is described as "good", and in the sense that it reduces churn I agree with that. But if in fact it means you continue to remain associated with an ap when another is available which is more viable, the user may experience poor performance. With regard to observations about the site. There were two sets of cisco ap's deployed, the hotel's were remotely managed by an offsite contrator (we had snmp read access and that's it), the others were managed by the noc staff using cisco wlse and other tools. Because the hotel's ap's were all configured for channel 6 (a questionable practice I agree) our map coloring problem was greatly simplified (if you saw a ch6 ap you knew whose it was). Each of the chantilly ballrooms had 4 hotel ap's in the ceiling, additional hotel ap's were in they foyer and in the boardrooms immediately upstairs. Two additional ap's (of our's) were located in chantilly west during the thursday plenary that were not present on wenesday, one located in the center of the south half of the room and one located in the center of the north half, signal from these ap's should have been highly attenuated during the meeting as they were located under chairs. these ap's were deployed due to the high number of associations with ap's at the north end of the room despite our efforts at attenuation. an example can be seen here: http://www.flickr.com/photos/64986983@N00/148888492/ Overall the existing hotel ap's were both a liability (hard to manage) and a serious asset in that they covered the lobby and the attrium's in a building that was approximately 550 meters in length. having them reduced the ap count that we had to come up with by order of 40-60 ap's and meant that ap's were already deployed in some fairly hard to reach places. Andrea G Forte wrote: > Dear all, > > as some of you may know, we have conducted some wireless measurements at > the 65th IETF in Dallas. Based on previous feedback, we were looking for > some anomalies in the wireless network. In our measurements we found > different behaviors for different wireless card vendors. > This is the link to the paper is: > http://www1.cs.columbia.edu/~andreaf/downloads/ietf_measures.pdf > > Regards, > -Andrea > > > _______________________________________________ > Ietf mailing list > Ietf@xxxxxxxx > https://www1.ietf.org/mailman/listinfo/ietf -- ------------------------------------------------------------------------ Joel Jaeggli Unix Consulting joelja@xxxxxxxxxxx GPG Key Fingerprint: 5C6E 0104 BAF0 40B0 5BD3 C38B F000 35AB B67F 56B2 _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf