Controllerless Networks

Hello, We are having issues with users being dropped in some of our IAP clusters. Cluster sizes range from 40-70 access points in a fairly dense layout. We have verified that coverage is fine and have performed some initial tuning the of RF environment to help clients roam more efficiently and not stick to an AP. The issue existed before this tuning was performed.

 

We are on the 6.5.1.5 code stream at the moment, but plan to upgrade very soon to 6.5.4.7. What I am curious about is how the IAP handles roaming. Does the cluster have a roaming cache that is shared amongst the access points that maintains the user session data as they roam or is it a hard roam every time they roam to a new IAP?

 

We use 802.1X  for internal users and an open SSID w/splash page and MAC caching via CPPM for customers. Issues occurs on both WLANs randomly. Switch ports and trunking has been verified etc. 

Roaming should be seamless within the cluster.

If you have roaming issues, I would consider lowering the arm maximum transmit power to encourage users to roan sooner.

Thanks. One of the firer things I did was to change the power settings so the access points could adjust as needed. I have a min of 12 and a max of 18 at the moment. 802.11k and v are enabled.’didn’t enable 802.11r since a lot of our devices are Apple mobile clients.

They seem to be roaming fine, we just have users at this one location that are getting kicked off for some reason. I have implemented these same settings at other locations and roaming behavior has improved but no drops. These are high traffic retail environments.

I’ve also got a TAC case open on this. I’ll post the solution if we are able to determine what the root cause is