I have noticed a slight problem regarding connectivity when roaming between access points.
Running a continuous ping to a random website (www.google.com), if I roam to another access point on my laptop, I will lose connectivity for 2-3 pings. Is this a normal thing? Once I gain connectivity again my IP address/gateway doesn't change, so I know the roam worked.. but I'm concerned about that packet loss.
I have aggressive load balancing enabled, though I'm not sure if that would be a problem at this time of year (low traffic levels).
The base of my concern is that we plan on having more VoIP functionality in the future, with the possibility of some wireless VoIP phones. The packet loss will become a big problem if this ends up going through.
Dropping a couple packets during a roam is normal but you should disable aggressive load-balancing to improve your roam times.
When aggressive load-balancing is enabled if the controller sees another AP in the area that is less loaded it will respond to the association request with a denial with code 17. If the client understands code 17, and most don't, it will try the other AP's. If the client doesn't understand the code then it will keep trying to associate to the same AP. After 3 attempts the controller wil allow the client on. In the latest versions of code aggressive load-balancing has been disabled by default.
1 - Upon roaming a new AP if the system gets an arp for a new ip address this can cause a delay and noticable gap in the voip traffic.
2 - If the system has a high encryption and security level then delay and audio loss can happen as the unit moves from access point to the other due to security re-negotiation.
3 - In the case of wpa2 the Cisco wireless system has a 1 second roam issue. See defect :
4 - Further problem occurr on the system when a unit walks out of coverage. For every packet not responded to by the hardware the Cisco Wireless controll send 64 packets. This affectivly creates a packet storm and can cause the network bandwidth to be congested.
"Development has looked at the sniffer trace for your original concern where there was what appeared to be an RTS storm after your device stopped responding. The conclusion was the controller/AP will continue to try to forward the wired side traffic to the client - for each new packet from the wired side there would be 8 data packet attempts followed by 56 rts attempts."
We do use WPA2, so it looks like it's point #3 that's applying here..
I looked up the bug, what are they referring to when they say it's fixed in version 12.4(3g)JA1? Is that an IOS for an wireless integrated 3750G switch? If so, I guess they haven't fixed it for a physical controller (WLC) yet.
IntroductionHow to use the Wireless LAN Controller Configuration Analyzer (WLCCA)
Javier Contreras is a Senior Tech Lead for the Wireless Business Unit in Cisco, with over 2 decades of experi...
< PRE >
(#)For this reason being that : - application that doesn't use multicast, sends one copy of each packet ( data unit of traffic at layer 3 ) to each client (" who seeks the traffic ).- application that does use multicast, sends ...
Transferring Crash file from standby:
Login to the Active WLC in HA.
(Cisco Controller) >transfer upload datatype crash
(Cisco Controller) >transfer upload filename <Desired filename>
(Cisco Controller) >transfer up...