MSM460 reboot

[PeteSahut]

Hello,

We have a lot of MSM460 access points (around 100) to which from time to time we have packet loss (95%-99% or 100%), which eventually is causing the AP to loose connectivity with the WLC and it's rebooting itself. This sometimes last for hours and then the next month there is not issue at all. When we ping other network devices, for example the core switch or the access switches we don't observe any packet loss. Also the APs are not connected to the same access switch but they're spread across the site and connected to lot of switches. When this happens only the APs are getting timed out and rebooted. The controller itself doesn't seems to have any issues, since they're other sites on that controller without any issue. (The WLC is not located on the site). The only one thing we observed is high broadcast traffic on the trunk link between the redundant core switches. 

Thanks for any help in advance.

[johnk3r]

What firmware Controller? Error Logs?

[parnassus]

A Network Topology diagram (related to your Wireless deployment) would be of help in understanding where the issue shows up (moreover considering you have a lot of Switches involved)...and knowing that your HP MSM7xx Controller and MSM APs are yet running a recent version (like the software version 6.6.4.0, as example) would also be benefical too (I mean: running updated software is a good starting point to diagnose what is happening so you have not to worry about issues already fixed).

[PeteSahut]

Unfortunately I don't have the exact software version on running on the WLC, but I did check the logs and there were no errors regarding this event (and the messages saying that the AP has been rebooted). We discovered a lot of broadcast traffic on the downlinks from the core. (Around 2k Frames/s peak on the interface). I will provide the software version of the WLC. Every floor has it's on VLAN. But it looks like it's related to the huge amount of broadcast traffic. We're not sure yet if the broadcast is coming from one source (I doubt that) or it is caused by the huge amount of devices connected in one VLAN. Btw. every floor has it's own VLAN. One possible solution would be limiting the broadcast traffic and then we will see if it helps.

ps: Regarding the network topology: It's happening on the whole site. The topology is simple. Distribution switches and downlinks to the access switches where the APs are connected. 

[parnassus]

OK, but each defined VLAN owns its Broadcast traffic (since the purpose of any VLAN is to define a broadcast domain), so...in which VLAN are you exactly observing this huge amount of broadcast traffic?