We have been noticing a problem in every version of 2.9 with wireless where a single client with a bad signal disconnecting and reconnecting raises the CPU on ALL APs connected to the switch and makes the APs and CPE unpingable for about 30 seconds at a time.
We are using RB532s for APs and different RBs for clients. (112, 153, 133, 532, etc) The APs are setup as AP Bridge and the CPE are setup as Station WDS.
I’m sure it’s a WDS problem but I don’t know how to stop it. Any ideas?
Have you installed a STP stop on your WDS network?
I would NEVER use WDS, unless there’s no other way, and if speed and low latency is not important at all.
If you have problems with clients disconnecting due to bad signal, you could consider changing your Ack. Timeout (Acknowledgment Timeout) to a value between 50 and 450. I find 380 is a good value in very dodgy areas. But, playing with Ack settings could cause other problems, only use as last resort.