Error %DOT11-4-BA_FLUSH on 3502i

[patoberli]

Hi all

On at least one AP 3502i in lightweight mode I have a LOT of these errors:

*Mar 14 12:59:35.320: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40240 200
*Mar 14 13:01:41.045: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40160 200
*Mar 14 13:02:36.754: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40240 200
*Mar 14 13:03:47.812: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40240 200
*Mar 14 13:04:45.728: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40260 200
*Mar 14 13:06:59.262: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40260 200
*Mar 14 13:13:08.235: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with 40160 200
*Mar 14 13:15:14.943: %DOT11-4-BA_FLUSH: Client:[Here various Mac Addresses] ba flush with C0240 200

It's connected to a 5520 WLC running 8.2.141.0. 

The users complain about slow speed and connection problems on this accesspoint. Do you have an idea what this error message could mean?

[Leo Laohoo]

CSCut26137 &/or CSCvb82264

[patoberli]

Thanks. Those two bugs should be fixed in 8.2.141.0 though, which I'm running.

I believe now it's a hardware fault of this AP, it appears to only occur on one of around 60 and also several reboots don't help to (temporary) fix it.

[patoberli]

Hi Leo

I have another question though, do you know if I can somehow forward these AP logs to another server/syslog/wlc so that I have an actual chance to see if other AP have the same problem?

I have a configuration on the WLC to send its logs to my syslog, but I currently can't check it for other reasons if it already also contains any entries from my APs.

[Leo Laohoo]

Not sure if this function to log all AP logs to a syslog server exist.  

[patoberli]

After some searching I found a way described here:

https://supportforums.cisco.com/discussion/11958496/ap-syslog-configuration

I haven't tested it yet, because I have to check my syslog server first for the possible additional load.