Session Drops and **RX stuck: issue1; workaround, HAL restart!

I'm having an issue on an ePMP1000 2.4 GHz AP where all or nearly all stations will drop with a minute or two.  During peak usage hours (6 PM to 11 PM), this will happen many times per hour.  During off hours (1 AM - 7 AM), it happens less, once every 3 to 5 hours.  This unit also shows the following message in the syslog:

**RX stuck: issue1; workaround, HAL restart! badkey_wbufdrop:2 rxstate_18:0 no_rxintr_count:0 rxeol:2 rxorn:0

I have 4 AP's in a frequency reuse arrangement.  This seems far worse on one sector than the other three.  I have 7 clients on this sector, with less on the others (5, 4 and 2 clients on others).

I thought this was an interference issue and have elimated any interfering wifi devices (client routers picked up by eDetect on the AP's).  This had little if any effect.   Anyone have any ideas or know what this error message means?  Please let me know if there is any other info I can provide to help diagnose.

Hi Chris,

Please contact me offline:

I will work with you to localize and debug that issue.

Thank you.