Friday, April 8, 2011

WGB acting up? Clean it up!

It's Friday and it's suppose to be the day to ease into the weekend.  Well I'm doing that by fixing problems with WGBs.  I  was pinged bright and early to take a look at a Cisco WGB that was having issues maintaining a stable connection to it's parent AP. 
Upon further inspection I noticed the WGB was roaming between several APs causing the device connected to it's Ethernet port to have a few issues.  The funny part is, is that this application controls the main gate to the facility, if the application's session is breaking then people aren't able to enter the facility to work

My first reaction was to use a command that I have used in the past to stabilize WGBs connections and keeping them from roaming.  On a Cisco Autonomous AP there is a command that can be applied to the radio interface that specifies the number allowed attempts the AP makes for sending a packet before giving up.  Optionally there is a command that can added that tells the AP to drop the packet rather then seek a new association when the max number is reached.
this command is as follows:

#conf t
(config)#interface dot11 0
(config-if)#packet retries 128 drop-packet

You can verify you command by doing a show run on the dot11 radio interface you applied the command to.  In my case it was the dot11 radio 0.

#sh running-config | begin interface Dot11
interface Dot11Radio0
 no ip address
 no ip route-cache
 !
 encryption mode ciphers aes-ccm
 !
 ssid lab
 !
 speed  basic-6.0 9.0 12.0 18.0 24.0 36.0 48.0 54.0
 packet retries 128 drop-packet
 station-role workgroup-bridge
 bridge-group 1
 bridge-group 1 subscriber-loop-control
 bridge-group 1 block-unknown-source
 no bridge-group 1 source-learning
 no bridge-group 1 unicast-flooding
 bridge-group 1 spanning-disabled

You have now successfully to the WGB to retry sending the data packet(frame) 128 times and when that value is reached, send it to the bit bucket rather than attempting to find a different AP.  Keep in mind this really doesn't address the root cause of the issue, the number of packet retries, but it help mitigate the effects.  To really fix that I would likely need to travel onsite and do some additional surveys and antenna adjustment.

There is another option that was pointed out to me that allows the WGB to be configured with parent APs.  This is not an option I have used or tested, so I plan to mock this up in my lab and write a short post about how to configure it.