Access Points not being associated with Controller

[tfbicking]

I have a 4400 series controller capable of 25 access points. I currently have 11 deployed and the controller is only seeing 10. Also if I have the need to reset a switch the AP is hanging from it can take days for the AP to find itself again, if it finds itself at all. Any Thoughts? I am new to this technology.

[Aaron Harrison]

Hi

Are you running your controller in Layer 2 or layer 3 mode?

Check that the native or access VLAN of the port you connect the AP to is in the same VLAN as the management interface of the controller. Try plugging a laptop into the same port and check you can ping the controller...

Regards

Aaron

Please rate helpful posts...

[Scott Fella]

DO you have one controller or multiple. When the AP is registered on the controller, make sure you configure the primary controller using the hostname and configure the secondary controller if you have another controller.

Is the AP and the Controller on the same subnet? If not, you might want to configure the switch port for the same vlan as the ap-manager and let the AP find the controller, then you can move it to a different subnet. Make sure you do not have a static IP configured though. Reset the AP if possible.

If you have issue with the AP registering with the controller right away especially on the same subnet, move the AP if possible to the same switch to stage the AP. If AP is mounted then maybe you have to troubleshoot connectivity from the switch port the ap is attached to, to the controller itself.

[tfbicking]

Hi,

Thanks for your reply. We have one controller. The controller and AP are on the same VLAN. We are using a static address for the controller. The controller is connected to a Catalyst 6513 on a 24 port Gig E blade. The port is set to 1 gig. I have tried placing the AP on the same blade in the 6513 but we are running PoE on the edge switches where the AP's are and the Blade will feed the needed power on the 6513 Blade. I just ordered an external power supply for the AP's to try your idea of connecting it to the same blade. I have tried recycling the AP by shuitting down the port in the remote buildings. Since running PoE that works well for recycling the AP.

Thanks,

Tom

[Scott Fella]

Since you have 10 of the 11 access points up and running and working properly, what I would do I'd take the access point that is giving you problems and reset it. then I woyld patch it in to a switch port that one of your 10 access points is plugged into and see if the access point works properly or not. you can also take a good access point and connect it to the switch port that the bad access point was connected to to verify that from that connection to the controller is good. just try to eliminated points of failure. it is weird to have 10 working fine and 1 that is not unless ap switch port or trunking back to the controller is configured wrong.

[jlwomeld]

try disabling CDP on the port where the AP in on.

[tfbicking]

Aaron,

Thanks for your reply. I have already tried the VLAN checks since I figured maybe I missed something. I am confused because they will work for a while but say we lose power and the switch dies, Once it comes up the AP takes a long time (Days sometimes) to reassociate to the controller. The configs on all switches have been saved so I know the configs are still correct after a power failure.

Tom

[Aaron Harrison]

Hi Tom

What config do you have on the controller port?

Normally I leave at Auto speed/duplex (not set to 1gb as I think you specified above). Also set it to a dot1q trunk (not negotiated, specifically configured) and make an unused VLAN the native VLAN. This makes all VLANs trunked and is more secure..

All your AP ports should be auto/auto in the WLC mgmt VLAN as mentioned before.

Do you have poor performance one the APs are working?

Aaron

Please rate helpful posts..

[tfbicking]

I double checked and it is set to Auto/Auto. It is connected to a Cisco Catalyst 6513 on a WS-X6454-ge-tx board. It's port is setup as a trunk but I am having difficulities setting it to a dot1q trunk. I was trying to use the "switchport trunk encapsulation dot1q" command but the only command that board will give me is a "switchport dot1q ethertype 600-FFFF" command. When I type in "switchport trunk ?" I get no encapulation only Allowed, Native and Pruning. I am using an unused VLAN for my trunk.I also double checked my ports my AP's are on and they too are set to Auto speed / duplex. As far as performance Since this is a new setup we have had limited users at a time testing but all have been getting about 54Mbps. Any other thoughts or guidance would be greatly appreciated. I have tried getting onto the Cisco site again this morning but it is DOG SLOW. Not sure what that is about. Everything else seems fine.

[jlwomeld]

check your port on the 6513 and make sure the enc match on both ports.