cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
435
Views
0
Helpful
4
Replies

WLC4404 failover doesn't return properly

robinwenham
Level 1
Level 1

I have the following scenario:

WLC01 on 172.16.72.200

AP on 172.16.72.4 (DHCP)

WLC02 on 172.16.32.200 (different subnet)

I have DHCP servers on each subnet, and routing between them is operational.

The two WLCs are on the same mobility group and they can see each other.

WLC01 is the primary controller for the AP, WLC02 is the secondary controller for the AP (this provides failover over a WAN).

I am not running the WLANs or APs in H-REAP, they are in fully tunnelled mode.

In my lab scenario this works fine but with a wrinkle as detailed below:

When I down the WLC01 controller, as expected the AP reboots and attaches to the WLC02 controller (the backup).

The client gets a DHCP address from the backup LAN - traffic works fine.

When I restore the WLC01 primary controller the AP re-attaches fine to that, but the client does something strange:

1) it retains the IP address of the backup LAN and retains its association with that controller. There are no APs associated with this controller.

2) it also associates with the primary controller but through the backup controller

- in this cases a look at the client properties on the backup controller shows that the AP Name that it is associated to is 172.16.72.200, which is actually the IP address of the primary controller.

This means that the routing fails.

On 172.16.72.200 (Primary) the Mobility role of the client is described as "Foreign".

On 172.16.32.200 (Secondary) the Mobility role of the clinet is described as "Anchor".

Now I'm sure that there is something I have missed here.

Does anyone know how to get the client to re-associate with the primary controller?

4 Replies 4

Scott Fella
Hall of Fame
Hall of Fame

That is how mobility works in an lwapp environment. Remove the wlc's on the mobility group and the user will drop and have to reassociate. No need to have the wlc's in the mobility if you want to have different dhcp scope.

-Scott
*** Please rate helpful posts ***

I tried that but the AP couldn't find the secondary controller over the WAN.

I have reverted to H-REAP for all APs and WLANSs. This ensures that the clients keep a local IP address in the event of failure and maintain continuity during the failover process.

That is the best way to go. You also could of used an IP-Helper and forward udp port 12223 on the l3 interface to help the ap find the second controller.

-Scott
*** Please rate helpful posts ***

eastman.rivai
Level 1
Level 1

What you just simulated was inter controller roaming. WLC v4 and above support auto anchor, which means whenever a client move to another AP on a different controller (in this case, same AP but move to different controller), the connection will be anchored to their original controler (in your case WLC2), thus it will keep the IP address on this site. So nothing wrong with this at all. repairing or reconnectiong the client will get them back to WLC1.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: