Can We Use Wireless LAN controllers (WLC) over WAN ?

[seanwy]

hi folks,

are we able to utilise the WLC over a WAN connection? Say the primary WLC is at the HQ, and the secondary WLC is at one of the remote branches.

If its able to support, what are the restrictions over the WAN? bandwidth capacity ?

I understand tt WLC utilises the (N+1) redundacy model. So am I right to assume tt the secondary WLC is only able to support ONLY 1 master WLC ?

appreciate any insights/feedbacks. Thank you!

regard,

Sean.

[Aaron Harrison]

Hi

You can indeed use WLCs over a WAN. First and most important is that you get the right APs - 1010 and 1020 APs put all traffic through the LWAPP tunnel to the WLC, so each AP could use maybe 6MB or so over your WAN.

1030 APs run in REAP (Remote Edge AP) mode which bridges traffic onto the local LAN, but all authentication and management is still done by the WLC.

You can configure each AP with a list of up to 3 WLCs to connect to. The 'Master' functionality on these WLCs is set on one controller per subnet, and determines which WLC APs will connect to by default.

Clustering on these devices is quite poor in my opinion - the WLCs communicate to facilitate roaming etc, but config is done on each WLC seperately so you need to make any changes to each WLC.

Regards

Aaron

Please rate helpful posts...

[jcornford]

I think what you're asking is can you have a secondary or "backup" WLC at a different physical site. Is that right? The short answer is yes, but my main concern would be your WAN bandwidth. If you've got something like a routed 100Mbit or more link between sites you should be ok to do it (without knowing your applications or WiFi client loads), but I'd have to suggest no if it's 10Mbit LES or less. Call your sites A and B. When site A's WLC fails in this design, all the normal LWAPP type APs would pass their traffic back to site B's WLC. This could result in WAN degradation. It’s all about your WAN speeds really. Also, you might need to look at any firewall rules between sites.

Regarding redundancy, think of it like this. APs have a primary, secondary (backup), and tertiary (last chance) setting. They normally connect to the controllers in that order (the actual process has a bit more to it than that but don’t worry). Don’t think of it in the sense that any given controller supports another as it’s master. It works more on the principal of a number of APs (X) could connect to a controller as their first choice (primary). If they can’t, they try the secondary controller, then the third. They move to the next controller when the first choice isn’t available. Let’s say you have 3 controllers on one site supporting 12 APs, and you have 24 APs. You’d configure the APs to use the controllers in this fashion.

First 8 APs > Controller A, Controller B, Controller C

Second 8 APs > Controller B, Controller C, Controller A

Third 8 APs > Controller C, Controller A, Controller B

This way, it’s load balanced, and when any controller fails, the APs will failover evenly to the other controllers. The N+1 means that you need enough controllers to support all your APs, then +1 is a live backup controller big enough to accommodate the largest of the active controllers that could fail. Does that make sense???

In your example you have a controller at your existing site. If you had to controller your second sites APs on that controller, could it handle them? If so, you’d buy another controller of equal power to act as the backup.

Oh and additionally, you might want to buy the WCS software so you can manage the configuration of all your controllers in one place.

Hope that helps!