I am looking for some assistance in designing a wireless solution for a customer of ours. A basic diagram of the network is attached. There are actually many more sites than are shown here but I just wanted to keep it simple.
The plan is to have one 1240AG access point at each remote site connected to the WLC across the WAN. The WCS will in time be managed by a WCS server.
Each site is already partitioned into 2 subnets and it is not possible to add a new one for the AP and it's clients. There will anyway be very little wireless traffic. The AP will operate in H-REAP mode but there will be no local communication required. All traffic is always destined for HQ (ie they use Citrix). Therefore do these APs actually need to be locally switched H-REAPs? If not then how would this work? Can a remote site AP work as a non H-REAP?
Some sites have their own 3560 switch and others have just a switch module inside a 2801 router.
There is no need to have VLANs routed across the WAN.
DHCP is provided by the router on each remote site.
I really need some advice on the best way to implement this solution including but not limited to:
Do I need to have a separate WLAN for each site?
Will DHCP from the router actually work?
How should the switch ports on the HWIC module be configured?
What benefits is there in having a WCS when there is just a single WLC? (I thought WCS was for managing multiple WLC).
I have skimmed the design guides and i am still unclear on the best way to do this.
Any help would be much appreciated
Not in H-REAP... the WLAN SSID you create on the controller will always be set to the management interface. This is how the AP and WLC will communicate. When you enable loacal switching, then you have the ability to map that ssid to a local vlan. I'm currently deploying 90 ap's in h-reap with 27 sites and 2 4404's at the core. I have 7 ssids and using wlan override. DHCP is local on the remote site and I'm using the same vlan id for mapping the ssid. Local native vlan for the ap is also the same throughout, but of couse the subnets are different. It is a bit different how you look at what interface you have to map the ssid to especially when doing h-reap.
Look at it this way... only when you have traffic comming back to the WLC and the WLC putting the traffic back on the network is when the interface mapping is important... like in a normal LWAPP install.