Hi, everyone.
I have some shorter questions today regarding how APs are deployed in an SD-Access fabric. My book says:
A fabric-enabled WLC connects APs and wireless endpoints to the SD-Access fabric. The
WLC is external to the fabric and connects to the SD-Access fabric through an internal border node.
While my CBT Nuggets resource says that the WLC is located within the fabric… so which one there is correct?
The control plane node maps the host EID to the current fabric access point and fabric edge node location the access point is attached to.
In the case of a fabric WLC, does it register the EIDs with the control node instead of the switch (fabric edge node)? If so, what information does it provide? The EID, the AP, and the fabric edge node location, all three together?
Fabric APs establish a VXLAN tunnel to the fabric edge to transport wireless client data
traffic through the VXLAN tunnel instead of the CAPWAP tunnel. For this to work, the
AP must be directly connected to the fabric edge or a fabric extended node.
So does the AP establish a VXLAN tunnel starting from itself all the way to the destination edge node? So the control plane remains centralized while the data plane isn’t?
If possible, could someone tell me how exactly is an AP and a wireless client onboarded? I don’t think I quite understand the full process. The AP comes online, builds a CAPWAP tunnel to the WLC, the WLC then registers this AP and any further clients… or? When exactly does the AP build the VXLAN tunnel?
Thank you.
David