To understand SD-Access, let me show you the traditional problems of enterprise campuses.
For decades, enterprise campuses have followed the same model.
VLANs per floor, subnets per VLAN, and ACLs based on IP addresses.This model worked when users stayed at their desks.
But modern users move, devices multiply, and policies need to follow people, not cables.Identity Tied to the Network
In a traditional campus, your identity on the network is defined by where you plug in.
Each floor has its own VLAN.
Each VLAN has its own subnet, and your IP address comes from that subnet.
Figure 1 – Identity follows the IP, not the user
This is fine until a user moves.
An employee walks from the 1st floor to the 3rd floor.
They land on a different VLAN and get a new IP address.
Every IP-based ACL and QoS policy needs to be rewritten to recognize them.The network identifies you by your location, not by who you are.
This is the same scalability problem as traditional ACLs, solved later by Cisco TrustSec.Answer the question below
In a traditional campus, what does your identity on the network depend on?
Configuration Doesn't Scale
The second problem is operational.
In a traditional campus, every switch is configured individually through CLI.
Figure 2 – One SSH session per switch
Adding a new VLAN means logging into every access switch.
You then type the same commands yourself, one device at a time.
Updating an ACL means doing it on every device that enforces it.In a campus with hundreds of switches, this approach creates errors and slows down deployments.
SD-Access exists to fix both of these problems.Answer the question below
In a traditional campus, how is each switch typically configured?
SD-Access (Software-Defined Access) solves these problems by building a fabric.
A fabric is a logical layer that turns your campus into a single managed network.
Inside this fabric, identity follows the user, and policies follow the user too.Inside the fabric, every wired device plays one of five roles.
Let's reveal them step by step.Fabric Edge — Where Users Connect
At every floor and every location, the traditional access switch is replaced.
It becomes a Fabric Edge node.
This is the device where your users plug in: laptops, IP phones, access points, printers.
Figure 3 – Fabric Edge nodes at every access location
The Fabric Edge plays two roles.
It is the point of attachment for your users.
It is also responsible for placing their traffic into the fabric correctly.Every Fabric Edge also acts as an anycast gateway.
A user can move between floors and keep the same default gateway, as if nothing had changed.But the Fabric Edge does not know where the other users are in the fabric.
It needs something that tracks every user in real time.
So who keeps that directory?Answer the question below
What is the name of the SD-Access node where end users plug in?
Answer the question below
What kind of gateway does every Fabric Edge act as?
Control Plane Node — The Brain
The Control Plane Node is the brain of the fabric.
It maintains a real-time database of every user and device in the fabric.
It also knows where each one is attached.
Figure 4 – Control Plane Node: the LISP directory
Internally, the Control Plane Node uses LISP (Locator/ID Separation Protocol).
Every Fabric Edge registers its connected users with it.
Each Fabric Edge then queries this directory whenever it needs to reach another user.This is what allows your fabric to follow the user instead of the IP.
When a user moves, only the Control Plane entry is updated.
No ACL changes, no policy rewrite.Your fabric now has access points and a directory.
But what happens when a user needs to reach something outside the fabric?Answer the question below
Which protocol does the Control Plane Node use internally to track users in the fabric?
Border Node — The Door to the Outside
The fabric is a closed bubble.
Users inside can talk to each other through the Fabric Edges and the Control Plane Node.
But anything outside the fabric is unreachable until you add an exit point.That exit point is the Border Node.

Figure 5 – Border Node: the gateway to the outside world
The Border Node connects the fabric to everything that is not part of it.
That includes the data center, the WAN, the Internet, and traditional VLANs not yet migrated.When a user needs to reach an external resource, their traffic is forwarded to the Border Node.
The Border Node then hands off the traffic to the outside network.Answer the question below
Which SD-Access component connects the fabric to networks outside of it?
Fabric Intermediate Nodes — The Plumbing
Until now, you have seen the fabric as if every Edge connected directly to the underlay bus.
In a real campus, that bus is built from many switches working together.
Figure 6 – The SD-Access fabric sits on the underlay
Traffic between Fabric Edges and the Border Node transits through Fabric Intermediate Nodes.
These are simple Layer 3 switches that route the underlay packets.They do not participate in LISP or in the policy.
That is why they are not shown as a separate component in our diagrams.Answer the question below
Fabric Intermediate Nodes are simple switches at which layer?
Answer the question below
3 buildings of 4 floors, one Fabric Edge per floor, plus CP and Border nodes: how many act as 10.50.0.1?
The wired side of the fabric is now complete.
But your campus also serves wireless users, and the whole system still needs to be configured and monitored.
Two more pieces complete the picture.Fabric WLC — The Wireless Anchor
Wired users have their Fabric Edge.
Wireless users need a different anchor: the Fabric WLC (Wireless LAN Controller).
Figure 7 – Fabric WLC anchors wireless users into the fabric
The Fabric WLC sits outside the fabric.
It onboards access points and registers wireless clients with the Control Plane Node.From the fabric's point of view, a wireless client behaves the same way as a wired one.
The same identity, the same policy, the same anycast gateway.Answer the question below
Which component anchors wireless users into the SD-Access fabric?
Catalyst Center — The Orchestrator
Five fabric roles handle the network itself.
But you still need a system to design, deploy, and monitor the whole thing.That system is Cisco Catalyst Center (formerly Cisco DNA Center).
It is a centralized platform that designs, deploys, and monitors every fabric device.From the Catalyst Center web GUI, you define the underlay and provision new switches.
You create policies, then push the configuration to every device automatically.
Figure 8 – Catalyst Center: the fabric orchestrator
You no longer need to SSH into each switch and type commands by hand.
Catalyst Center also integrates with Cisco ISE (Identity Services Engine), the Cisco identity server that handles user authentication and security policies.
This is what makes your fabric software-defined.
The entire campus is configured from a single dashboard, not box by box.Answer the question below
Which SD-Access component provides the centralized GUI to design, deploy, and monitor the fabric?
You now know the five fabric roles plus the orchestrator.
Time to put everything together and look ahead.The Three Fabric Planes
Inside the fabric, three planes work together.
Each one has a single responsibility.Control plane uses LISP to track where every user is
Data plane uses VXLAN, an encapsulation that wraps user traffic into a tunnel across the fabric
Policy plane uses Cisco TrustSec to enforce group-based security and QoS
Three planes, three technologies, one fabric.
You already know LISP from a previous lesson.Answer the question below
Which protocol does the SD-Access data plane rely on?
Answer the question below
Which technology does the SD-Access policy plane rely on?
All Components in Place
Here is the full architecture in one view:
Fabric Edge connects wired users and acts as anycast gateway
Control Plane Node tracks every user with LISP
Border Node connects the fabric to external networks
Fabric WLC anchors wireless users into the fabric
Catalyst Center orchestrates everything from a single dashboard

Figure 9 – The complete Cisco SD-Access architecture
All these components run on top of a physical underlay.
The underlay is a routed Layer 3 network of switches that provides simple IP connectivity between fabric devices.
On our diagrams, the horizontal bus is just a simplified representation of this routed network.You now know the architecture.
But knowing the components is only half the story.
How does a Fabric Edge actually find another user? How is your traffic carried from one site to another? How does VXLAN really wrap a packet?
Time to open the hood. The next lesson shows you how the control plane and the data plane operate together to make your fabric work.Answer the question below
How many fabric roles make up the SD-Access overlay?