PlantUML DiagramRouter-on-a-Stick
A structured lesson workspace with readable content, hands-on examples, and a clean path to completion.
A small office has five VLANs and a single Cisco router with one spare Ethernet port. The router's other interfaces are all in use for WAN and internet. You need to route between the VLANs. Do you buy another router, or do you make the one port do the work?
Router-on-a-stick is the answer to that question. One physical router interface, multiple logical subinterfaces — one per VLAN. The router treats each subinterface as if it were a separate physical port. Inter-VLAN traffic flows to the router over a trunk, gets routed, and comes back over the same trunk.
The switch port facing the router is a trunk carrying all three VLANs. The router has one physical interface (Gi0/0) with three subinterfaces (Gi0/0.10, Gi0/0.20, Gi0/0.30), each configured with 802.1Q encapsulation for the corresponding VLAN.
That traffic path is the key limitation: every packet crosses the same physical link twice — once from source switch to router, once from router back to switch. If the link is 1 Gbps, effective inter-VLAN bandwidth is at most 500 Mbps when traffic is bidirectional and heavy.
For a small office, none of these are blockers. For a campus with 50 VLANs and heavy inter-VLAN traffic, router-on-a-stick is the wrong design — you'd move to a Layer 3 switch (the next lesson after this one).
Router-on-a-stick routes between VLANs using one physical link and multiple subinterfaces. It's simple and works, but the shared link is a bandwidth bottleneck. Next lesson covers the actual configuration.
Finish the lesson once you have worked through the material. This awards ★ 15 XP.