Is it possible to use Stackwise Virtual technology at Intermediate Node as it is a point of aggregation for access layer and it can still support layer-3 routing.
Yes you can use SWV as intermediate node, that's fine. But, why would you do this instead of running two separate L3 switches as intermediate nodes? SWV exists to solve L2 connectivity scenarios (MEC, no STP blocking, etc.), and SDA underlay is a pure L3 environment. If we do not need SWV then the recommendation is don't use it. No SWV = less protocols in the network, which increases simplicity and stability. Please share a diagram if possible to explain why SWV is used as intermediate node. Best regards, Jerome
Thanks for your reply.
Currently the existing network is running with two core devices and running HSRP. The connected access layer switches are still legacy devices at access layer which doesn't support the Layer 3 protocol.
A new infrastructure facility with SDA solution is being created for new building. However, the existing legacy network also requires migration which is the planned for future. Since the core switch devices are almost nearing to EoL, it requires immediate replacement and hence will be replaced with two new Cat 9K switches.
Replacing the whole of old network with SDA is not possible at this stage.
Keeping the existing network and extending to the new network forming the Intermediate Node provides easy migration from the old switched environment to new routed network later on. However, for a smooth migration of these legacy switches requires the replication of existing configuration due to access layer limitation and also due to the span of VLANs across multiple wiring closet.
Using SWV at this layer should be resolving both issue, the legacy traditional switched network and also the ability to join the SDA network incrementally. The role of Intermediate Node is just pass the IP packets in an routed environment with the major prerequisite of Jumbo MTU size.
Based on the above situation, I think SWV is the best solution but if you guys has better ideas you are welcome. The major challenge is to replace the existing switches seamless with minimum downtime as the network availability is critical.
Hi hafeezali, sounds like a well reasoned plan. I do the same thing in the migration scenario in DGTL-BRKENS-3822 , https://www.ciscolive.com/global/on-demand-library.html?search=brkens-3822#/session/1570575336196001v4R5