Access Layer Redundancy Design

Dave.Messenger — Tue, 22 Oct 2019 09:30:14 GMT

I'm currently researching Data Centre designs and I'm looking specifically at the 9K series of switches. I have large port count requirements and high bandwidth east-west traffic flows.

The design I'm considering is the following:

Core - 9500 100G

Aggregation - 9500 40G/100G

Access - 9400 10G

My goal is to keep the majority of traffic flows within the access layer.

My concern however is the lack of layer 2 redundancy on the 9400 access layer. The advice I see is that virtual stackwise can and should be used at the distribution / core layers but not at the access layer. Am I looking at the wrong switches to provide layer 2 redundancy to downstream servers? I'm comparing to the VPC feature of the Nexus line. I'm looking specifically for a solution which will provide multi-chassis etherchannel. Would the stackable 9300 series switches actually be a better fit here? Perhaps it is possible to move the layer 3 down to the access layer and provide FHRP and multi-chassis etherchannel and I've missed something.

Would someone be able to point me in the right direction?

topic Access Layer Redundancy Design in Server Networking

Access Layer Redundancy Design