Estimate server-to-server traffic inside a data centre and how much of it actually crosses the fabric.
Modern application tiers, storage replication and container overlays push 70–80% of data-centre traffic east-west. Only the part that leaves its own rack loads the spine layer, so rack-affinity in workload placement directly reduces fabric utilisation. Fabric upgrades are triggered by east-west growth, not internet growth, and this shows how much headroom the spine layer really has.
East-West Traffic
east-west = servers × per-server rate × east-west share; fabric load = east-west × share that leaves the rack.
east-west = servers × per-server rate × east-west share; fabric load = east-west × share that leaves the rack. Modern application tiers, storage replication and container overlays push 70–80% of data-centre traffic east-west. Only the part that leaves its own rack loads the spine layer, so rack-affinity in workload placement directly reduces fabric utilisation.
Fabric upgrades are triggered by east-west growth, not internet growth, and this shows how much headroom the spine layer really has.
This calculator takes 5 inputs: Servers in the fabric, Busy-hour traffic per server, Share that is east-west, East-west share leaving its own rack, Fabric bandwidth available. The pre-filled defaults are a realistic starting point — replace them with figures from your own environment for a result you can act on.