Leaf-Spine Network Topology and Top-of-Rack (ToR) Switching
Plain-English definition
A non-blocking two-tier data center network architecture where every Top-of-Rack (ToR) leaf switch connects upward to every spine switch via high-speed optical links, giving server-to-server ('east-west') traffic predictable two-hop latency and multi-path resilience.
Why it matters on shift
Understanding redundant ToR-A / ToR-B switch placement and air-intake direction (port-side intake vs port-side exhaust) prevents operators from installing reversed-airflow switches that suck 42°C hot-aisle exhaust straight over optical transceivers.
Illustrative operational example
In an illustrative cabinet installation, dual ToR leaf switches draw cool air from the cold aisle using front-to-rear fan trays and uplink over 400G/800G fiber trunks distributed across four independent spine switches.
Scope, code & safety boundary
Always verify fan-tray airflow arrows on replacement ToR switches before racking; one backward fan module pulls hot exhaust into the cold aisle and overheats optical modules.
Related courses & interactive lab practice
Facility, rack & safety foundations
Air, chilled-water & liquid cooling
High-density AI cooling & commissioning
Practice live readings and switching in "NOC, Spine-Fiber & White-Space Walkdown" with no login required.
Primary standards & technical references