What is a disadvantage of the traditional three-tier architecture model when east west traffic between different pods must go through the distribution and core layers?

Low bandwidth is not the specific disadvantage of east-west traffic traversing the distribution and core layers, as those layers typically feature high-capacity uplinks designed for aggregated traffic.

Security is a general concern in any network design but is not the technical disadvantage specifically introduced by routing east-west traffic through the distribution and core layers.

Scalability is a known limitation of the three-tier model overall, but the scenario described - traffic being forced through upper layers between pods - specifically illustrates a latency problem, not a scalability constraint.

When east-west traffic between servers in different pods must travel up from the access layer through the distribution layer and potentially the core layer before descending back to the destination pod, each additional hop adds forwarding and queuing delay. This multi-hop path is the defining latency disadvantage of the three-tier model for east-west traffic patterns, which dominate modern data center workloads. Spine-leaf architectures were specifically designed to address this by guaranteeing that any two endpoints communicate within exactly two hops regardless of pod placement.