mediumMultiple Choice
CCNP Practice Question: An enterprise is migrating its data center to a…
An enterprise is migrating its data center to a leaf-spine architecture. The design must provide high availability and support for east-west traffic patterns. Which design choice best meets these requirements?
⚠ Common exam trap
Cisco often tests the misconception that a three-tier design is always more reliable or that a collapsed core is sufficient for modern data centers, but the trap here is that candidates overlook the specific requirement for east-west traffic patterns, which demands a flat, non-blocking fabric like leaf-spine with ECMP rather than traditional hierarchical or ring topologies.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Implement a leaf-spine topology with multiple spine switches and ECMP.
A leaf-spine topology with multiple spine switches and Equal-Cost Multi-Path (ECMP) routing provides high availability by eliminating single points of failure and supports east-west traffic patterns by ensuring that any leaf switch can reach any other leaf switch with a consistent number of hops (typically one hop via a spine). ECMP allows load balancing across all available spine links, maximizing bandwidth and redundancy for data center east-west flows.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Deploy a collapsed core with a single pair of core switches.
Why it's wrong here
A collapsed core design funnels all inter-rack traffic through a single pair of core switches, creating a severe oversubscription bottleneck for east-west flows common in virtualized and microservices environments. The architecture offers only a single point of failure at that pair, and its fixed capacity cannot scale linearly as compute and storage nodes are added, unlike leaf-spine which provides numerous parallel paths.
- ✗
Use a three-tier hierarchical design with access, distribution, and core layers.
Why it's wrong here
The classic three-tier access/distribution/core hierarchy was originally engineered for north-south traffic, where clients reach servers and responses return the same path. In modern data centers, server-to-server east-west traffic must traverse up to the distribution layer, introducing additional latency and contention as every flow crosses that tier. The distribution layer becomes a constrained chokepoint, and the design lacks the any-to-any, equal-cost multipath capabilities of leaf-spine needed for efficient east-west forwarding.
- ✓
Implement a leaf-spine topology with multiple spine switches and ECMP.
Why this is correct
Leaf-spine is a two-stage topology where each leaf switch connects to every spine switch, forming a full mesh that provides predictable, uniformly low latency between any two servers. Equal-cost multipath (ECMP) enables traffic to be hash-distributed across all available spine uplinks, maximizing aggregate bandwidth while avoiding link oversubscription. Adding spine or leaf switches scales bandwidth and port density linearly, and the multiple active paths deliver fast failover and high availability, making this design ideal for heavy east-west virtualization traffic.
- ✗
Use a ring topology connecting all switches in a loop.
Why it's wrong here
A ring topology connects switches in a loop with exactly two links per device, so east-west traffic may hop through many intermediate switches, increasing latency and consuming bandwidth on each transit segment. The total effective bandwidth is constrained by the ring's shared links, and a single link or switch failure requires spanning tree or similar protocols to reconverge, causing unpredictable outages. This design offers no parallel forwarding paths and is therefore unsuitable for the high-throughput, low-latency, and resilient requirements of modern data center east-west communication.
Visual reference
About these practice questions
Courseiva writes every 350-401 question from scratch — 1,923 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This 350-401 practice question is part of Courseiva's free Cisco certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the 350-401 exam.