How do vertical scaling and horizontal scaling differ, and when would you choose each?

• A. Vertical scaling adds capacity to a single node; horizontal scaling adds more nodes; choose vertical for simpler architectures; horizontal for elasticity and fault tolerance.
• B. Vertical scaling means adding more servers to the system.
• C. Horizontal scaling reduces fault tolerance.
• D. Vertical scaling is always more cost-effective than horizontal scaling.

Answer: (A)

Vertical scaling means upgrading a single machine’s capacity—more CPU, RAM, or storage—so you handle more work on one node. Horizontal scaling means adding more machines to share the work, usually with load balancing and a distributed approach. The simplest way to grow is to scale up a single node when the workload fits within its hardware limits and you want to avoid the complexity of a distributed system. This keeps things straightforward, with fewer moving parts and easier maintenance. But it has a ceiling, and relying on one machine can become a single point of failure.

Scaling out, or horizontally, is preferred when you need elasticity and fault tolerance. You can add or remove nodes on demand, distribute load across multiple machines, and keep the system running even if some nodes fail. The trade-off is greater architectural complexity: you have to manage load balancing, state, consistency, and data distribution across nodes.

In short, choose vertical scaling when you want a simpler setup and your workload stays within a single machine’s capacity. Choose horizontal scaling when you need scalable capacity with better resilience and you’re ready to manage a distributed system.