What is the difference between horizontal and vertical scaling?

Vertical scaling is the fastest first step: a bigger database instance or more memory often solves a short-term bottleneck without code changes. The limit is cost and hardware size, and a single machine remains a single failure domain. Horizontal scaling spreads load across replicas and can survive a node loss.

What horizontal scaling requires

• Stateless application instances or externalized session state
• A load balancer or service discovery layer
• A shared database, cache or queue that can handle the connection count
• Idempotent operations and safe retries across instances
• Distributed tracing and centralized logs to diagnose a fleet
• A plan for database scaling, which is often the real bottleneck

const sessions = new Map();

app.post("/login", (req, res) => {
  sessions.set(req.body.id, req.body);
  res.json({ ok: true });
});

Scaling the application tier is usually easier than scaling the database. Adding ten app instances can exhaust the connection pool or saturate the primary database. Read replicas, partitioning, caching and query optimization may be needed before more app replicas help.

Vertical scaling still matters for databases and stateful systems, where horizontal sharding adds significant complexity. A common path is to scale up first, then scale out the stateless tier, then address data partitioning when measurements justify it.