What does it mean for a service to be stateless, and why does it matter for scaling?

A stateless service keeps no client-specific data in its own memory or on its local disk between requests.

Some data must survive after a request ends. Examples:

• sessions
• uploaded files
• caches that several instances must share

This data lives in a backing service, such as Redis, a database or object storage. Or it travels with each request as a signed token, like a JWT.

Why it matters for scaling: any instance can now serve any request. This means:

• The load balancer can use simple algorithms.
• You can add or remove instances freely. This is called autoscaling.
• A crashed instance loses nothing important. So deploys and failures do not log users out.

The Twelve-Factor App (a well-known set of rules for web apps) says it directly: processes are stateless and share nothing. Persistent data goes to a backing service.