Scaling: Bigger or More Servers

13 min read

One server can answer only so many requests each second. When more users come, you must scale. You can make the server bigger, or you can add more servers. This lesson shows both ways and their limits.

New words

Scaling
Making a system able to handle more users or more work.
Vertical scaling
Using one bigger and stronger machine.
Horizontal scaling
Using more machines that share the work.
Load balancer
A helper that sends each request to one of the servers.
Stateless
A server that keeps no user data between requests.
Bottleneck
The slowest part, which limits the speed of the whole system.

What you will learn

  1. The 10 am Tatkal rush
  2. Vertical scaling: a bigger machine
  3. Horizontal scaling: more machines
  4. Stateless servers
  5. Find the bottleneck
  6. Try it: survive the rush
  7. Auto-scaling
  8. Bigger or more servers: what do you gain and lose?
  9. In a real system
  10. Interview tip
  11. Key takeaways

10 quiz questions at the end.

Start this lesson

The full lesson, with interactive demos and the quiz, opens after you sign in with Google. It is free.

All lessons · Back to the roadmap