Skip to content
BENIBIX Learn

Autoscaling

Infrastructure☁️ Virtualization & CloudCloud Basics

Also known as Auto Scaling, Auto-Scaling, Elastic Scaling

More servers when busy, fewer when quiet. It happens on its own.

Think of it like

It is like a shop that opens more tills when the queue gets long. It closes them again when it is quiet.

Example

A ticket site runs 4 servers on a normal day. When tickets go on sale, autoscaling grows it to 40, then shrinks it back by evening.

Why it matters

You pay for the servers you need now, not for the busiest hour of the year all year round.

Definition

Autoscaling adds or removes servers on its own as the load changes. You set a rule, such as keeping the average processor use near 50 percent. When traffic rises, it starts new instances from a machine image. When traffic falls, it shuts some down. A load balancer spreads the work across whatever is running.
When the load rises, autoscaling starts more copies from the image. When it falls, it stops them.

Common mistake

Expecting it to handle a sudden spike at once. New servers take minutes to start, so keep some spare room.

Key takeaways

  1. 1
    Follows the load. More servers when busy.
  2. 2
    Rules decide. Set a target, like processor use.
  3. 3
    Needs a balancer. To share work across them.

Compare

Learn it properly