Auto-Scaling Patterns — Growing and Shrinking on Demand

Your application normally handles 100 requests per second. Then a marketing campaign goes viral, and suddenly you are getting 10,000 requests per second. Your two servers are drowning. Pages load slowly, then not at all. Users see error pages. Your site is effectively down during the moment it matters most.

Or the opposite scenario: you provisioned ten servers to handle Black Friday traffic. Black Friday is over. Now nine servers sit idle for the next 364 days, burning money.

Auto-scaling solves both problems by automatically adjusting your infrastructure capacity based on demand. More traffic? Add servers. Less traffic? Remove them. You pay for what you need, when you need it.

Vertical vs Horizontal Scaling

There are two fundamental ways to add capacity:

**Vertical scaling (s