Why you don't need to worry about scaling your Spring webapp

Marco Behler - JetBrains

Ever asked yourself: “Will my Spring app scale?” Did you have a good answer to that, backed with data, or just a general gut feeling that auto-scaling in the cloud will solve it?

This session will teach processes, tools & data interpretation skills to find out if and how your Spring app will scale.

When writing new Spring webapps, teams often have some nagging scaling doubts:

• How many users can my application simultaneously handle?
• How fast will my controllers dash out those JSON responses?
• How much memory, CPU or I/O does my application need?
• What is the cheapest, but most suitable instance type for my app?

Surprisingly, there’s very little advice out there on how to sensibly approach these questions in a practical way - apart from that deceiving gut-feeling which tells you that everything will be fine, by simply auto-scaling random Kubernetes pods into oblivion.

In this #almostnoslides session we’re going to use an IDE, a pen tablet and a few OS tools & custom code to get a deep, practical understanding of the following scaling topics:

• Throughput: How to reliably find out to how many users your application will scale to
• Memory & CPU: How much (or how little) memory & CPU your Spring app actually needs
• Latency: How your Spring webapp would compare to a Counterstrike game server
• Servers: What the cheapest Hetzner server can do for you
• And much more

By the end of this talk, you’ll have learned all the tools and, more importantly, solid load-testing processes that make you never worry about scale again. You will understand how to find out if your application is CPU bound, memory bound, or I/O bound and where most time in your application is spent. And you will understand how to collect performance data, correctly interpret that data and thus understand exactly what happens in your Spring application in various load testing scenarios.

See you there!