A live coding demonstration shows how to build a standard blocking API with Spring Boot and how its performance degrades under load.
Creating thousands of traditional platform threads demonstrates their high memory consumption, 1:1 mapping to OS threads, and eventual OutOfMemoryError.
Refactoring the blocking API to use asynchronous programming with CompletableFuture solves the performance bottleneck but introduces significant code complexity.
Virtual threads are lightweight, JVM-managed threads that run on a small pool of carrier threads, enabling millions of concurrent tasks without high memory overhead.
A single configuration flag in Spring Boot 3.2 is all that's needed to switch the application to use virtual threads for handling requests.
StructuredTaskScope provides a modern API to run multiple blocking operations concurrently in separate virtual threads, simplifying parallel code.
Virtual threads can be "pinned" to their carrier thread by synchronized blocks, native calls, or CPU-intensive tasks, which negates their benefits.
Jobs that call for the skills explored in this talk.