Spring Boot 4 Performance in Practice

Most performance books teach you what the knobs do. This one teaches you which knobs to turn, and proves it with numbers.

I take CinéTrack, a real Spring Boot 4 monolith, and push it from its baseline throughput to fifty times that over twenty-eight chapters. Every chapter starts with a profiler, finds an actual bottleneck, applies a single change, and measures the result side by side. Some changes buy 30%. Some buy 3%. Some buy nothing. I say so.

You will learn the measurement discipline that separates engineers who tune from engineers who guess: how to read a flame graph, how to interpret p99 latency, how to tell a GC problem from a thread-pool problem from a query problem before you touch a line of code.

Then you will learn the tuning stack: virtual threads, ZGC, Hibernate batch loading, connection pool sizing, Redis caching, Kafka back-pressure, GraalVM native image. Not as a menu of options but as a sequence of decisions, each grounded in what the profiler actually showed.

For engineers who have built something with Spring Boot and now have to make it fast.