Backpressure Strategies

Backpressure is where reactive design stops being abstract. Once producers emit faster than consumers can keep up, every pipeline needs a concrete policy for what happens next: accumulate, slow down, drop work, or stop requesting more data.

This section compares the main backpressure responses in Java so overload behavior is explicit instead of emerging accidentally from default operator behavior.

In this section

• Buffering in Reactive Java
  Use buffering in reactive Java pipelines when short bursts are acceptable and queue growth remains bounded and observable.
• Dropping and Throttling in Reactive Java
  Apply dropping and throttling in reactive Java when preserving responsiveness matters more than processing every emitted item.
• Requesting Strategies in Reactive Java
  Control demand explicitly in reactive Java streams so subscribers request work at a rate they can actually process.