Skip to content

BENCHMARKS.md

Latest commit

 

History

History
15 lines (11 loc) · 575 Bytes

BENCHMARKS.md

File metadata and controls

15 lines (11 loc) · 575 Bytes
Raw

StreamForge Benchmarks

The canonical benchmark instructions are in docs/PERFORMANCE_TESTING.md.

StreamForge distinguishes:

  • Criterion microbenchmarks for isolated code paths;
  • synthetic pipeline benchmarks that do not include Kafka;
  • Kafka-backed completion-throughput runs produced by scripts/benchmarks/run_throughput_test.sh.

Historical reports under docs/benchmarks/results/ are retained as historical artifacts. They are not current baselines unless the workload, environment, configuration, and measurement method match.