This repository contains curated performance benchmarks used to validate architectural decisions in the FlowtraceX Network Detection & Response (NDR) platform.
The benchmarks focus on real-time stream processing, scalability, and cost-efficiency for security detection workloads.
Path: flink/
These benchmarks evaluate:
- Pure Flink SQL execution
- SQL with light and heavy Python UDFs
- DataStream vs SQL performance characteristics
- Serialization vs computation overhead in PyFlink
Detailed methodology, results, and analysis are documented in:
➡️ flink/README.md
- Production-scale stream processing
- Security detection workloads (NDR use cases)
- Architecture and cost trade-offs
Additional benchmark categories will be added incrementally.
- Curated and reproducible
- Focused on real production constraints
- Designed to inform architectural decisions
-
FlowtraceX Core Platform
https://github.com/flowtracex/flowtracex-core