fortio/fortio
Fortio
Fortio is a performance testing tool written in Go that generates constant request rates to measure latency histograms and response percentiles. It functions as a network performance benchmarker for analyzing system performance across HTTP and gRPC endpoints, and includes a specialized HTTP echo server for mirroring requests.
The project distinguishes itself through a web-based HTTP load testing API used for triggering asynchronous tests and visualizing result data through graphs. It also provides server behavior simulation, allowing the echo server to inject artificial delays and vary response sizes based on weighted probabilities to mimic real-world network conditions.
Broad capabilities cover load generation and latency analysis, DNS performance benchmarking, and network traffic simulation. The toolset also includes traffic management utilities for API rate limiting, TLS connection termination, and protocol proxying for HTTP and gRPC services.
Monitoring and observability are handled through telemetry data export, structured JSON logging, and the generation of machine-readable load result exports for external analysis.
Features
- Constant Throughput Generators - Generates network requests at a constant rate per second to measure latency and system stability.
- HTTP Load Testing - Generates a constant rate of queries per second to measure response latency histograms and system performance.
- Concurrent Network Workers - Implements concurrent Go-routines to generate high-volume network traffic for performance benchmarking.
- gRPC Latency Measurements - Measures round-trip time and clock skew between services using gRPC ping messages.
- Latency Histograms - Records response times in buckets to calculate accurate percentiles and tail latency distributions.
- Tail Latency Measurement - Measures request percentiles and tail latency to identify performance bottlenecks in web services and network endpoints.
- Load Testing - Performs simulated high-volume request generation against network endpoints to evaluate system stability.
- Go-Based Load Testers - Performance testing utility implemented in Go for generating high-concurrency network traffic.
- Network Performance Benchmarks - Analyzes system performance by recording response times and tail latency across HTTP and gRPC endpoints.
- Analysis Result Exporters - Exports performance metrics and telemetry data into machine-readable JSON formats for external analysis.
- Behavioral Simulation - Implements an echo server that injects artificial delays and varies response sizes based on probability.
- Latency Tracking - Sends queries to DNS servers to track packet loss and calculate response latency.
- Echo Servers - Provides a specialized server that mirrors requests and injects artificial delays to simulate network conditions.
- Single Port Protocol Multiplexing - Routes both HTTP and gRPC traffic through a single entry point to different backend destinations.
- Response Simulation - Simulates real-world network conditions by injecting artificial delays and varying status codes based on probabilities.
- Asynchronous Test Triggering - Provides a web interface and API to trigger and monitor load tests asynchronously for automation.
- DNS Performance Analysis - Sends queries to DNS servers to track packet loss and calculate response timing and latency.
- Test Orchestration APIs - Offers a web-based API and interface for triggering asynchronous load tests and visualizing results.
- Response Simulators - Runs an echo server that injects artificial delays and varies response sizes to test client resilience.
- Load Ramping - Supports executing load tests using scripting to perform complex ramp-up scenarios.
- Test Result Visualizers - Provides visual graphing of saved JSON test data to compare response time percentiles.
- Networking Libraries - Load testing library and echo server with UI.
- Networking - Load testing and network benchmarking tool.
- 网络库 - Load testing and echo server tool.
- Testing and Benchmarking - Load testing library and tool for microservices.
