giltene/wrk2
Wrk2
wrk2 is a high-performance HTTP benchmarking tool and constant throughput load generator. It is designed to measure server stability and throughput by sending a fixed number of requests per second, functioning as a high-resolution latency profiler and a Lua-scriptable performance tester.
The tool distinguishes itself through a scriptable interface that uses a high-performance just-in-time compiler to define complex request formats and custom response reporting logic. It employs high-resolution histograms to capture precise response time distributions and high-percentile latency data.
The project covers broad capabilities in load generation, high-precision latency analysis, and workload simulation. These include the ability to record detailed percentile spectrums to identify performance bottlenecks and the execution of custom HTTP workload scripting to automate benchmark results.
Features
- Performance Benchmarkers - Functions as a tool for measuring and comparing the performance and stability of web services under constant workloads.
- Scriptable Benchmark Engines - Uses an embedded Lua scripting engine to define complex performance test scenarios and request formats.
- Custom Request Logic Execution - Executes custom Lua logic to generate tailored HTTP requests and implement custom response reporting.
- JIT-Compiled Request Scripting - Provides an embedded LuaJIT engine to define complex request formats and custom response reporting logic with minimal overhead.
- Load Generation - Generates high volumes of network requests across multiple connections to benchmark server stability.
- Constant Throughput Generators - Sends a fixed number of requests per second to evaluate how server stability holds under a constant load.
- Constant-Rate Request Schedulers - Implements a precise timer mechanism to dispatch requests at a fixed frequency regardless of target response time.
- Latency Histograms - Uses high-resolution histograms to aggregate response time data for precise percentile calculation.
- Latency Profilers - Employs histograms to record precise response time distributions and extract high-percentile latency data.
- High-Precision Latency Analysis - Records detailed response time distributions to provide precise high-percentile timing data and identify bottlenecks.
- HTTP Endpoint Benchmarking - Measures the throughput and response speed of HTTP endpoints using a high-performance, scriptable engine.
- HTTP Load Testing - Generates a steady volume of HTTP requests to evaluate the stability and capacity of web servers.
- Workload Simulations - Allows the definition of specific request patterns through scriptable logic to mimic realistic production traffic.
- Custom Benchmark Scenarios - Provides scriptable definitions of performance tests to simulate specific application behaviors and request formats.
- Network I/O Multiplexing - Utilizes system-level multiplexing to efficiently handle thousands of concurrent network connections.
- Multi-Threaded Event Distribution - Distributes network I/O across multiple worker threads to maximize throughput and minimize measurement interference.
- Tail Latency Measurement - Tracks high-percentile response times using high-resolution histograms to identify tail latency bottlenecks.
- Percentile Distribution Computation - Calculates value distributions and percentiles using histograms to produce detailed response time spectrums.
- Command Line Tools - Listed in the “Command Line Tools” section of the The Book Of Secret Knowledge awesome list.
- Garbage Collection - A constant throughput, correct latency recording variant of wrk.
- Load Testing Tools - Constant throughput variant of a classic benchmarking tool.
- HTTP Security - Constant throughput variant of wrk with latency recording.
Star history
How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.
