This is a command-line tool for performing HTTP load testing, designed to evaluate the performance of web services by simulating a specified number of requests per second (QPS) over a set duration.
- Supports all HTTP Methods(GET, POST, PUT , PATCH and DELETE) requests.
- Allows specifying custom headers and payloads.
- Calculates various performance metrics such as average latency, standard deviation, and error rate.
- Configurable request timeout.
- Docker
First, build the Docker image:
docker build -t hammerhead .To run the load testing tool, use the following command:
docker run -it --rm hammerhead <url> [options]- url (required): The URL to test.
- --qps (optional): Queries per second. Default is 1.
- --duration (optional): Duration of the test in seconds. Default is 10.
- --method (optional): HTTP method (GET or POST). Default is GET.
- --headers (optional): HTTP headers in the format "Header1:Value1,Header2:Value2".
- --payload (optional): HTTP payload as a string.
- --max_timeout (optional): Maximum timeout for any HTTP request in milliseconds. Default is 10,000.
docker run -it --rm hammerhead http://www.example.com --duration=10 --qps=2docker run -it --rm hammerhead http://www.example.com --duration=10 --qps=2 --method=POST --headers="Content-Type:application/json,Authorization:Bearer mytoken" --payload='{"key1":"value1","key2":"value2"}'The tool logs the following information:
- Start and completion of the load test.
- Latency of each successful request.
- Summary statistics after the test, including:
- Minimum, maximum, and average latency.
- Standard deviation of latency.
- 95th and 99th percentile latency.
- Error rate.
- Total number of requests.
- Allow user to define maximum concurrent threads for load testing.
- Generate a detailed report (in formats such as CSV or JSON) that can be analyzed later.
- Using a configuration file (YAML/JSON) for load test parameters
- Allow specifying headers and payload as JSON data
- Add More Unit test cases to cover all corner cases
- Explore using a thread pool for worker thread management
- Implement different logging levels and potentially separate log files
- Implement more sophisticated error handling and retry mechanisms.