🐍 python-http-libraries-benchmark

This project benchmarks popular Python HTTP libraries to compare their performance across core networking metrics. The goal is to offer insights on how each library performs under real-world request loads.

---

📦 HTTP Libraries Benchmarked

This benchmark includes five widely-used Python HTTP libraries—requests, urllib3, httpx, aiohttp, and pycurl—chosen for their relevance, diversity, and real-world usage. requests is the most popular and beginner-friendly library, while urllib3 offers low-level control and powers requests internally. httpx and aiohttp provide modern asynchronous support for high-concurrency applications. Finally, pycurl wraps the high-performance C-based libcurl library, often used in system-level or legacy environments. Together, they represent a broad spectrum of HTTP client capabilities in Python, from ease of use to advanced performance tuning.

• aiohttp
• httpx
• requests
• urllib3
• pycurl

---

📊 Metrics Captured

Each library is evaluated using the following metrics:

• Requests per Second (req/sec): Measures throughput. A higher number indicates the library handles more traffic efficiently.
• Total Time: Measures how long the library takes to complete a full batch of requests.
• Average Connection Time: Indicates the average time spent on establishing connections for each request.

These metrics are chosen to reflect practical usage in applications such as REST APIs, microservices, and integrations.

---

🎯 Benchmarking Strategy

To ensure fair and unbiased comparisons:

• Each run sends a fixed number of requests (NUM_REQUESTS = 100) to a local HTTP server (http://localhost by default).
• The script performs 101 complete benchmark runs, where the first is used as warm-up and excluded from the CSV output.
• Randomized Execution Order: For every run, the order in which HTTP libraries are tested is randomized. This prevents any one library from benefiting from system or network caching effects due to consistent positioning.

---

📁 Output

Benchmark results are stored in benchmark_results.csv. Each row represents one full benchmarking round and includes:

• Start and end timestamps
• Requests per second, total duration, and average connection time for each library
• Number of requests executed

---

▶️ How to Run

1. Install requirements:

pip install -r requirements.txt

2. Start a local HTTP server (in a separate terminal):

python -m http.server 8080

3. Run the benchmark:

BENCHMARK_URL=http://localhost:8080 python benchmark.py

4. Generate charts:

python benchmark_analytics.py

Environment Variables

Variable	Default	Description
BENCHMARK_URL	http://localhost	Target URL for requests
BENCHMARK_RUNS	101	Total number of runs (first is warm-up)

Example with custom values:

BENCHMARK_URL=http://localhost:8080 BENCHMARK_RUNS=11 python benchmark.py

---

⚙️ CI

A GitHub Actions workflow runs the benchmark automatically every Sunday and commits updated results and charts back to the repository. It can also be triggered manually via the Actions tab, with optional inputs for Python version and number of runs.

Results

Environment:

OS: Windows 11 Home

Processor: Intel(R) Core(TM) i5-10400 CPU @ 2.90GHz 2.90 GHz

RAM: 12.0 GB

Location Type: Localhost