README.md

FastAPI Pystone Benchmark + Wasmer

This example runs the classic Pystone CPU benchmark behind a FastAPI endpoint on Wasmer Edge.

Demo

https://<your-subdomain>.wasmer.app/ (call /?loops=50000 to adjust the workload)

How it Works

src/main.py contains both the FastAPI app and the benchmark:

• The module defines app = FastAPI() so Wasmer can load main:app.
• Pystone logic (records, procedures, and helper functions) is imported from the historical benchmark implementation.
• LOOPS = 50000 sets the default number of iterations; the / route accepts an optional loops query parameter (capped to twice the default).
• The handler calls pystones(loops) and returns JSON containing the runtime, loop count, and benchmark version.
• Running the module directly starts uvicorn on 0.0.0.0:8000.

This exposes an easy way to gather performance data from Wasmer Edge.

Running Locally

python -m venv .venv
source .venv/bin/activate
pip install fastapi uvicorn
uvicorn src.main:app --host 0.0.0.0 --port 8000 --reload

Then visit http://127.0.0.1:8000/ or http://127.0.0.1:8000/?loops=100000 to trigger the benchmark.

Deploying to Wasmer (Overview)

1. Ensure the deployment points to src.main:app.
2. Deploy to Wasmer Edge.
3. Call https://<your-subdomain>.wasmer.app/?loops=50000 and inspect the JSON benchmark results.