🔧 Node.js Thread Pool Benchmark

This project demonstrates how Node.js handles heavy tasks using the internal libuv thread pool, even though it's single-threaded at its core. It benchmarks how different thread pool sizes (UV_THREADPOOL_SIZE) affect total execution time using the crypto.pbkdf2 function.

📈 What does it show?

Node.js uses a thread pool behind the scenes to offload expensive operations (I/O, crypto, compression, etc). This benchmark simulates a CPU-heavy task using crypto.pbkdf2 to show how concurrent workloads are distributed across available threads.

📂 Project Structure

.
├── main.js       # Runs the benchmark
├── plot.js            # Generates a graph from the results
├── results.csv        # Stores benchmark results (auto-generated)
├── Dockerfile         # Containerized setup
├── Makefile           # Shortcut to run with different thread sizes
└── README.md          # You're here

🚀 How to Run

1. Build Docker image

docker build -t threadpool-benchmark .

# makefile
make build

2. Run with different thread pool sizes

make run THREADS=2
make run THREADS=4
make run THREADS=8

Each run appends a line like 4,1.82 (threads,time) to results.csv.

3. Generate the chart

node plot.js

This generates a file named result.png, showing execution time vs thread pool size.

📊 Benchmark chart result

💻 Example Output

🧵 UV_THREADPOOL_SIZE = 4
🚀 Firing 10 tasks with pbkdf2...

✅ Task 1 finished in 0.63s
✅ Task 2 finished in 0.64s
...
🏁 All tasks concluded in 1.82s

📦 Saved in results.csv: [4,1.82]

📦 Docker Volume Tip

If you're running this in Docker and want to persist results.csv, mount the current folder:

docker run --rm -e UV_THREADPOOL_SIZE=4 -v $(pwd):/app threadpool-benchmark

🧠 Takeaway

This benchmark visually proves how the thread pool offloads work in parallel — reducing total execution time without blocking the Node.js main thread.

---

Made with 🧵 by @Artur