feat: add bun-cluster preset

[Saeid-Za]

🔗 Linked issue

resolves #3057

❓ Type of change

• 📖 Documentation (updates to the documentation, readme, or JSdoc annotations)
• 🐞 Bug fix (a non-breaking change that fixes an issue)
• 👌 Enhancement (improving an existing functionality like performance)
• ✨ New feature (a non-breaking change that adds functionality)
• 🧹 Chore (updates to the build process or auxiliary tools and libraries)
• ⚠️ Breaking change (fix or feature that would cause existing functionality to change)

📚 Description

This PR tries to create a new preset to support clustering mode similar to node-cluster in bun.

📝 Checklist

• I have linked an issue or discussion.
• I have updated the documentation accordingly.

[pi0]

Thanks for the PR. Since bun already supports node:http and node:cluster, can you please try a benchmark to see if output of node-cluster on bun is faster or slower than this new bun-cluster?

(since it adds complexity to the codebase, I prefer to recommend node-cluster to bun users. it makes a difference in nitro v3 with native fetch support which we can consider)

[Saeid-Za]

sure, I'll provide a benchmark soon, in the meantime, pooya, would you please expand on native fetch support?
Don't both bun and node support global fetch natively?
Thanks in advance 🙏

[pi0]

Yes indeed (and we do use native fetch for long time).

I meant Bun.serve({ fetch() => {} }) (as server interface). While it is more convenient, node:http implementation in Bun is sometimes faster.

[Saeid-Za]

I used oha to benchmark a simple GET endpoint built with both node-cluster and bun-cluster.
Here is the index.ts in routes:

let number = 1

export default eventHandler(async () => {
  return "Hello World! (request number: " + number++ + ")";
});

There were some unexpected/interesting results.

1. Using bun to run node-cluster : 31K Req/s

2. Using Bun to run bun-cluster : 61K Req/s

3. Using node to run node-cluster: 72K Req/s

---

This means than if we set Bun running with bun-cluster as baseline:

1. bun with bun-cluster is slower than node with node-cluster (14.55% slower)
2. bun with bun-cluster is faster than running node-cluster preset (93.91% faster)

Additional Context

1. The benchmark results might differ depending on your system's configuration. If possible, I kindly suggest testing it yourself
2. I believe that using a simple GET endpoint for benchmarking does not accurately reflect real-world scenarios
3. There were two runs in each scenario, one warm-up and the second for the data collection.
4. I noticed that this route function was using async but did not need it, I also tested without async, there were no significant diffrence.

[Saeid-Za]

Based on the benchmark data, it appears that using the node-cluster preset with the Node.js runtime is more performant than Bun, even when utilizing Bun's own HTTP server and clustering mechanism. This contradicts claims that Bun's approach is faster.

Given this information, it seems that this pull request primarily adds code complexity without providing significant benefits. We could revisit this PR at a later stage after additional releases from Bun to see if future benchmark results change.

I will close this for now.

Thank you for your input!