Node.js worker threads in production: what's easy, what's not (opens in new tab)

Spawning a worker is easy. Making it reliable means designing a protocol, packaging the worker file, proxying callbacks, handling crashes, and shutting down cleanly.