---
title: "Comment by Node Edge Ops on \"Node.js Internals: Event Loop & Thread Pool Explained\" | Hashnode"
description: "Treat the event loop and the libuv thread pool as separate capacity budgets. Keep CPU-bound and large sync filesystem work off the main tick, and raise the pool"
url: "https://hashnode.com/posts/node-js-internals-event-loop-thread-pool-explained/69d2076fc5b050516ee95103/comment/6aec7e82520975999b9cc2e1"
---

[Discussion](https://hashnode.com/posts/node-js-internals-event-loop-thread-pool-explained/69d2076fc5b050516ee95103)

Comment [Node.js Internals: Event Loop & Thread Pool Explained](https://hashnode.com/posts/node-js-internals-event-loop-thread-pool-explained/69d2076fc5b050516ee95103)

[Node Edge Ops](https://hashnode.com/@nodeedgeops)

[Nov 5](https://hashnode.com/posts/node-js-internals-event-loop-thread-pool-explained/69d2076fc5b050516ee95103/comment/6aec7e82520975999b9cc2e1)

Treat the event loop and the libuv thread pool as separate capacity budgets. Keep CPU-bound and large sync filesystem work off the main tick, and raise the pool size only when you have measured fs or crypto saturation — a larger pool does not help pure JavaScript CPU. Prefer streaming and async APIs so handlers return quickly; long sync loops starve timers and inbound accepts even when the process looks idle on CPU charts.

Ops prove-out after a change that touches heavy I/O or crypto: run a mixed load of light request paths and the heavy path while watching event-loop lag and p99 on the light routes. Lag should stay flat if offload worked. Saturate the heavy path alone and confirm pool wait metrics climb before the main thread stalls. If light routes still jitter, something sync is still on the main tick — shrinking concurrency alone will not fix it.
