Node.js worker_threads Out of Memory - Fix Worker Heap Crash in CI

Each Worker runs its own V8 isolate with its own heap. A worker can exhaust memory and abort even while the main thread looks fine - and the limit you set on the parent does not automatically apply to the worker.

What this error means

A job using worker_threads (or a tool that spawns workers, like some test runners and bundlers) crashes with a worker-side heap-out-of-memory or ERR_WORKER_OUT_OF_MEMORY. The parent process did not hit its own limit; the worker did.

<--- Last few GCs --->
FATAL ERROR: Reached heap limit Allocation failed -
JavaScript heap out of memory
Error [ERR_WORKER_OUT_OF_MEMORY]: Worker terminated due to reaching
memory limit: JS heap out of memory

Common causes

The worker has its own heap limit

Setting --max-old-space-size on the parent does not raise the worker’s limit. A worker created with a low resourceLimits.maxOldGenerationSizeMb (or the default) hits its own ceiling.

Too many concurrent workers for the runner

Spawning many workers in parallel multiplies heap usage; collectively they exceed the runner’s physical memory and one is OOM-terminated.

How to fix it

Raise the worker’s own limit

Set resourceLimits when creating the Worker, or pass execArgv to the worker.

new Worker('./task.js', {
  resourceLimits: { maxOldGenerationSizeMb: 2048 },
})

Reduce concurrency and peak memory

1. Lower the number of workers a pool spawns at once.
2. Stream or chunk large inputs instead of holding everything in the worker heap.
3. Use a runner with more RAM if the workload genuinely needs it.

How to prevent it

• Size worker resourceLimits relative to runner memory.
• Cap worker-pool concurrency in CI.
• Process large data in chunks rather than all at once.

Frequently asked questions

Related guides