jest heap out of memory on large test suites

TL;DR

Each Jest worker has a heap limit, and large suites exceed it. Raise the limit with NODE_OPTIONS, reduce workers so each gets more room, and find the suite that leaks.

Error

FATAL ERROR: Reached heap limit Allocation failed - JavaScript heap out of memory

Steps

1. Run with NODE_OPTIONS=--max-old-space-size=4096 npx jest. Expected: the run gets further.
2. Reduce workers: npx jest --maxWorkers=2. Expected: fewer workers means more memory each.
3. Find the leaking suite: run files individually and watch memory. Expected: one file grows without bound.
4. Common leaks: huge snapshots, unclosed servers, module-level caches. Fix the suite. Expected: memory returns to baseline.
5. For permanently large suites, split the CI job by path. Expected: each job fits in memory.

When to use

• heap out of memory naming node/Jest.
• Suites with big snapshots or many files.

When not to use

• Worker child-process exceptions without OOM (worker skill).
• Cypress heap OOM (different process model).

Tool compatibility

• Jest 27 through 30; NODE_OPTIONS, --maxWorkers.

Variant phrasings

Jest ran out of memory in CI

CI runners have less RAM; the same suite passes on a laptop.

Allocation failed in Jest worker

The worker-specific form; reduce workers first.

Why it happens

Workers share the machine's RAM. Big suites plus many workers exceed it, and the default heap limit is modest.

Edge cases

• --logHeapUsage shows per-suite memory; use it to find the leak.
• --workerIdleMemoryLimit (Jest 29+) restarts workers that grow too big automatically.
• Snapshot files of many megabytes are a common hidden cause.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_8Ai2oRvBkobxGksj0u0uuQ