@futurelastic/colab-handbook 1.12.0-rc.43 → 1.12.0-rc.44

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/package.json CHANGED
@@ -27,5 +27,5 @@
27
27
  "engines": {
28
28
  "node": ">=18"
29
29
  },
30
- "version": "1.12.0-rc.43"
30
+ "version": "1.12.0-rc.44"
31
31
  }
@@ -361,6 +361,15 @@ jobs:
361
361
  if: steps.scripts.outputs.lint == 'true'
362
362
  run: npm run lint
363
363
 
364
+ # EDIT: cap your test runner's workers in CI, in its own config (#478). vitest,
365
+ # jest, node:test and playwright start one worker per CPU the job can SEE. On a
366
+ # self-hosted pool slot that is often the whole host's count, while the slot
367
+ # caps only memory. A host upgrade then adds workers until the slot is
368
+ # throttled. There is no OOM, the tests time out, and trunk goes red with no
369
+ # code change. Use a fixed number sized to the slot's memory, e.g. vitest
370
+ # `maxWorkers: process.env.CI ? 4 : undefined`, jest `--maxWorkers=4`,
371
+ # node:test `--test-concurrency=4`. GitHub Actions sets CI=true. Why and how
372
+ # to size it: CONVENTIONS.md §7, "Self-hosted runners".
364
373
  - name: Test
365
374
  if: steps.scripts.outputs.test == 'true'
366
375
  run: npm test