Last active
April 19, 2024 19:46
-
-
Save imhoffd/b4ca0a94c2496d81303ebb00063a863d to your computer and use it in GitHub Desktop.
Parallelizing Jest in GitHub Actions
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
name: CI | |
on: [push] | |
jobs: | |
setup: | |
runs-on: ubuntu-latest | |
outputs: | |
test-chunks: ${{ steps['set-test-chunks'].outputs['test-chunks'] }} | |
test-chunk-ids: ${{ steps['set-test-chunk-ids'].outputs['test-chunk-ids'] }} | |
steps: | |
- uses: actions/checkout@v2 | |
- run: npm install | |
- id: set-test-chunks | |
name: Set Chunks | |
run: echo "::set-output name=test-chunks::$(npx jest --listTests --json | jq -cM '[_nwise(length / 2 | ceil)]')" | |
- id: set-test-chunk-ids | |
name: Set Chunk IDs | |
run: echo "::set-output name=test-chunk-ids::$(echo $CHUNKS | jq -cM 'to_entries | map(.key)')" | |
env: | |
CHUNKS: ${{ steps['set-test-chunks'].outputs['test-chunks'] }} | |
test: | |
runs-on: ubuntu-latest | |
name: test (chunk ${{ matrix.chunk }}) | |
needs: | |
- setup | |
strategy: | |
matrix: | |
chunk: ${{ fromJson(needs.setup.outputs['test-chunk-ids']) }} | |
steps: | |
- uses: actions/checkout@v2 | |
- run: npm install | |
- name: jest | |
run: echo $CHUNKS | jq '.[${{ matrix.chunk }}] | .[] | @text' | xargs npx jest | |
env: | |
CHUNKS: ${{ needs.setup.outputs['test-chunks'] }} |
@imhoffd nice, two questions
- how do you make checks with dynamic names (test (chunk 0)) as a required check?
- if you know that GHA already has 2 cores, why do you need to calculate the number of runs dynamically?
oh and one more: you need to run npm install
at least 3 times, do you think it's faster than jest?
- My understanding is the check has to actually run first, then it will appear in Repo Settings -> Branches -> Edit -> Require status checks to pass before merging automatically.
- I think you may be conflating concurrency with parallelism. Jest will automatically use up the available cores on the VM. This script will run jest tests on separate VMs entirely.
- You can cache your
node_modules
directory in Github Actions if you find it's becoming a problem. There will never be a silver bullet for "fastest performance". You have to do your own testing for your own test suite.
Thanks, @imhoffd! I managed to solve those problems with caching, as for required checks, I just made chunks static, so I know how many of those I'm going to use
FYI, Jest is potentially getting a --shard
option which makes this way easier: jestjs/jest#12546
Thanks! seems it will be available in jest@28
i got this error when using the above github actions
Error reading JToken from JsonReader. Path '', line 0, position 0.,.github/workflows/chunks.yml (Line: 75, Col: 16): Unexpected type of value '', expected type: Sequence.
any idea how to resolve this?
Do you have to specify the length manually, or is there a way to set up that automatically? For instance, for every 100 tests, create a new test chunk
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
@imhoffd dude this is awesome