Skip to content

perf(ci): split Test Web over two runners with vitest --shard (modelled: 56 s median off a web PR's longest job, 109 s on web-only PRs) #31035

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Pyseph wants to merge 3 commits into from
+12 −3
Closed

perf(ci): split Test Web over two runners with vitest --shard (modelled: 56 s median off a web PR's longest job, 109 s on web-only PRs) #31035

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
052b5f5
perf(ci): split Test Web over two runners with vitest --shard
Pyseph Nov 4, 2026
b979a85
Merge origin/main into perf/ci-test-web-shards
Pyseph Nov 5, 2026
cc87444
Merge remote-tracking branch 'github/main' into perf/ci-test-web-shards
maria-rcks Nov 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 12 additions & 3 deletions .github/workflows/ci.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -322,16 +322,25 @@ jobs:

# The web suite is the largest outside the server. In the shared Test job it
# started last and ran alone for about 40s after everything else finished.
# release.yml mirrors this test, test_web, and test_server split.
# It is also the longest job of web pull requests, so vitest's `--shard` splits
# its test files (unit and browser alike, by a hash of the path) over two
# runners. The shards repeat only checkout, install and the Chromium restore,
# about 20s. The `Check` job needs `test_web` as a whole, so this adds no
# required status. release.yml mirrors the test, test_web, and test_server
# split but not the web shards: it is not a pull request wait.
test_web:
permissions:
contents: read
id-token: write
name: Test Web
name: Test Web ${{ matrix.shard }}
needs: changes
if: ${{ !cancelled() && needs.changes.outputs.test_web != 'false' }}
runs-on: blacksmith-8vcpu-ubuntu-2404
timeout-minutes: 10
strategy:
fail-fast: false
matrix:
shard: [1, 2]
steps:
- name: Checkout
uses: actions/checkout@v6
Expand Down Expand Up @@ -373,7 +382,7 @@ jobs:
fi

- name: Test
run: vp run --filter @t3tools/web test
run: vp run --filter @t3tools/web test --shard ${{ matrix.shard }}/${{ strategy.job-total }}

# apps/server sets `fileParallelism: false`, so its test files run strictly
# one at a time. Sharding spreads them over separate runners instead of
Expand Down
Loading