feat(infra): Add backend selective testing workflow

.github/workflows/backend-selective.yml

@@ -0,0 +1,263 @@
+name: backend (selective)
+
+on:
+  pull_request:
+
+# Cancel in progress workflows on pull_requests.
+# https://docs.github.com/en/actions/using-jobs/using-concurrency#example-using-a-fallback-value
+concurrency:
+  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
+
+# hack for https://github.com/actions/cache/issues/810#issuecomment-1222550359
+env:
+  SEGMENT_DOWNLOAD_TIMEOUT_MINS: 3
+  SNUBA_NO_WORKERS: 1
+
+jobs:
+  files-changed:
+    name: detect what files changed
+    runs-on: ubuntu-24.04
+    timeout-minutes: 3
+    outputs:
+      api_docs: ${{ steps.changes.outputs.api_docs }}
+      backend: ${{ steps.changes.outputs.backend_all }}
+      backend_dependencies: ${{ steps.changes.outputs.backend_dependencies }}
+      backend_api_urls: ${{ steps.changes.outputs.backend_api_urls }}
+      backend_any_type: ${{ steps.changes.outputs.backend_any_type }}
+      migration_lockfile: ${{ steps.changes.outputs.migration_lockfile }}
+    steps:
+      - uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
+
+      - name: Check for backend file changes
+        uses: dorny/paths-filter@0bc4621a3135347011ad047f9ecf449bf72ce2bd # v3.0.0
+        id: changes
+        with:
+          token: ${{ github.token }}
+          filters: .github/file-filters.yml
+
+  prepare-selective-tests:
+    if: needs.files-changed.outputs.backend == 'true'
+    needs: files-changed
+    name: prepare selective tests
+    runs-on: ubuntu-24.04
+    timeout-minutes: 10
+    permissions:
+      contents: read
+      id-token: write
+    outputs:
+      has-coverage: ${{ steps.find-coverage.outputs.found }}
+      coverage-sha: ${{ steps.find-coverage.outputs.coverage-sha }}
+      changed-files: ${{ steps.changed-files.outputs.files }}
+      test-count: ${{ steps.compute-tests.outputs.test-count }}
+      has-selected-tests: ${{ steps.compute-tests.outputs.has-selected-tests }}
+    steps:
+      - uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
+        with:
+          fetch-depth: 0 # Need full history for git diff
+
+      - name: Setup Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.13.1'
+
+      - name: Authenticate to Google Cloud
+        id: gcloud-auth
+        uses: google-github-actions/auth@v2
+        with:
+          project_id: sentry-dev-tooling
+          workload_identity_provider: ${{ secrets.SENTRY_GCP_DEV_WORKLOAD_IDENTITY_POOL }}
+          service_account: ${{ secrets.COLLECT_TEST_DATA_SERVICE_ACCOUNT_EMAIL }}
+
+      - name: Find coverage data for selective testing
+        id: find-coverage
+        env:
+          GCS_BUCKET: sentry-coverage-data
+        run: |
+          set -euo pipefail
+
+          # Get the base commit (what the PR branches from)
+          BASE_SHA="${{ github.event.pull_request.base.sha }}"
+
+          echo "Looking for coverage data starting from base commit: $BASE_SHA"
+
+          COVERAGE_SHA=""
+          for sha in $(git rev-list "$BASE_SHA" --max-count=30); do
+            # Check if coverage exists in GCS for this commit
+            if gcloud storage ls "gs://${GCS_BUCKET}/${sha}/" &>/dev/null; then
+              COVERAGE_SHA="$sha"
+              echo "Found coverage data at commit: $sha"
+              break
+            fi
+            echo "No coverage at $sha, checking parent..."
+          done
+
+          if [[ -z "$COVERAGE_SHA" ]]; then
+            echo "No coverage found in last 30 commits, will run full test suite"
+            echo "found=false" >> "$GITHUB_OUTPUT"
+          else
+            echo "found=true" >> "$GITHUB_OUTPUT"
+            echo "coverage-sha=$COVERAGE_SHA" >> "$GITHUB_OUTPUT"
+          fi
+
+      - name: Download coverage database
+        id: download-coverage
+        if: steps.find-coverage.outputs.found == 'true'
+        env:
+          COVERAGE_SHA: ${{ steps.find-coverage.outputs.coverage-sha }}
+        run: |
+          set -euxo pipefail
+          mkdir -p .coverage
+          gcloud storage cp "gs://sentry-coverage-data/${COVERAGE_SHA}/*" .coverage/ || true
+
+          # Find the coverage file (could be .coverage.* format)
+          COVERAGE_FILE=$(ls .coverage/.coverage.* 2>/dev/null | head -1 || true)
+          if [[ -z "$COVERAGE_FILE" ]]; then
+            echo "Warning: No coverage file found in downloaded data"
+            ls -la .coverage/ || true
+            echo "coverage-file=" >> "$GITHUB_OUTPUT"
+          else
+            echo "Downloaded coverage file: $COVERAGE_FILE"
+            echo "coverage-file=$COVERAGE_FILE" >> "$GITHUB_OUTPUT"
+          fi
+
+      - name: Get changed files
+        id: changed-files
+        run: |
+          # Get files changed between base and head of PR
+          BASE_SHA="${{ github.event.pull_request.base.sha }}"
+          HEAD_SHA="${{ github.event.pull_request.head.sha }}"
+
+          CHANGED_FILES=$(git diff --name-only "$BASE_SHA" "$HEAD_SHA" | tr '\n' ' ')
+          echo "Changed files: $CHANGED_FILES"
+          echo "files=$CHANGED_FILES" >> "$GITHUB_OUTPUT"
+
+      - name: Compute selected tests
+        id: compute-tests
+        if: steps.download-coverage.outputs.coverage-file != ''
+        run: |
+          python3 .github/workflows/scripts/compute-selected-tests.py \
+            --coverage-db "${{ steps.download-coverage.outputs.coverage-file }}" \
+            --changed-files "${{ steps.changed-files.outputs.files }}" \
+            --output .artifacts/selected-tests.txt \
+            --github-output
+
+      - name: Upload coverage database artifact
+        if: steps.download-coverage.outputs.coverage-file != ''
+        uses: actions/upload-artifact@v4
+        with:
+          name: coverage-db-${{ github.run_id }}
+          path: .coverage/
+          retention-days: 1
+          include-hidden-files: true
+
+      - name: Upload selected tests artifact
+        if: steps.compute-tests.outputs.has-selected-tests == 'true'
+        uses: actions/upload-artifact@v4
+        with:
+          name: selected-tests-${{ github.run_id }}
+          path: .artifacts/selected-tests.txt
+          retention-days: 1
+
+  calculate-shards:
+    if: needs.files-changed.outputs.backend == 'true'
+    needs: [files-changed, prepare-selective-tests]
+    name: calculate test shards
+    runs-on: ubuntu-24.04
+    timeout-minutes: 5
+    outputs:
+      shard-count: ${{ steps.calculate-shards.outputs.shard-count }}
+      shard-indices: ${{ steps.calculate-shards.outputs.shard-indices }}
+    steps:
+      - uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
+
+      - name: Setup sentry env
+        uses: ./.github/actions/setup-sentry
+        id: setup
+        with:
+          mode: backend-ci
+          skip-devservices: true
+
+      - name: Download selected tests artifact
+        if: needs.prepare-selective-tests.outputs.has-selected-tests == 'true'
+        uses: actions/download-artifact@v4
+        with:
+          name: selected-tests-${{ github.run_id }}
+          path: .artifacts/
+
+      - name: Calculate test shards
+        id: calculate-shards
+        env:
+          SELECTED_TESTS_FILE: ${{ needs.prepare-selective-tests.outputs.has-selected-tests == 'true' && '.artifacts/selected-tests.txt' || '' }}
+          SELECTED_TEST_COUNT: ${{ needs.prepare-selective-tests.outputs.test-count }}
+        run: |
+          python3 .github/workflows/scripts/calculate-backend-test-shards.py
+
+  backend-test-selective:
+    if: needs.files-changed.outputs.backend == 'true'
+    needs: [files-changed, prepare-selective-tests, calculate-shards]
+    name: backend test
+    runs-on: ubuntu-24.04
+    timeout-minutes: 60
+    permissions:
+      contents: read
+      id-token: write
+      actions: read
+    strategy:
+      fail-fast: false
+      matrix:
+        instance: ${{ fromJSON(needs.calculate-shards.outputs.shard-indices) }}
+
+    env:
+      MATRIX_INSTANCE_TOTAL: ${{ needs.calculate-shards.outputs.shard-count }}
+      TEST_GROUP_STRATEGY: roundrobin
+
+    steps:
+      - uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
+        with:
+          fetch-depth: 0
+
+      - name: Setup sentry env
+        uses: ./.github/actions/setup-sentry
+        id: setup
+        with:
+          mode: backend-ci
+
+      - name: Download coverage database artifact
+        if: needs.prepare-selective-tests.outputs.has-coverage == 'true'
+        uses: actions/download-artifact@v4
+        with:
+          name: coverage-db-${{ github.run_id }}
+          path: .coverage/
+
+      - name: Find coverage file
+        id: find-coverage-file
+        if: needs.prepare-selective-tests.outputs.has-coverage == 'true'
+        run: |
+          COVERAGE_FILE=$(ls .coverage/.coverage.* 2>/dev/null | head -1 || true)
+          echo "coverage-file=$COVERAGE_FILE" >> "$GITHUB_OUTPUT"
+
+      - name: Run backend tests (${{ steps.setup.outputs.matrix-instance-number }} of ${{ steps.setup.outputs.matrix-instance-total }})
+        id: run_backend_tests
+        run: make test-python-ci
+        env:
+          CHANGED_FILES: ${{ needs.prepare-selective-tests.outputs.changed-files }}
+          COVERAGE_DB_PATH: ${{ steps.find-coverage-file.outputs.coverage-file }}
+
+      - name: Inspect failure
+        if: failure()
+        run: |
+          if command -v devservices; then
+            devservices logs
+          fi
+
+      - name: Collect test data
+        uses: ./.github/actions/collect-test-data
+        if: ${{ !cancelled() }}
+        with:
+          artifact_path: .artifacts/pytest.json
+          gcs_bucket: ${{ secrets.COLLECT_TEST_DATA_GCS_BUCKET }}
+          gcp_project_id: ${{ secrets.COLLECT_TEST_DATA_GCP_PROJECT_ID }}
+          workload_identity_provider: ${{ secrets.SENTRY_GCP_DEV_WORKLOAD_IDENTITY_POOL }}
+          service_account_email: ${{ secrets.COLLECT_TEST_DATA_SERVICE_ACCOUNT_EMAIL }}
+          matrix_instance_number: ${{ steps.setup.outputs.matrix-instance-number }}

.github/workflows/scripts/calculate-backend-test-shards.py

@@ -5,17 +5,20 @@
 import re
 import subprocess
 import sys
+from pathlib import Path
 
 TESTS_PER_SHARD = 1200
 MIN_SHARDS = 1
 MAX_SHARDS = 22
 DEFAULT_SHARDS = 22
 
-PYTEST_ARGS = [
+PYTEST_BASE_ARGS = [
     "pytest",
     "--collect-only",
     "--quiet",
-    "tests",
+]
+
+PYTEST_IGNORE_ARGS = [
     "--ignore=tests/acceptance",
     "--ignore=tests/apidocs",
     "--ignore=tests/js",
@@ -24,9 +27,30 @@
 
 
 def collect_test_count():
+    """Collect test count, optionally filtering to selected test files."""
+    selected_tests_file = os.environ.get("SELECTED_TESTS_FILE")
+
+    if selected_tests_file:
+        path = Path(selected_tests_file)
+        if not path.exists():
+            print(f"Selected tests file not found: {selected_tests_file}", file=sys.stderr)
+            return None
+
+        with path.open() as f:
+            selected_files = [line.strip() for line in f if line.strip()]
+
+        if not selected_files:
+            print("No selected test files, running 0 tests", file=sys.stderr)
+            return 0
+
+        print(f"Counting tests in {len(selected_files)} selected files", file=sys.stderr)
+        pytest_args = PYTEST_BASE_ARGS + selected_files
+    else:
+        pytest_args = PYTEST_BASE_ARGS + ["tests"] + PYTEST_IGNORE_ARGS
+
     try:
         result = subprocess.run(
-            PYTEST_ARGS,
+            pytest_args,
             capture_output=True,
             text=True,
             check=False,
@@ -40,7 +64,6 @@ def collect_test_count():
             print(f"Collected {count} tests", file=sys.stderr)
             return count
 
-        # If no match, check if pytest failed
         if result.returncode != 0:
             print(
                 f"Pytest collection failed (exit {result.returncode})",
@@ -85,7 +108,6 @@ def calculate_shards(test_count):
 def main():
     test_count = collect_test_count()
     shard_count = calculate_shards(test_count)
-    # Generate a JSON array of shard indices [0, 1, 2, ..., shard_count-1]
     shard_indices = json.dumps(list(range(shard_count)))
 
     github_output = os.getenv("GITHUB_OUTPUT")