commit-metadata.yml

# Uploads commit metadata (an empty ingest envelope -- no benchmark records) on
# every push to develop, to both the v3 server and the v4 Postgres, so the
# `commits` dim stays populated even when no benchmark ran. Needed by both
# backends, hence not named for either.

name: Commit metadata

on:
  push:
    branches: [develop]
  workflow_dispatch: { }

permissions:
  id-token: write  # enables AWS-GitHub OIDC for the best-effort v4 ingest step
  contents: read

jobs:
  commit-metadata:
    runs-on: ubuntu-latest
    timeout-minutes: 10
    steps:
      - uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10  # v6
        with:
          fetch-depth: 2

      - name: Ingest commit metadata to v3 server
        if: vars.V3_INGEST_URL != ''
        shell: bash
        env:
          INGEST_BEARER_TOKEN: ${{ secrets.INGEST_BEARER_TOKEN }}
        run: |
          echo -n > empty.jsonl
          python3 scripts/post-ingest.py empty.jsonl \
            --server "${{ vars.V3_INGEST_URL }}" \
            --commit-sha "${{ github.sha }}" \
            --benchmark-id "commit-metadata" \
            --repo-url "${{ github.server_url }}/${{ github.repository }}"

      # v4 (Postgres) dual-write -- BEST-EFFORT (see bench.yml rationale). Empty records:
      # post-ingest.py --postgres upserts the commit row only. v3 above stays required;
      # a v4 failure never fails the job (promoted to required at cutover, PR-5.1).
      # Gated on the ingest-role ARN var (the assume-role input that MUST exist) so
      # it no-ops until v4 infra is wired.
      #
      # ORDER MATTERS: "Install uv" runs BEFORE "Configure AWS credentials".
      # configure-aws-credentials persists the assumed ingest-role (rds-db:connect
      # only) as the job's ambient AWS creds; the uv setup compiles via sccache
      # (S3-backed), so running it after the role switch fails with S3 AccessDenied.
      # Installing uv first keeps sccache on the original S3-capable creds; the role
      # is assumed immediately before the ingest, which needs only rds-db:connect.
      - name: Install uv for v4 ingest
        if: vars.GH_BENCH_INGEST_ROLE_ARN != ''
        continue-on-error: true
        uses: spiraldb/actions/.github/actions/setup-uv@a746510eafaa926484c354541cfc49b2ec06cc63  # 0.18.6
      - name: Configure AWS credentials for v4 ingest (OIDC)
        if: vars.GH_BENCH_INGEST_ROLE_ARN != ''
        continue-on-error: true
        uses: aws-actions/configure-aws-credentials@e7f100cf4c008499ea8adda475de1042d6975c7b  # v6
        with:
          role-to-assume: ${{ vars.GH_BENCH_INGEST_ROLE_ARN }}
          aws-region: ${{ vars.RDS_BENCH_REGION }}
      - name: Ingest commit metadata to v4 Postgres (best-effort)
        if: vars.GH_BENCH_INGEST_ROLE_ARN != ''
        continue-on-error: true
        shell: bash
        env:
          RDS_BENCH_INSTANCE_ENDPOINT: ${{ vars.RDS_BENCH_INSTANCE_ENDPOINT }}
          RDS_BENCH_DB_NAME: ${{ vars.RDS_BENCH_DB_NAME }}
          AWS_REGION: ${{ vars.RDS_BENCH_REGION }}
          BENCH_SITE_BASE_URL: ${{ vars.BENCH_SITE_BASE_URL }}
          BENCH_REVALIDATE_TOKEN: ${{ secrets.BENCH_REVALIDATE_TOKEN }}
        run: |
          set -Eeuo pipefail
          echo -n > empty.jsonl
          curl -fsSL https://truststore.pki.rds.amazonaws.com/global/global-bundle.pem \
            -o "${RUNNER_TEMP}/rds-global-bundle.pem"
          DSN="postgresql://bench_ingest@${RDS_BENCH_INSTANCE_ENDPOINT}:5432/${RDS_BENCH_DB_NAME}?sslmode=verify-full&sslrootcert=${RUNNER_TEMP}/rds-global-bundle.pem"
          uv run --no-project --with 'psycopg[binary]' --with boto3 --with xxhash \
            scripts/post-ingest.py empty.jsonl \
            --postgres "${DSN}" \
            --commit-sha "${{ github.sha }}" \
            --region "${AWS_REGION}"