tpch.yml

name: TPC-H Benchmarks
on:
  workflow_dispatch:
    inputs:
      scale:
        description: 'Scale Factor'
        required: true
        default: 10000
        type: choice
        options:
          - 10000
          - 1000
          - 100
          - 10
          - 1
      dask:
        description: 'Dask'
        required: true
        default: true
        type: boolean
      duckdb:
        description: 'DuckDB'
        required: true
        default: true
        type: boolean
      polars:
        description: 'Polars'
        required: true
        default: false
        type: boolean
      pyspark:
        description: 'PySpark'
        required: true
        default: true
        type: boolean

# concurrency:
#   # Only allow a single run at a time to rate limiting
#   group: tpch

defaults:
  # Required shell entrypoint to have properly activated conda environments
  run:
    shell: bash -l {0}

jobs:
  tpch:
    name: TPC-H
    runs-on: ubuntu-latest

    steps:
    - name: Checkout
      uses: actions/checkout@v4

    - name: Set up environment
      uses: conda-incubator/setup-miniconda@v3
      with:
        miniforge-variant: Mambaforge
        use-mamba: true
        condarc-file: ci/condarc
        python-version: "3.9"
        environment-file: ci/environment.yml

    - name: Pip Compile
      run: pip-compile ci/requirements-2tpch-non-dask.in

    - name: Add Environment dependencies
      run: python -m pip install -r ci/requirements-2tpch-non-dask.txt

    - name: Dump environment
      run: |
        # For debugging
        echo -e "--\n--Conda Environment (re-create this with \`conda env create --name <name> -f <output_file>\`)\n--"
        mamba env export | grep -E -v '^prefix:.*$'

    - name: Add Dask to benchmark if enabled
      if: ${{ inputs.dask }}
      run: |
        echo PYTEST_BENCHMARKS="${{ env.PYTEST_BENCHMARKS }} tests/tpch/test_dask.py" >> $GITHUB_ENV

    - name: Add DuckDB to benchmark if enabled
      if: ${{ inputs.duckdb }}
      run: |
        echo PYTEST_BENCHMARKS="${{ env.PYTEST_BENCHMARKS }} tests/tpch/test_duckdb.py" >> $GITHUB_ENV

    - name: Add Polars to benchmark if enabled
      if: ${{ inputs.polars }}
      run: |
        echo PYTEST_BENCHMARKS="${{ env.PYTEST_BENCHMARKS }} tests/tpch/test_polars.py" >> $GITHUB_ENV

    - name: Add PySpark to benchmark if enabled
      if: ${{ inputs.pyspark }}
      run: |
        echo PYTEST_BENCHMARKS="${{ env.PYTEST_BENCHMARKS }} tests/tpch/test_pyspark.py" >> $GITHUB_ENV

    - name: Run TPC-H benchmarks (except polars)
      env:
        DASK_COILED__TOKEN: ${{ secrets.COILED_BENCHMARK_BOT_TOKEN }}
        AWS_ACCESS_KEY_ID: ${{ secrets.RUNTIME_CI_BOT_AWS_ACCESS_KEY_ID }}
        AWS_DEFAULT_REGION: us-east-2 # this is needed for boto for some reason
        AWS_SECRET_ACCESS_KEY: ${{ secrets.RUNTIME_CI_BOT_AWS_SECRET_ACCESS_KEY }}
        COILED_RUNTIME_VERSION: ${{ matrix.runtime-version }}
        DB_NAME: tpch_${{ inputs.scale }}.db
        CLUSTER_DUMP: always
        DASK_DATAFRAME__QUERY_PLANNING: True
      run: |
        pytest --benchmark \
            ${{ env.PYTEST_BENCHMARKS }} \
            -n 4 --dist loadscope \
            --scale ${{ inputs.scale }}

    - name: Upload benchmark results
      uses: actions/upload-artifact@v4
      if: always()
      with:
        name: tpch-benchmark
        path: |
          tpch_${{ inputs.scale }}.db
          mamba_env_export.yml