perftest.yml

# This workflow runs the SimpleStorage performance test
name: Performance Test

on:
  push:
    branches:
      - main
      - v0.*
  pull_request:
    branches:
      - main
      - v0.*

jobs:
  perftest:
    runs-on: ubuntu-latest
    timeout-minutes: 10
    strategy:
      fail-fast: false
      matrix:
        python-version: ["3.11"]

    steps:
    - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@0b93645e9fea7318ecaed2b359559ac225c90a2b # v5.3.0
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
        pip install -e .
    - name: Start Ray cluster
      run: |
        # Get the host IP
        HOST_IP=$(hostname -I | awk '{print $1}')
        echo "Host IP: $HOST_IP"
        # Start Ray with node resource
        ray start --head --resources="{\"node:$HOST_IP\":1}"
        ray status
    - name: Run SimpleStorage performance test
      run: |
        # Get the host IP
        HOST_IP=$(hostname -I | awk '{print $1}')
        echo "Host IP: $HOST_IP"
        # Run the perftest with small batch size for quick test
        cd scripts/performance_test
        python perftest.py \
          --backend_config=../../transfer_queue/config.yaml \
          --device=cpu \
          --global_batch_size=128 \
          --field_num=4 \
          --seq_len=1024 \
          --head_node_ip=$HOST_IP \
          --output_csv=results.csv
    - name: Stop Ray cluster
      run: |
        ray stop
      if: always()