docker-network-health.yml

name: Docker Network Health Check

on:
  workflow_dispatch:
  pull_request:
    paths:
      - '**'
      - '!documentation/**'
      - '!scripts/**'
      - '!tools/**'
      - 'tools/genesis-snapshot/**'
  push:
    branches:
      - develop

concurrency: 
  group: run-and-check-group
  cancel-in-progress: false

jobs:
  run-and-check:
    runs-on: self-hosted

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

    - name: Run network, wait and check health
      run: |
        set -x

        # Run network
        cd ./tools/docker-network
        timeout 10m ./run.sh &
        RUN_PID=$!
        
        # Wait for node-4 to be created before querying it
        timeout 10m bash -c 'until docker ps | grep docker-network-node-4; do sleep 5; done' &

        # Wait for any of the two processes to exit
        wait -n || exit 1
        
        # Additional 10 seconds wait to allow the API to come up
        sleep 10

        # Health check
        SUCCESS=false
        while true; do
          OUTPUT=$(curl -o /dev/null -s -w "%{http_code}\n" http://localhost:8080/health)
          if [[ $OUTPUT -eq 200 ]]; then
            SUCCESS=true
            kill -s SIGINT $RUN_PID
            break
          # curl will return a connection refused when the network is tear down from the timeout.
          elif [[ $OUTPUT -eq 000 ]]; then
            echo "Connection refused. Failing the action."
            break
          fi
          sleep 5
        done

        if [[ ! $SUCCESS ]]; then
          echo "Health check never returned 200. Failing the action."
          exit 1
        fi

    - name: Cleanup
      run: |
        cd ./tools/docker-network
        docker compose kill || true
        docker compose down -t 1 -v || true