Pipeline work

.github/actions/build-action/action.yaml

@@ -29,7 +29,7 @@ runs:
   using: composite
   steps:
     - name: Common SSH and Azure CLI Setup
-      uses: ./.github/actions/common-steps
+      uses: ./.github/actions/common-ssh-and-azure-steps
       with:
         VM_IP: ${{ inputs.VM_IP }}
         SSH_PRIVATE_KEY: ${{ inputs.SSH_PRIVATE_KEY }}
@@ -44,61 +44,4 @@ runs:
 
     - name: Build project
       shell: bash
-      run: |
-        ssh -T -o ConnectTimeout=10 ${{ inputs.VM_USERNAME }}@${{ inputs.VM_IP }} << 'EOF'
-          set -e # Exit on error
-          cd ~/aperi-mech
-
-          compose_file=docker-compose.yml
-          service_name=aperi-mech-development
-          # Check if GPU build is enabled and code coverage is not enabled, overriding gpu flag if code coverage is enabled
-          if [ ${{ inputs.gpu }} == 'true' ] && [ ${{ inputs.code-coverage }} == 'false' ]; then
-            compose_file=docker-compose_nvidia_t4_gpu.yml
-            service_name=aperi-mech-gpu-development
-          fi
-          # Debugging
-          echo "On VM, Compose file: $compose_file"
-          echo "On VM, Service name: $service_name"
-
-          docker-compose -f $compose_file run --rm $service_name /bin/bash -c '
-            configure_flag=""
-            build_root=build
-            if [ ${{ inputs.with-protego }} == "true" ]; then
-              build_root=protego-mech/build
-              configure_flag=" --protego-mech"
-            fi
-            build_path="${build_root}/${{ inputs.build-type }}"
-            if [ ${{ inputs.gpu }} == "true" ]; then
-              configure_flag="${configure_flag} --gpu"
-              build_path="${build_path}_gpu"
-            fi
-
-            # Check if code coverage is enabled, overriding gpu flag if code coverage is enabled
-            if [ ${{ inputs.code-coverage }} == "true" ]; then
-              configure_flag="${configure_flag} --code-coverage"
-              build_path="${build_path}_cov"
-            fi
-            # Debugging
-            echo "In container, inputs.build-type: ${{ inputs.build-type }}"
-            echo "In container, inputs.gpu: ${{ inputs.gpu }}"
-            echo "In container, inputs.code-coverage: ${{ inputs.code-coverage }}"
-            echo "In container, configure_flag: $configure_flag"
-            echo "In container, build_path: $build_path"
-
-            cd ~/aperi-mech
-            echo "Setting up Spack environment..."
-            . ~/spack/share/spack/setup-env.sh
-            spack env activate aperi-mech
-
-            echo "Configuring project..."
-            # do_configure will set build type to Debug if code coverage is enabled
-            ./do_configure --build-type ${{ inputs.build-type }} ${configure_flag}
-
-            echo "Building project..."
-            cd $build_path
-            make -j 4
-            if [ ${{ inputs.code-coverage }} == "true" ]; then
-              make coverage
-            fi
-          '  || { echo "Build project step failed"; exit 1; }
-        EOF
+      run: python3 ./.github/actions/build-action/build.py --ip ${{ inputs.VM_IP }} --username ${{ inputs.VM_USERNAME }} --build-type ${{ inputs.build-type }} --gpu ${{ inputs.gpu }} --code-coverage ${{ inputs.code-coverage }} --protego ${{ inputs.with-protego }}

.github/actions/build-action/build.py

@@ -0,0 +1,109 @@
+#!/usr/bin/env python3
+
+import argparse
+
+import paramiko
+
+
+def str_to_bool(value):
+    return value.lower() in ("true", "1", "t", "y", "yes")
+
+
+def run_build(vm_ip, vm_username, gpu, build_type, code_coverage, with_protego):
+    ssh = paramiko.SSHClient()
+    ssh.set_missing_host_key_policy(paramiko.AutoAddPolicy())
+
+    # Establish the SSH connection with increased timeout and keepalive options
+    ssh.connect(vm_ip, username=vm_username, timeout=60)
+    transport = ssh.get_transport()
+    transport.set_keepalive(30)  # Send keepalive messages every 30 seconds
+
+    # Construct the compose file and service name
+    compose_file = "docker-compose.yml"
+    service_name = "aperi-mech-development"
+    if gpu and not code_coverage:
+        compose_file = "docker-compose_nvidia_t4_gpu.yml"
+        service_name = "aperi-mech-gpu-development"
+
+    # Construct the configure flag and build path
+    configure_flag = ""
+    build_root = "build"
+    if with_protego:
+        build_root = "protego-mech/build"
+        configure_flag += " --protego-mech"
+    build_path = f"{build_root}/{build_type}"
+    if gpu:
+        configure_flag += " --gpu"
+        build_path += "_gpu"
+
+    run_coverage = ""
+    if code_coverage:
+        configure_flag += " --code-coverage"
+        build_path += "_cov"
+        run_coverage = "make coverage"
+
+    # Construct the command to run on the VM
+    commands = f"""
+    set -e
+    cd ~/aperi-mech
+
+    docker-compose -f {compose_file} run --rm {service_name} /bin/bash -c '
+        cd ~/aperi-mech
+        . ~/spack/share/spack/setup-env.sh
+        spack env activate aperi-mech
+
+        ./do_configure --build-type {build_type} {configure_flag}
+
+        cd {build_path}
+        make -j 4
+
+        {run_coverage}
+
+    ' || {{ echo "Build project step failed"; exit 1; }}
+    """
+
+    print("Executing the following commands on the VM:")
+    print(commands)
+
+    stdin, stdout, stderr = ssh.exec_command(commands, get_pty=True)
+    for line in stdout:
+        print(line.strip())
+    for line in stderr:
+        print(line.strip())
+
+    ssh.close()
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Run build on remote VM")
+    parser.add_argument("--ip", default="127.0.0.1", help="IP address of the VM")
+    parser.add_argument("--username", default="azureuser", help="Username for the VM")
+    parser.add_argument("--gpu", default="false", help="GPU flag (true/false)")
+    parser.add_argument(
+        "--build_type",
+        "--build-type",
+        dest="build_type",
+        default="Release",
+        help="Build type",
+    )
+    parser.add_argument(
+        "--code_coverage",
+        "--code-coverage",
+        dest="code_coverage",
+        default="false",
+        help="Code coverage flag (true/false)",
+    )
+    parser.add_argument(
+        "--protego", default="false", help="With Protego flag (true/false)"
+    )
+
+    args = parser.parse_args()
+
+    run_build(
+        args.ip,
+        args.username,
+        str_to_bool(args.gpu),
+        args.build_type,
+        str_to_bool(args.code_coverage),
+        str_to_bool(args.protego),
+    )

.github/actions/common-runner-config-steps/action.yaml

@@ -0,0 +1,15 @@
+name: Common Runner Config Steps
+description: Setup python on the runner and install dependencies
+runs:
+  using: composite
+  steps:
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: 3.x
+
+    - name: Install Dependencies
+      shell: bash
+      run: |
+        python -m pip install --upgrade pip
+        pip install paramiko

.github/actions/run-aperi-mech-performance-tests/action.yaml

@@ -24,47 +24,4 @@ runs:
 
     - name: Run aperi-mech performance tests
       shell: bash
-      run: |
-        ssh -T -o ConnectTimeout=10 ${{ inputs.VM_USERNAME }}@${{ inputs.VM_IP }} << 'EOF'
-          set -e # Exit on error
-          cd ~/aperi-mech
-
-          compose_file=docker-compose.yml
-          service_name=aperi-mech-development
-          if [ ${{ inputs.gpu }} == 'true' ]; then
-            compose_file=docker-compose_nvidia_t4_gpu.yml
-            service_name=aperi-mech-gpu-development
-          fi
-          # Debugging
-          echo "On VM, Compose file: $compose_file"
-          echo "On VM, Service name: $service_name"
-
-          docker-compose -f $compose_file run --rm $service_name /bin/bash -c '
-            test_flags="--release"
-            if [ ${{ inputs.gpu }} == "true" ]; then
-              test_flags="$test_flags --gpu"
-            else
-              test_flags="$test_flags --cpu"
-            fi
-            if [ ${{ inputs.parallel }} == "true" ]; then
-              test_flags="$test_flags --parallel"
-            else
-              test_flags="$test_flags --serial"
-            fi
-            # Debugging
-            echo "In container, inputs.gpu: ${{ inputs.gpu }}"
-            echo "In container, inputs.parallel: ${{ inputs.parallel }}"
-            echo "In container, test_flags: $test_flags"
-
-            echo "Setting up Spack environment..."
-            . ~/spack/share/spack/setup-env.sh
-            spack env activate aperi-mech
-
-            echo "Running aperi-mech performance tests..."
-            cd ~/aperi-mech/test/
-            ./run_regression_tests.py --directory ./performance_tests/aperi-mech --build-dir ~/aperi-mech/build/ $test_flags --write-json --no-preclean
-            ./run_regression_tests.py --directory ./performance_tests/aperi-mech --build-dir ~/aperi-mech/build/ $test_flags --clean-logs
-            ./run_regression_tests.py --directory ./performance_tests/aperi-mech --build-dir ~/aperi-mech/build/ $test_flags --clean-results
-
-          '  || { echo "Performance test step failed"; exit 1; }
-        EOF
+      run: python3 ./.github/actions/run-aperi-mech-performance-tests/run_aperi-mech_performance_tests.py --ip ${{ inputs.VM_IP }} --username ${{ inputs.VM_USERNAME }} --gpu ${{ inputs.gpu }} --parallel ${{ inputs.parallel }}

.github/actions/run-aperi-mech-performance-tests/run_aperi-mech_performance_tests.py

@@ -0,0 +1,78 @@
+import argparse
+
+import paramiko
+
+
+def str_to_bool(value):
+    return value.lower() in ("true", "1", "t", "y", "yes")
+
+
+def run_aperi_mech_performance_tests(vm_ip, vm_username, gpu, parallel):
+    ssh = paramiko.SSHClient()
+    ssh.set_missing_host_key_policy(paramiko.AutoAddPolicy())
+    ssh.connect(vm_ip, username=vm_username)
+
+    # Convert string inputs to boolean
+    gpu = str_to_bool(gpu)
+    parallel = str_to_bool(parallel)
+
+    # Construct the compose file and service name
+    compose_file = "docker-compose.yml"
+    service_name = "aperi-mech-development"
+    if gpu:
+        compose_file = "docker-compose_nvidia_t4_gpu.yml"
+        service_name = "aperi-mech-gpu-development"
+
+    # Construct the test flags
+    test_flags = "--release"
+    if gpu:
+        test_flags += " --gpu"
+    else:
+        test_flags += " --cpu"
+    if parallel:
+        test_flags += " --parallel"
+    else:
+        test_flags += " --serial"
+
+    # Construct the command to run on the VM
+    commands = f"""
+    set -e
+    cd ~/aperi-mech
+
+    docker-compose -f {compose_file} run --rm {service_name} /bin/bash -c '
+        echo "Setting up Spack environment..."
+        . ~/spack/share/spack/setup-env.sh
+        spack env activate aperi-mech
+
+        echo "Running aperi-mech performance tests..."
+        cd ~/aperi-mech/test/
+        ./run_regression_tests.py --directory ./performance_tests/aperi-mech --build-dir ~/aperi-mech/build/ {test_flags} --write-json --no-preclean
+        ./run_regression_tests.py --directory ./performance_tests/aperi-mech --build-dir ~/aperi-mech/build/ {test_flags} --clean-logs
+        ./run_regression_tests.py --directory ./performance_tests/aperi-mech --build-dir ~/aperi-mech/build/ {test_flags} --clean-results
+    ' || {{ echo "Performance test step failed"; exit 1; }}
+    """
+
+    print("Executing the following commands on the VM:")
+    print(commands)
+
+    stdin, stdout, stderr = ssh.exec_command(commands)
+    for line in stdout:
+        print(line.strip())
+    for line in stderr:
+        print(line.strip())
+
+    ssh.close()
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Run performance tests on remote VM")
+    parser.add_argument("--ip", default="127.0.0.1", help="IP address of the VM")
+    parser.add_argument("--username", default="azureuser", help="Username for the VM")
+    parser.add_argument("--gpu", default="false", help="GPU flag (true/false)")
+    parser.add_argument(
+        "--parallel", default="false", help="Parallel flag (true/false)"
+    )
+
+    args = parser.parse_args()
+
+    run_aperi_mech_performance_tests(args.ip, args.username, args.gpu, args.parallel)

.github/actions/run-gtest-performance-tests/action.yaml

@@ -24,38 +24,4 @@ runs:
 
     - name: Run gtest performance tests
       shell: bash
-      run: |
-        ssh -T -o ConnectTimeout=10 ${{ inputs.VM_USERNAME }}@${{ inputs.VM_IP }} << 'EOF'
-          set -e # Exit on error
-          cd ~/aperi-mech
-
-          compose_file=docker-compose.yml
-          service_name=aperi-mech-development
-          if [ ${{ inputs.gpu }} == 'true' ]; then
-            compose_file=docker-compose_nvidia_t4_gpu.yml
-            service_name=aperi-mech-gpu-development
-          fi
-          # Debugging
-          echo "On VM, Compose file: $compose_file"
-          echo "On VM, Service name: $service_name"
-
-          docker-compose -f $compose_file run --rm $service_name /bin/bash -c '
-            build_path=~/aperi-mech/build/${{ inputs.build-type }}
-            if [ ${{ inputs.gpu }} == "true" ]; then
-              build_path=~/aperi-mech/build/${{ inputs.build-type }}_gpu
-            fi
-            # Debugging
-            echo "In container, inputs.build-type: ${{ inputs.build-type }}"
-            echo "In container, inputs.gpu: ${{ inputs.gpu }}"
-            echo "In container, build_path: $build_path"
-
-            cd $build_path
-            echo "Setting up Spack environment..."
-            . ~/spack/share/spack/setup-env.sh
-            spack env activate aperi-mech
-
-            echo "Running gtest performance tests..."
-            ./performance_tests
-
-          '  || { echo "Gtest performance test step failed"; exit 1; }
-        EOF
+      run: python3 ./.github/actions/run-gtest-performance-tests/run_gtest_performance_tests.py --ip ${{ inputs.VM_IP }} --username ${{ inputs.VM_USERNAME }} --gpu ${{ inputs.gpu }} --build-type ${{ inputs.build-type }}