Workflow Executor

A Python microservice that implements a workflow execution engine with HTTP and gRPC interfaces.

Features

• HTTP server for health checks and metrics
• gRPC server implementing the ExecutorService defined in proto/executor.proto
• Prometheus metrics integration
• Concurrent workflow execution

Requirements

• Python 3.8+
• Make (for using the Makefile)

Setup

1. Create a virtual environment and install dependencies:

make setup

2. Install the package in development mode:

source venv/bin/activate
make install

Running the Service

Start the service with:

make run

This will start:

• HTTP server on port 8000 (configurable via HTTP_PORT env var)
• gRPC server on port 50051 (configurable via GRPC_PORT env var)
• Prometheus metrics server on port 9090 (configurable via PROMETHEUS_PORT env var)

API Endpoints

HTTP

• GET /health - Health check endpoint
• GET /metrics - Custom metrics endpoint

gRPC

The service implements the following gRPC methods:

• ExecuteWorkflow - Execute a workflow with streaming response
• GetWorkflowStatus - Get status of a workflow
• CancelWorkflow - Cancel a workflow

Development

Regenerating Proto Files

If you modify the proto file, regenerate the Python code with:

make proto

Running Tests

make test

Linting

make lint

Cleaning Up

make clean

Environment Variables

• GRPC_PORT - Port for the gRPC server (default: 50051)
• HTTP_PORT - Port for the HTTP server (default: 8000)
• PROMETHEUS_PORT - Port for the Prometheus metrics server (default: 9090)

License

Copyright (c) 2025 Windsurf