rafay-workflow-sdk 0.0.21

Rafay workflow handler Python function SDK

Workflow Handler Python Function SDK

This SDK provides a framework for building Python functions that can be deployed as part of Rafay workflows.

Features

• FastAPI-based web framework for high performance
• Built-in logging and activity tracking
• Error handling with retry mechanisms
• Async/await support with thread pool execution for synchronous handlers
• Configurable concurrency for handling multiple requests

Concurrency Model

The SDK uses FastAPI with async request handling. Since user handler functions are typically synchronous, the SDK runs them in a ThreadPoolExecutor to avoid blocking the event loop.

Environment Variables

• FUNCTION_NAME: Name of the function (default: "default-function-name")
• LOG_LEVEL: Logging level (default: "INFO")
• LOG_BUFFER_CAPACITY: Number of log messages to buffer (default: 10)
• MAX_WORKERS: Thread pool size for handling concurrent requests (default: 40)
• LOG_FLUSH_TIMEOUT: Timeout for flushing logs (default: 10)
• skip_tls_verify: Skip TLS verification (default: "false")

Concurrency Explanation

• FastAPI Event Loop: Handles incoming HTTP requests asynchronously
• ThreadPoolExecutor: Executes synchronous user handler functions in separate threads
• MAX_WORKERS=40: Up to 40 concurrent handler executions (similar to Flask's WSGI_THREADS=40)

This design provides excellent concurrency for I/O-bound operations while maintaining compatibility with synchronous handler functions.