-
Notifications
You must be signed in to change notification settings - Fork 2
Home
mirazola-ik edited this page Dec 9, 2025
·
6 revisions
The Serverless Runtime is a Python-based service designed to execute offloaded functions within the COGNIT ecosystem. When a device uses the device-runtime to offload a task, the Serverless Runtime is the component responsible for receiving, executing, and returning the result of that task.
It exposes a REST API to handle function execution requests, allowing developers to offload computation from resource-constrained devices.
- Function Execution: Receives serialized functions via a REST API endpoint, executes them in an isolated environment, and returns the result.
- Service Integration: Provides pre-configured clients for interacting with essential services, such as MinIO for object storage, directly from within the offloaded function.
- Stateless and Scalable: Designed to be stateless, making it easy to deploy and scale across multiple nodes in an edge-cloud continuum.
- Centralized Logging: Integrates with a standardized logger, allowing logs generated by the offloaded function to be captured and viewed alongside the runtime's own logs.
- Prerequisites
- Environment Setup
- Running the Serverless Runtime
- Main Application (
app.py) - Core Modules
- Data Models
- High-Level Design
- Request Flow
- Component Interaction
- How to check function execution metrics
- How to update a SR template image