Stateful Function-as-a-Service for Edge Computing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing FaaS platforms are limited in their ability to define and deploy stateful functions at the edge, leading to latency issues and complexity in managing local device state, particularly in edge computing environments where functions need to operate on local data without centralized hosting.
Innovation Solution
Implementing a FaaS platform that allows for the definition and deployment of stateful functions directly on edge devices, using a fabric manager and API gateway to manage function deployment, local state awareness, and device addressing, enabling functions to operate on local data without relying on central hosts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If FaaS functions are hosted centrally in the cloud, then service management and scaling are simplified, but latency increases and local state awareness is lost
Solution Approach 1:
The patent segments the FaaS platform into distributed function runtime environments deployed on edge devices throughout the network, rather than consolidating all functions in a single centralized cloud location. This segmentation allows functions to execute locally near their data sources, reducing latency while maintaining centralized management capabilities through the fabric manager that orchestrates the distributed segments.
Solution Approach 2:
The patent introduces a new dimension of spatial distribution by deploying function runtimes across multiple geographic locations at the network edge, moving from a single-dimensional centralized model to a multi-dimensional distributed architecture. This enables functions to be executed in the dimension closest to the data and users, reducing latency without sacrificing management simplicity.
2Loss of time
If FaaS functions are deployed at the edge on local devices, then latency is reduced and local state awareness is improved, but device complexity and deployment management become more challenging
Solution Approach 1:
The fabric manager serves as an intermediary between the centralized service provider and distributed edge devices. It automatically handles the complexity of deploying, configuring, and managing function runtimes across multiple edge locations, shielding service providers from deployment management challenges while enabling low-latency edge execution.
Solution Approach 2:
The function runtime environments on edge devices are designed to be self-configuring and self-managing to some extent, automatically registering themselves with the fabric manager and managing their own execution contexts. This reduces the operational complexity for service providers while maintaining the benefits of distributed edge deployment.
3Productivity
If stateful functions are implemented on edge devices, then local data processing capability is improved, but reliability and state persistence become more difficult to ensure
Solution Approach 1:
The patent merges the function execution environment with the edge device's local data storage and processing capabilities, creating an integrated stateful function runtime. This combination allows functions to directly access and process local data with persistent state, improving productivity while the fabric manager ensures reliability through coordinated state management across the distributed system.
Data Source
AI summary
One method includes receiving a user-defined function, generating a binary based on the user-defined function, copying the binary to a target host where the user-defined function will run, in response to a lookup request, received from an API gateway based on a function invocation by a user, looking up the target host and a route, and returning, to the API gateway, a target host ID and the route, so as to enable the API gateway to request the target host to respond to the function invocation.