Edge Function Bursting via Service Registry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional edge computing approaches face challenges in resource utilization, real-time decision-making, and function deployment due to inadequate awareness of available resources, dynamic resource consumption, and the complexity of deploying functions across heterogeneous environments.
Innovation Solution
The implementation of a function bursting mechanism that utilizes a topology category and registration mechanism to dynamically determine the best placement for functions, allowing for automated deployment and orchestration across edge and cloud environments, thereby optimizing resource use and cost-performance balance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If each edge-station and user equipment operate as independent entities, then system independence and simplicity are maintained, but awareness of available resources across the network is lost
Solution Approach 1:
A service registry is introduced as an intermediary component that enables edge-stations and user equipment to discover and register available services and resources. The registry acts as a central information hub where services publish their availability and consumers can query for needed resources, solving the information loss problem while preserving operational independence.
Solution Approach 2:
Services perform preliminary registration with the service registry to advertise their availability before being needed. This advance publication of service information enables subsequent discovery and invocation without requiring continuous active communication between all entities, maintaining independence while ensuring resource awareness.
2Productivity
If dynamic real-time decision making is implemented for function placement, then resource optimization is improved, but decision-making speed becomes insufficient for sub-1 second function execution requirements
Solution Approach 1:
Edge-stations and user equipment maintain a pre-populated list of available services in the service registry during idle periods or lower-priority operations. When a function needs to be executed, the system can quickly query the pre-registered services and make placement decisions without performing complex real-time analysis, achieving both optimization and speed requirements.
Solution Approach 2:
The decision-making process is segmented into two phases: (1) Service registration and listing in the registry (performed in advance), and (2) Quick query and selection from pre-registered services (performed at execution time). This segmentation allows comprehensive resource awareness while enabling rapid decision-making at the moment of function invocation.
3Ease of manufacture
If traditional FaaS trigger-based execution models are used, then function execution is simplified, but deployment complexity increases due to the need for all software vendors to deploy functions to all edge-FaaS providers
Solution Approach 1:
The service registry acts as an intermediary that decouples function invocation from specific execution providers. Instead of requiring direct integration between all software vendors and edge-FaaS providers, the registry serves as a universal discovery and routing mechanism, significantly reducing deployment complexity while maintaining execution simplicity.
Solution Approach 2:
The service registry provides a universal interface that all software vendors and edge-FaaS providers can use without requiring vendor-specific integration. Services register their availability in the registry, and any consumer can query and invoke appropriate services through the same registry, creating a multi-functional, vendor-agnostic system that reduces deployment complexity.
Data Source
AI summary
One example method includes determining that local resources at an edge site are inadequate to support performance of a function needed by software running on the edge site, invoking a client agent, in response to invoking the client agent, receiving an execution manifest, determining, by the client agent, where to execute the function, wherein the determining comprises identifying a target execution environment for the function and the determining is based in part on information contained in the execution manifest, and transmitting, by the client agent, the execution manifest to a server agent of the target execution environment, and the execution manifest facilitates execution of the function in the target execution environment.