Edge Bare Metal Interface System for Distributed Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monolithic models for managing edge bare metal (EBM) services are not suitable for significant scaling, as they require a single instance of metal as a service (MAAS) to handle inventory and operation of all EBM services.
Innovation Solution
An interface and abstraction system is implemented to provide a uniform interface between multiple EBM servers located at network edges and external systems, allowing for asynchronous interaction and distributed MAAS instances, each with its own MAAS instance for coupling with the interface system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single instance of MAAS is used to handle inventory and operation of all EBM services, then the system is simple and consistent for upstream interactions, but the system is not suitable for significant scaling
Solution Approach 1:
The patent divides the monolithic MAAS system into multiple distributed MAAS instances, each managing a subset of EBM servers. The interface system segments the unified inventory into multiple inventories accessible by different MAAS instances, enabling horizontal scaling while maintaining a consistent interface layer for upstream systems.
2Adaptability or versatility
If multiple distributed MAAS instances are deployed for scaling, then the system can handle significant scaling, but the complexity of managing multiple instances increases
Solution Approach 1:
The interface system acts as an intermediary layer between upstream systems and multiple distributed MAAS instances. It provides a unified interface that abstracts the complexity of multiple instances, allowing upstream systems to interact with a single consistent endpoint while the interface system handles the distribution and coordination of requests across multiple MAAS instances.
3Ease of operation
If a monolithic MAAS model is used, then upstream interactions are simple and consistent, but wait times increase and responsiveness decreases
Solution Approach 1:
The patent segments the single MAAS instance into multiple parallel MAAS instances, each capable of handling requests independently. This segmentation enables parallel processing of requests, reducing wait times and improving responsiveness while the interface system maintains a unified interface for upstream interactions.
Solution Approach 2:
The patent transitions from a single-dimension monolithic architecture to a multi-dimensional distributed architecture by adding the dimension of parallelism. Multiple MAAS instances operate in parallel, processing requests simultaneously, which reduces overall wait time while the interface system maintains a consistent single-point interface for upstream systems.
Data Source
AI summary
Novel tools and techniques are provided for implementing interface and abstraction system for edge bare metal (“EBM”) service. In various embodiments, an interface system receives first input requesting interaction with one or more EBM servers, each located at a network edge and each configured to provide bare metal server resources. The interface system provides a uniform interface between a plurality of EBM servers and external systems and establishes a connection with each EBM server. The interface system generates commands (e.g., instructions and/or queries, etc.) for each EBM server based on the first input and sends the commands to each EBM server over the corresponding established connection. The interface system accesses and abstracts data from the EBM servers via the established connection with each EBM server, generates data views based on the abstracted data, and presents the data views to a requesting system among the one or more external systems.