Plugin Framework for Zero-Touch Management of Heterogeneous Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing modern, geographically distributed data centers with multi-vendor equipment is complex due to the lack of scalable, hierarchical infrastructure management solutions, particularly in handling event and telemetry data across diverse components and vendors.
Innovation Solution
A plugin framework that facilitates management of geographically distributed, multi-vendor infrastructure by aggregating resources through a Redfish-based API, providing a unified interface for fault and configuration management, and mediating communications between north-bound clients and heterogeneous infrastructure resources using a distributed plugin framework and Event Message Bus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multi-vendor equipment is deployed to reduce upfront costs and avoid vendor lock-in, then cost and flexibility are improved, but device complexity and management difficulty increase
Solution Approach 1:
The patent introduces a plugin framework as an intermediary layer between the resource aggregator and heterogeneous infrastructure resources. This framework includes vendor-specific plugins that translate between a unified management interface and diverse vendor protocols, enabling multi-vendor support while maintaining simplified centralized management. The plugin framework mediates communication between different systems, resolving the complexity of managing multiple vendor equipment types.
Solution Approach 2:
The management system is segmented into modular components: a core resource aggregator that handles unified management logic, and separate vendor-specific plugins that handle protocol translation. This segmentation allows the system to support multiple vendors without increasing core complexity, as each vendor's specificities are isolated in independent plugin modules that can be added or removed as needed.
2Reliability
If distributed data center architecture is implemented to improve reliability and service continuity, then system reliability is improved, but infrastructure complexity increases
Solution Approach 1:
The resource aggregator implements a universal management interface that can manage diverse infrastructure resources across multiple geographically distributed data centers through a single unified system. The Redfish-based API provides multi-functional access to different resource types and locations, simplifying the management of distributed infrastructure while maintaining the reliability benefits of geographic distribution.
Solution Approach 2:
The plugin framework serves as an intermediary that abstracts the complexity of distributed, heterogeneous infrastructure from the management system. It handles protocol translation and resource abstraction, allowing the resource aggregator to manage geographically distributed data centers through a unified interface without being burdened by the underlying infrastructure complexity.
3Ease of operation
If Zero Touch Management is implemented to reduce operational complexity, then ease of operation is improved, but system complexity increases due to need for sophisticated automation
Solution Approach 1:
The system implements Zero Touch Management capabilities where infrastructure resources can be automatically discovered, provisioned, and managed without manual intervention. The plugin framework enables automatic protocol negotiation and resource configuration, allowing the system to serve itself and reducing operational complexity while the automation framework handles the sophisticated coordination in the background.
Solution Approach 2:
The Event Message Bus implements feedback mechanisms that automatically monitor infrastructure resources and trigger appropriate management actions. When events are detected from managed resources, the system automatically processes and responds to these events through the plugin framework, enabling Zero Touch Management where operational decisions are made automatically based on real-time feedback from the infrastructure.
4Productivity
If centralized management of large-scale distributed resources is implemented to improve operational efficiency, then productivity is improved, but handling synchronous events from numerous resources becomes challenging
Solution Approach 1:
The Event Message Bus acts as an intermediary event handling system between the resource aggregator and numerous infrastructure resources. It provides centralized subscription and event routing capabilities that can efficiently handle synchronous events from tens to thousands of resources. The message bus abstracts the complexity of event subscription and routing, allowing centralized management while efficiently processing large volumes of events through its mediator architecture.
Data Source
AI summary
Example implementations relate to a plugin framework that facilitates management of heterogenous infrastructure resources residing in distributed data centers by multiple resource aggregators (RAs). According to an example, the plugin framework is logically interposed between the RAs and the heterogeneous infrastructure resources. The plugin framework includes multiple plugins each of which exposes a public API through which the RAs interact with the heterogeneous infrastructure resources. Each of the plugins corresponds to a particular type of resource of the heterogeneous infrastructure resources. Responsive to a first request issued to an RA by a north-bound client to a collection of the heterogeneous infrastructure resources, a plugin receives via its public API a second request including an RA-native message directed to a resource of the collection. The plugin directs the resource to take an action specified by the second request by translating the RA-native message to an infrastructure-native message understood by the resource.