Disaggregated Network Fabric Management via Distributed Fabric Managers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Centralized network management systems face challenges such as single points of failure, lack of cost-effective scalability, and increased complexity as network size grows, and decentralized management introduces scaling issues due to coordination complexity.
Innovation Solution
A network fabric management approach where each network element can assume fabric management functions, allowing for distributed management, migration of responsibilities, and secure communication through management channels, enabling redundancy and scalability without a central authority.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a centralized management system is used, then management control is simplified, but the system becomes a single point of failure and lacks scalability
Solution Approach 1:
The patent divides the centralized management system into multiple distributed fabric managers, each responsible for specific management functions. This segmentation eliminates the single point of failure by distributing management responsibilities across multiple nodes, while maintaining operational simplicity through standardized management interfaces and protocols.
2Ease of operation
If a centralized management system is used, then management control is simplified, but the system lacks cost-effective scalability
Solution Approach 1:
The management system is segmented into multiple independent fabric managers that can be added or removed based on network size requirements. Each manager handles a portion of the fabric, allowing the system to scale incrementally without requiring complete system replacement, thus providing cost-effective scalability while maintaining simplified control through consistent management protocols.
3Adaptability or versatility
If decentralized management units are used, then scalability is improved, but coordination complexity increases
Solution Approach 1:
The patent introduces fabric manager intermediaries that coordinate between distributed management units and network elements. These intermediaries simplify coordination by providing standardized interfaces and protocols, reducing the complexity of inter-unit communication while enabling scalable deployment of multiple management units across the fabric.
4Adaptability or versatility
If management functions are distributed to network elements, then redundancy and scalability are improved, but the risk of losing all management functions simultaneously increases
Solution Approach 1:
The patent implements prior cushioning by distributing management functions across multiple fabric managers before any failure occurs. This distribution creates inherent redundancy where if one manager fails, others continue to provide management services. The system also includes failover mechanisms and backup protocols that activate beforehand to ensure continuous management functionality.
Data Source
AI summary
Systems and methods for disaggregated management of a network fabric are presented. Network elements composing the network fabric can operate as a fabric manager with respect to one or more management functions while also continuing to operate as a communication conduit among hosts using the fabric. The roles or responsibilities of a fabric manager can migrate from one network element to another to preserve management coherency as well as to secure management of the network. Additionally, fabric managers communicate with the network fabric through one or more management channels.