Multi-Area SPB Network Segmentation for LSDB Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Shortest-Path Bridging (SPB) networks face limitations in scalability due to the growth of node numbers, leading to increased Link State Database (LSDB) sizes, which require more computing resources and result in unreliable convergence and network fluctuations, especially when changes occur in the network.
Innovation Solution
The implementation of a multi-area SPB architecture that uses virtual nodes and boundary nodes to manage and synchronize network information across multiple areas, allowing for a hierarchical or flexible topology without the need for a single flat SPB fabric, thereby reducing resource requirements and improving network resilience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If the number of nodes in the SPB network is increased, then the network size is expanded, but the Link State Database (LSDB) size increases requiring higher computing resources
Solution Approach 1:
The patent divides the SPB network into multiple areas using Area Boundary Nodes (ABNs). Each area maintains its own LSDB, so increasing network size by adding nodes to different areas does not proportionally increase the LSDB size at any single node. The ABN aggregates information from multiple areas, enabling network expansion without uniform LSDB growth across all nodes.
2Volume of moving object
If the LSDB size is increased to support more nodes, then network coverage is expanded, but computing resources such as disc space, memory, and processing power are consumed
Solution Approach 1:
By segmenting the network into areas with ABNs, each node only maintains an LSDB for its local area rather than the entire network. This reduces the computing resources required at each node while still providing comprehensive network coverage through the hierarchical area structure.
Solution Approach 2:
The Area Boundary Node acts as an intermediary between areas. It collects and aggregates routing information from multiple areas and distributes it to interior nodes, reducing the computing burden on individual nodes while maintaining comprehensive network awareness.
3Device complexity
If a flat network topology is used, then network simplicity is maintained, but network fluctuations occur when changes happen in any part of the network
Solution Approach 1:
The patent segments the flat network into hierarchical areas. When changes occur in one area, only that area and its boundary nodes need to reconverge, rather than the entire flat network. This isolation of convergence domains maintains network stability while preserving relative simplicity through structured organization.
4Loss of information
If all nodes maintain the complete link state database, then complete network awareness is achieved, but higher level of capability is required for all nodes
Solution Approach 1:
The patent segments network awareness into area-level knowledge. Interior nodes only need to understand the topology within their area, while ABNs maintain awareness of multiple areas. This hierarchical knowledge distribution reduces the capability level required at interior nodes while maintaining comprehensive network awareness at the boundary nodes.
Data Source
AI summary
Provided herein are systems and methods for providing a MAC-based redistribution policy between networks in a multi-area network. A network can have a boundary node that communicates to neighboring networks. Boundary nodes can receive policy updates that identify which services are redistributable across network boundaries. Boundary nodes can receive a packet for a service, translate the packet's encapsulation, and forward the packet across the boundary towards a destination node. Boundary nodes can forward the packet such that it originates in the second network from a virtual node.