Distributed Router Architecture with Resource Broker
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current routers have a unitary architecture that limits geo-resiliency and resource sharing, leading to constraints in service provision when demand varies across locations, as resources are tightly coupled within a single chassis and not easily shared between routers.
Innovation Solution
A distributed router architecture is implemented, where elements are virtualized and geographically distributed across multiple locations, forming a resource grid that allows dynamic allocation of resources and capabilities based on demand, using a resource broker to manage and coordinate resource sharing between physically disparate resource planes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If resources are tightly coupled within a single router chassis, then routing function is reliable and simple, but geo-resiliency and resource sharing capability are limited
Solution Approach 1:
The router is divided into multiple independent resource planes (first resource plane, second resource plane, etc.) that can be geographically distributed. Each resource plane contains routing resources that can be independently managed and shared, enabling the system to achieve resource sharing capability while maintaining manageable complexity through modular segmentation.
Solution Approach 2:
A resource broker is introduced as an intermediary component that mediates between different resource planes and the routing function. The resource broker manages resource allocation, sharing, and coordination across geographically distributed planes, enabling complex resource sharing while abstracting the complexity from the routing decisions.
2Reliability
If routing resources are concentrated in a single location, then local routing performance is optimized, but service availability is constrained when demand increases
Solution Approach 1:
Multiple resource planes located at different geographies are merged into a unified resource pool through the resource broker. This merging enables the system to aggregate routing capacity from multiple locations, providing both high availability through distribution and high productivity through combined resource capacity.
Solution Approach 2:
The system dynamically allocates routing resources across different resource planes based on real-time demand patterns. When demand increases at a particular location, the resource broker can dynamically allocate additional resources from other planes, enabling the system to adapt to varying loads while maintaining service availability.
3Adaptability or versatility
If a unitary router architecture is used, then device simplicity is maintained, but geo-resiliency is not supported
Solution Approach 1:
The unitary router architecture is segmented into multiple resource planes that can be geographically distributed. Each plane operates semi-independently but coordinates through the resource broker, enabling geo-resiliency while maintaining operational simplicity through standardized interfaces and centralized coordination.
Data Source
AI summary
An apparatus for providing distributed router architecture includes a processing element and a resource availability element. The processing element may be configured to receive indications of receipt of data associated with a service for routing to a destination address. The resource availability element may be in communication with the processing element and may be configured to monitor resource usage over a plurality of distributed resource planes. The processing element may be further configured to allocate a resource associated with routing the data based on the monitored resource usage.