Router Linecard Routing Table Assignment Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for assigning routing tables to router linecards in VPNs are inefficient, leading to high memory usage and computational burdens due to the storage of unnecessary VRF tables on all linecards, even though not all VPNs are coupled to each linecard, and the handling of numerous multicast groups.
Innovation Solution
The method involves computing costs to determine the optimal assignment of VPNs to subsets of linecards using an uncapacitated facility location problem approach, where a schema identifier defines subsets of linecards, a cover analyzer calculates assignment costs, and an optimizer decides the assignment based on these costs to minimize memory usage and processor overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If routing information for all VPNs is assigned to all linecards, then any linecard can handle any VPN traffic, but memory consumption and processor utilization increase significantly
Solution Approach 1:
The patent segments the routing information distribution by dividing VPNs into different subsets and assigning each subset to specific linecards. Instead of distributing all VPN routing information to all linecards, the system creates multiple routing table subsets (e.g., VRF tables 1-N) and assigns them to different linecards based on their capabilities and the VPN traffic patterns. This segmentation reduces the memory burden on each individual linecard while maintaining overall system versatility.
Solution Approach 2:
The patent implements local quality by making each linecard have different routing information based on its specific role and the VPNs it needs to handle. Rather than uniform distribution of all routing tables to all linecards, the system tailors the routing information stored on each linecard to its local requirements. Linecards that handle specific VPNs only store routing information for those VPNs, optimizing memory usage while ensuring each linecard has the necessary information for its assigned traffic.
2Adaptability or versatility
If routing information for all VPNs is assigned to all linecards, then routing flexibility is improved, but computational burden increases
Solution Approach 1:
The patent segments the routing decision-making process by dividing VPN handling into different subsets assigned to different linecards. Each linecard processes only the VPN traffic assigned to it, rather than all VPN traffic. The router control logic segments the routing table into multiple subsets and distributes them accordingly, reducing the computational burden on each linecard's processor while maintaining overall routing flexibility through the distributed architecture.
Solution Approach 2:
The patent applies local quality by enabling each linecard to make routing decisions independently for its assigned VPN subset using locally stored routing information. This eliminates the need for centralized processing of all routing decisions, reducing overall processor utilization. Each linecard has the specific routing information it needs locally, allowing distributed and efficient routing decisions without requiring all linecards to process all VPN routing logic.
3Reliability
If VRF tables are stored on all linecards, then routing information availability is improved, but memory usage increases
Solution Approach 1:
The patent segments the VRF tables into multiple subsets (VRF tables 1-N) and distributes these subsets across different linecards rather than duplicating all VRF tables on all linecards. Each linecard stores only the subset of VRF tables relevant to the VPNs it handles. This segmentation maintains routing information availability for each specific VPN on the linecards that need it, while significantly reducing total memory usage across the router system.
Solution Approach 2:
The patent implements local quality by storing routing information locally on each linecard only for the VPNs that linecard is configured to handle. Instead of uniform distribution of all VRF tables to all linecards, the system tailors the local routing information storage to each linecard's specific VPN assignment. This ensures routing information is available where needed while avoiding redundant storage and reducing overall memory consumption.
Data Source
AI summary
An example involves determining a first cost associated with assigning routing information for a subset of virtual private networks (VPNs) to a first subset of linecards of a router by determining a cost as a sum of: a first product of (1) a first number of routes associated with a first VPN of the first subset of the VPNs and (2) a number of the linecards included in the first subset of the linecards; and a second product of (1) a second number of routes associated with a second VPN of the first subset of the VPNs and (2) the number of the linecards included in the first subset of the linecards. The first subset of the linecards is configured with routing information for the first subset of the VPNs when the first subset of the VPNs is assigned to the first subset of the linecards based on the cost.