Multi-Tiered Nexthop Failover for Network Packet Forwarding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network devices experience undesirable packet forwarding behavior due to stale programming when network topology changes, leading to packet loss and inefficient routing, as they continue to use inoperable links until their packet processors are updated.
Innovation Solution
A multi-tiered approach is implemented to manage packet processors, involving pre-programmed updates, reprogramming with pre-computed next hops, and standards-based reprogramming to quickly adapt forwarding behavior to changes in network topology, ensuring optimal packet routing and minimizing packet loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network devices continue to use existing programming after link failure, then packet forwarding stability is maintained, but packet loss increases due to stale routing information
Solution Approach 1:
The system pre-computes backup next hops and stores them in the forwarding information base before link failures occur. When a failure is detected, these pre-computed backup paths are immediately activated without requiring complex real-time computation, thus preventing packet loss while maintaining forwarding stability.
Solution Approach 2:
The invention prepares compensatory measures in advance by maintaining multiple next hop candidates in the forwarding information base. When a primary next hop fails, the system has already-positioned backup next hops ready to take over, cushioning against the harmful effect of link failures and preventing packet loss.
2Productivity
If packet processors are updated frequently to reflect network topology changes, then routing efficiency improves, but system complexity increases
Solution Approach 1:
The system segments the next hop information into primary next hops and backup next hops, stored separately in the forwarding information base. This segmentation allows the control plane to manage complex routing logic while the data plane uses simple lookup tables, reducing the complexity burden on packet processors while maintaining routing efficiency.
Solution Approach 2:
The invention introduces a forwarding information base as an intermediary layer between the control plane and data plane. The FIB stores pre-computed primary and backup next hops, allowing the control plane to handle complex routing computations offline while the data plane performs simple table lookups, thus improving routing efficiency without increasing packet processor complexity.
3Loss of time
If backup next hops are pre-computed and stored, then failover time is reduced, but memory usage in forwarding hardware increases
Solution Approach 1:
The system applies local quality by storing backup next hop information selectively in the forwarding information base based on the criticality and failure probability of specific links. Not all next hops require backup entries, so the system optimizes memory usage by maintaining backup information only where most needed, reducing overall memory consumption while still achieving fast failover for critical paths.
Data Source
AI summary
In general, in one aspect, embodiments relate to a network device for forwarding packets as part of a network comprising an adjacent device connected to the network device via a link. The network device includes a switching system for directing the packets between ports of the network device, wherein a port of the ports is operably connected to the adjacent device via the link, and a switching system manager programmed to: identify a failure of the link, in response to identifying the failure of the link, perform a multi-tiered next hop failover of the switching system based on the failure of the link to obtain an updated switching system that does not forward the packets using the failed link, and forward a portion of the packets using the updated switching system.