Proxy Next Hop Entries for Rapid Network Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional router and switch devices face significant challenges in quickly reconfiguring hundreds of thousands of next hop entries when a network failure occurs, leading to prolonged downtime and traffic loss due to the time-consuming process of individually reprogramming each entry.
Innovation Solution
The implementation of proxy structures and state information collections within network devices allows for rapid reconvergence of next hop entries by associating incoming IP traffic with data structures that include forwarding information and state indicators, enabling efficient redirection of traffic to alternative paths without the need for individual reprogramming.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual reprogramming of each next hop entry is performed, then forwarding accuracy is maintained, but reconvergence time increases significantly
Solution Approach 1:
The patent segments the next hop entry reprogramming process by introducing proxy next hop entries that group multiple original next hop entries under a single proxy entry. When failover is needed, only the proxy entry needs to be reprogrammed rather than individually reprogramming each original next hop entry, thus maintaining forwarding accuracy while significantly reducing reconvergence time.
Solution Approach 2:
The patent introduces proxy next hop entries as intermediary structures between the control plane and the forwarding plane. These proxy entries act as mediators that can be quickly reprogrammed to redirect traffic, while the original next hop entries remain unchanged. This intermediary layer enables fast failover without compromising the precision of individual forwarding decisions.
2Reliability
If hundreds of thousands of next hop entries are reprogrammed individually, then complete failover coverage is achieved, but processing overhead increases
Solution Approach 1:
The patent merges multiple original next hop entries that share the same backup path under a single proxy next hop entry. This consolidation means that when a failure occurs, the control plane only needs to reprogram the proxy entry rather than individually updating hundreds of thousands of original entries, thus achieving complete failover coverage while dramatically reducing processing overhead.
Solution Approach 2:
The proxy next hop entry serves multiple functions: it acts as a forwarding entry for normal operation, serves as a grouping mechanism for failover, and provides a simplified interface for control plane management. This multi-functionality allows the system to achieve comprehensive failover coverage without the excessive processing overhead of individual entry reprogramming.
3Reliability
If traditional double barrel next hop entries are used, then backup path redundancy is provided, but reconvergence speed is insufficient for large-scale failures
Solution Approach 1:
The patent adds a new dimensional layer to the traditional double barrel next hop structure by introducing proxy next hop entries that sit above the original entries. This dimensional change allows the system to maintain the backup path redundancy of traditional approaches while enabling rapid reconvergence through proxy entry reprogramming, solving the speed limitation for large-scale failures.
Data Source
AI summary
Exemplary methods performed by a first network device (ND) include generating first and second prefix entries associating incoming Internet Protocol (IP) traffic to first and second data structures (DSs), respectively. Generating the first DS includes generating a first proxy including forwarding information causing incoming IP traffic to be forwarded to a second ND, and generating a second proxy referencing a third DS. Generating the second DS includes generating a first proxy including forwarding information causing incoming IP traffic to be forwarded to the second ND, and generating a second proxy referencing the third DS. The methods include generating the third DS including forwarding information causing the incoming IP traffic to be forwarded to a third ND, the third DS further including first state information indicating whether the forwarding information included in the first proxies of the first and second DSs should be used for forwarding the incoming IP traffic.