Coordinated Tunnel Rerouting in Distributed Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed path computation in computer networks is inefficient for optimizing tunnel placement and rerouting, leading to issues like burst signaling, preemption cascading, and fragmentation of bandwidth, which can result in unavailable paths for tunnels during link or node failures, making it difficult to coordinate timing, placement, and bandwidth reservation across competing tunnels.
Innovation Solution
Head-end nodes receive a list of tunnels to be rerouted from an intermediate node and determine if they can reroute all tunnels, informing other nodes if unable; they then execute a shared algorithm to compute paths for all tunnels, applying a bandwidth scaling factor if necessary, ensuring coordinated rerouting and fair bandwidth allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If distributed path computation is used for tunnel rerouting, then flexibility and scalability are improved, but coordination efficiency and bandwidth utilization deteriorate
Solution Approach 1:
The patent combines multiple independent distributed path computation processes into a coordinated group computation process. When link failures occur, affected head-end nodes exchange information about their tunnels and compute paths collectively rather than independently, merging their computation efforts to achieve better coordination and bandwidth utilization while maintaining the distributed architecture's flexibility and scalability.
2Device complexity
If distributed independent rerouting is performed, then system complexity is reduced, but bandwidth fragmentation and preemption cascading increase
Solution Approach 1:
The patent introduces feedback mechanisms where head-end nodes exchange information about their tunnel rerouting needs and bandwidth requirements. This feedback loop allows nodes to adjust their path computation decisions based on the state of other nodes, preventing bandwidth fragmentation and preemption cascading while keeping the system architecture relatively simple and avoiding the need for a complex centralized controller.
3Speed
If head-end nodes compute paths independently, then response time to topology changes is improved, but overall tunnel placement optimization deteriorates
Solution Approach 1:
The patent implements preliminary action by having head-end nodes proactively exchange information about their tunnels and bandwidth requirements before actual rerouting is needed. This preparation allows for optimized tunnel placement decisions to be made more efficiently when link failures occur, improving both the speed of response and the quality of tunnel placement optimization without requiring complex real-time coordination during failures.
Data Source
AI summary
In one embodiment, head-end nodes receive a list of tunnels to be rerouted from a particular link of an intermediate node. If a head-end node is unable to reroute a tunnel for which it is the head-end node using conventional distributed routing, each head-end node executes the same algorithm to compute paths for all tunnels in the list (e.g., potentially applying bandwidth scaling).