Predictive SD-WAN Traffic Rerouting for SLA Compliance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional reactive failure detection in SD-WANs leads to traffic disruption until a backup tunnel is activated, and predictive failure detection using machine learning is hindered by imbalanced datasets causing false positives, which can result in rerouting traffic to tunnels with inferior performance.
Innovation Solution
A device predicts tunnel failures using machine learning and proactively reroutes a subset of traffic to a backup tunnel that can satisfy service level agreements (SLAs), even if no backup tunnel fully meets the SLAs, by employing predictive routing and what-if evaluation models to identify suitable backup paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reactive failure detection using keep-alive mechanisms is used, then tunnel failures are detected and traffic is rerouted, but traffic is affected by the failure until the traffic is moved to another tunnel
Solution Approach 1:
The patent performs preliminary actions by proactively rerouting traffic to backup tunnels before the primary tunnel actually fails. The system evaluates backup tunnel performance and pre-establishes rerouting paths, so that when a failure occurs, traffic can be immediately redirected without disruption. This transforms the reactive keep-alive approach into a proactive system that eliminates traffic disruption time.
2Reliability
If predictive failure detection using machine learning is used, then proactive routing is enabled, but false positives occur due to imbalanced training datasets
Solution Approach 1:
The patent applies partial action by evaluating multiple backup tunnels and selecting only those that meet performance thresholds. Instead of making a binary prediction decision, the system partially activates rerouting only when backup tunnels are available and suitable, reducing false positives while maintaining proactive routing benefits for genuine failures.
Solution Approach 2:
The system implements feedback mechanisms where prediction outcomes and actual tunnel performance are continuously monitored. This feedback loop allows the machine learning model to learn from both true positives and false positives, improving prediction accuracy over time by adjusting to the imbalanced dataset characteristics.
3Reliability
If traffic is rerouted onto a backup tunnel due to false positive predictions, then traffic is protected from actual failures, but traffic is rerouted to tunnels with inferior performance
Solution Approach 1:
The patent applies local quality by evaluating and selecting specific backup tunnels based on their individual performance characteristics. Rather than treating all backup tunnels uniformly, the system identifies which backup tunnels can satisfy SLAs for specific traffic types, enabling precise rerouting decisions that maintain service quality while providing protection.
Solution Approach 2:
The system changes parameters by dynamically adjusting rerouting decisions based on real-time tunnel performance metrics and SLA requirements. When a false positive occurs, the system can change the routing state back to the primary tunnel if performance degradation is detected, thereby maintaining SLA satisfaction while still providing protection against actual failures.
Data Source
AI summary
In one embodiment, a device predicts a failure of a first tunnel in a software-defined wide area network (SD-WAN). The device determines that no backup tunnel for the first tunnel exists in the SD-WAN that can satisfy one or more service level agreements (SLAs) of traffic on the first tunnel, were the traffic rerouted from the first tunnel onto that tunnel. The device predicts, using a machine learning model, that a backup tunnel for the first tunnel exists in the SD-WAN that can satisfy an SLA of a subset of the traffic on the first tunnel, in response to determining that no backup tunnel exists in the SD-WAN that can satisfy the one or more SLAs of the traffic on the first tunnel. The device proactively reroutes the subset of the traffic on the first tunnel onto the backup tunnel, in advance of the predicted failure of the first tunnel.