Policy-Driven Traffic Steering for Cloud Application Transitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current approaches for traffic steering between blue and green environments in cloud-based applications are inadequate, as they struggle to control traffic percentage splits based on application-related conditions and provide effective feedback on performance and functionality.
Innovation Solution
A system comprising an analytics server and a steering server that generates and adjusts steering policies to guide traffic decisions between current and updated cloud-based code environments, using rules based on HTTP request expressions and smartflows to manage traffic splitting, mirroring, redirecting, and rate limiting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If DNS endpoints are manipulated to split traffic between blue and green environments, then traffic steering is achieved, but precise control of traffic percentage and conditional splitting based on application conditions is lost
Solution Approach 1:
The patent introduces a traffic management service as an intermediary component between the DNS and the application environments. This service receives traffic steering requests, applies policy-based rules for conditional splitting, and dynamically controls traffic percentage distribution to blue and green environments, thereby enabling precise and adaptable traffic management that DNS manipulation alone cannot provide
Solution Approach 2:
The system implements dynamic traffic steering capabilities where traffic splitting ratios can be adjusted in real-time based on application conditions, performance metrics, and policy rules. The traffic management service continuously monitors environment health and dynamically redistributes traffic percentages between blue and green environments, transforming static DNS-based routing into a dynamic, adaptive system
2Ease of manufacture
If manual configuration changes are made to split traffic between environments, then traffic steering is implemented, but verification of performance and functional correctness becomes difficult and time-consuming
Solution Approach 1:
The patent implements comprehensive feedback mechanisms where the traffic management service continuously collects performance metrics, functional correctness data, and health status from both blue and green environments. This feedback is automatically analyzed to verify the effectiveness of traffic splitting configurations, detect issues early, and trigger automatic adjustments or alerts, eliminating the need for manual verification of performance and functional correctness
Solution Approach 2:
The system enables self-verification through automated monitoring and validation capabilities. The traffic management service automatically detects configuration errors, validates traffic splitting effectiveness, and performs self-diagnosis of environment health, reducing the burden of manual verification and enabling rapid detection of performance and functional issues
3Reliability
If a small percentage of traffic is diverted to verify new code, then risk is reduced, but the ability to quickly verify performance and functionality across different traffic splits is limited
Solution Approach 1:
The patent implements progressive traffic verification strategies where traffic is initially diverted in small controlled percentages to verify new code safety, then automatically increased in staged increments as confidence builds. The system can simultaneously test multiple traffic split scenarios and automatically select the optimal verification approach, balancing risk reduction with accelerated verification throughput through parallel testing capabilities
Data Source
AI summary
Provided are methods and systems for transition between a current cloud-based code environment and an updated cloud-based code environment. A method for transition between a current cloud-based code environment and an updated cloud-based code may commence with generating a steering policy. The steering policy may include a set of rules to guide steering decisions between a current cloud-based code environment and an updated cloud-based code environment. The method may further include sending the steering policy to a steering server. The steering server may make steering decisions to steer, based on the steering policy, service requests between the current cloud-based code environment and the updated cloud-based code environment. The method may continue with receiving feedback concerning actual steering decisions made by the steering server. The method may further include automatically adjusting the steering policy in response to the feedback.