Predictive DNS Resolution for Cloud Application Experience
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern software as a service (SaaS) applications delivered via public cloud infrastructure experience varying user experiences due to region-based service delivery, leading to inconsistent application performance across different geographic locations, despite having multiple Points of Presence (POP) globally.
Innovation Solution
A predictive application aware routing engine is deployed to collect application experience metrics, predict future performance for each provider endpoint, and adjust Domain Name System (DNS) resolution to route traffic through the endpoint offering the best predicted experience, using machine learning models and DNS dynamic reconfiguration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If cloud providers use region-based service delivery with multiple Points of Presence globally, then service coverage is improved, but application experience consistency deteriorates
Solution Approach 1:
The system performs preliminary actions by collecting application experience metrics from multiple provider endpoints before a user actually accesses the application. Machine learning models predict future performance metrics in advance, allowing the DNS resolution system to pre-determine the optimal endpoint before the user's request arrives, thus ensuring consistent application experience across different geographic locations.
Solution Approach 2:
The patent introduces an intermediary DNS resolution system that sits between the user and the cloud provider endpoints. This intermediary collects metrics, predicts performance, and makes intelligent routing decisions by adjusting DNS resolution, thereby mediating between the user's request and the multiple provider endpoints to ensure consistent application experience regardless of user location.
2Productivity
If DNS resolution is adjusted dynamically based on predicted metrics, then application performance is improved, but system complexity increases
Solution Approach 1:
The system implements self-service by automatically collecting application experience metrics from multiple provider endpoints, using machine learning models to predict future performance, and dynamically adjusting DNS resolution without human intervention. The entire process is automated, with the system serving itself by making intelligent routing decisions based on predicted metrics, thereby improving application performance while managing complexity through automation.
Solution Approach 2:
The patent incorporates feedback mechanisms by continuously collecting application experience metrics from provider endpoints and using this feedback to train machine learning models. These models then predict future performance metrics, which feed back into the DNS resolution system to dynamically adjust routing decisions, creating a closed-loop system that continuously optimizes application performance.
3Measurement precision
If machine learning models predict future performance metrics, then endpoint selection accuracy is improved, but computational requirements increase
Solution Approach 1:
The system applies partial action by using machine learning models to predict only the specific future performance metrics that are most relevant for endpoint selection, rather than analyzing all possible parameters. This selective prediction approach maintains endpoint selection accuracy while reducing computational requirements by focusing only on the most critical performance indicators.
Solution Approach 2:
The patent employs parameter changes by using machine learning models to transform historical application experience metrics into predicted future performance metrics. The models change the temporal parameter of the metrics (from past to future predictions) and select only the most relevant parameters for endpoint selection, thereby improving accuracy while managing computational complexity through intelligent parameter transformation.
Data Source
AI summary
In one embodiment, a device obtains application experience metrics for an online application. The device predicts, based on the application experience metrics, future application experience metrics for each of a set of provider endpoints for the online application. The device selects, based on the future application experience metrics, a particular provider endpoint from among the set of provider endpoints. The device provides, to a Domain Name System (DNS) resolver, resolution information for one or more of the set of provider endpoints that causes a query for one of those provider endpoints to resolve to an address of the particular provider endpoint.