Network Impairment Replication for Application State Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network administrators face challenges in identifying the cause of poor application quality of experience (QoE) due to limited visibility and control over end-to-end traffic paths in distributed and dynamic network environments, where network conditions intersect with application and endpoint-level issues, making it difficult to distinguish between network path degradations and application or endpoint problems.
Innovation Solution
A device receives application performance metrics and replicates network impairments to associate specific network conditions with application states, using a combination of predictive networking and trial-and-error testing to identify the cause of QoE issues, employing machine learning models and simulation environments to mimic network phenomena and gather application telemetry for analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If network administrators monitor connectivity between diverse end user locations and dynamic applications distributed to multiple cloud regions, then visibility and control over network paths improve, but system complexity and resource requirements increase significantly
Solution Approach 1:
The patent introduces an intermediary monitoring system that acts as a mediator between network administrators and the complex distributed network infrastructure. This intermediary automatically collects, aggregates, and analyzes network performance data from multiple sources, providing simplified visibility into end-to-end connectivity without requiring administrators to directly manage the complexity of monitoring diverse cloud regions and endpoints.
Solution Approach 2:
The patent employs synthetic transactions and digital twins to create virtual copies of user interactions with applications. These synthetic transactions replicate real user behavior patterns to test and monitor application performance across distributed cloud regions, enabling comprehensive visibility without the need to instrument every physical endpoint and application instance.
2Measurement precision
If network administrators implement comprehensive monitoring of end-to-end traffic paths, then ability to identify QoE causes improves, but cost and resource consumption increase
Solution Approach 1:
The patent implements selective monitoring that focuses resources on critical performance indicators and problematic network paths rather than attempting to monitor every possible parameter and endpoint. The system identifies and prioritizes monitoring of key performance metrics that have the highest impact on user experience, applying excessive detail only where problems are detected rather than uniformly across the entire network.
Solution Approach 2:
The patent performs preliminary analysis and filtering of network data to identify potential QoE issues before they impact users. By proactively detecting degradation patterns and predicting potential failures, the system can focus monitoring resources on preemptively addressing problems rather than exhaustively monitoring all parameters continuously, thereby reducing overall resource consumption while maintaining high detection accuracy.
3Reliability
If network administrators rely on SaaS vendors to monitor application infrastructure, then vendor responsibility for application performance increases, but network administrator visibility into application-level issues decreases
Solution Approach 1:
The patent implements a multi-layer feedback mechanism where synthetic transactions and monitoring data flow from the application level through the network to administrators. The system collects application performance metrics from SaaS vendors, correlates them with network conditions, and provides feedback loops that enable administrators to understand both application-level and network-level contributions to QoE issues, maintaining visibility while leveraging vendor monitoring capabilities.
Data Source
AI summary
In one embodiment, a device receives application performance metrics generated by an online application accessible via a network and indicative of a particular state of the online application. The device performs testing of the online application by replicating a plurality of network impairments. The device associates, based on results of the testing, a particular network impairment from plurality of network impairments with the particular state of the online application. The device provides an indication of the particular network impairment as a cause of the particular state of the online application.