Browser Incident Correlation via DOM Snapshot and Annotation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud-based applications face challenges in identifying and resolving user incidents due to complexity, diversity of user technical knowledge, and varying reporting capabilities, leading to difficulties in capturing necessary information and predicting the root cause of issues.
Innovation Solution
Implementing a system that collects incident data at the user's browser using client-side JavaScript APIs, including HTML DOM, device, and log data, with an annotation tool for user input, and a rendering module to analyze the data on the server side, enabling quick and complete incident resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If incident data is collected manually from users with diverse technical knowledge, then user involvement is required, but information completeness and accuracy deteriorate
Solution Approach 1:
The system automatically collects incident data including DOM snapshots, console logs, network requests, and device information without requiring user intervention. The JavaScript library injects into the browser and passively gathers all relevant incident information, eliminating the need for users to manually provide technical details while ensuring complete information capture
Solution Approach 2:
The patent replaces manual user reporting mechanisms with automated programmatic data collection. Instead of relying on users to describe incidents textually, the system uses JavaScript APIs to automatically capture browser state, logs, and technical metrics, substituting human effort with automated computational processes
2Measurement precision
If detailed incident data is collected from diverse user environments, then incident analysis accuracy improves, but system complexity increases
Solution Approach 1:
The patent implements a universal JavaScript library that functions across all modern browsers and devices. This single multi-functional solution collects DOM data, console logs, network requests, device information, and performance metrics uniformly, eliminating the need for separate collection mechanisms for different browser types or devices while maintaining comprehensive data capture
Solution Approach 2:
The system dynamically adjusts which data parameters are collected based on the specific incident context and user environment. The JavaScript library can enable or disable collection of specific data types (such as detailed console logs versus basic device information) to optimize between data completeness and resource consumption, adapting to varying system conditions
3Reliability
If comprehensive incident data is collected and analyzed, then root cause identification improves, but time to resolve incident increases
Solution Approach 1:
The system proactively collects and stores incident data in structured formats immediately when incidents occur, before analysis is needed. DOM snapshots, console logs, and device information are captured and preserved in their original state, creating a ready-to-analyze data repository that eliminates preparation time when incidents need investigation
Solution Approach 2:
The patent creates simplified copies and representations of complex incident data. The JavaScript library generates condensed incident reports that summarize key information from extensive data collections, and the server generates reconstructed incident scenarios that replicate user environments, allowing rapid analysis without processing entire raw data sets
Data Source
AI summary
An information collection module receives an incident report request from a user of a browser running a network application on a client system to thereby provide a page in the browser. The information collection module collects incident data including user interactions with the page, client system features and operations associated with executing the network application, and a document object model (DOM) of the page. An annotation tool of the network application receives annotations of the page from the user for inclusion in the incident data, and a server-side incident manager receives the incident data. The server-side incident manager includes a rendering engine that renders the document object model and the annotations, and an incident analyzer that displays the collected user interactions, and client system features and operations, and the rendered document object model and annotations.