Application Abnormality Analysis via Scene Information Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to identify and repair abnormalities, such as Application Not Responding (ANR), in application programs due to the difficulty in pinpointing the cause of the issue within the application itself, as most stack information in trace files relates to system internals rather than specific user actions or service operations.
Innovation Solution
A method that records scene information during application program operation, including life cycle nodes, event information, and service interactions, allowing for the extraction and analysis of specific scene information when an abnormality occurs, enabling the identification of the affected functional modules and user interfaces, and facilitating abnormality analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional trace file analysis is used to diagnose application abnormalities, then system-level information is captured, but the specific cause related to user actions and service operations cannot be identified
Solution Approach 1:
The patent segments the application execution process into distinct scene information units, including life cycle nodes, event information, and service interaction data. This segmentation allows precise tracking of specific user actions and service operations that lead to abnormalities, rather than analyzing undifferentiated trace files. By dividing the execution flow into manageable scene segments, the system can pinpoint exact causes of abnormalities.
Solution Approach 2:
The patent implements preliminary recording of scene information during application execution, capturing life cycle nodes, events, and service interactions before abnormalities occur. This preliminary action ensures that when an abnormality happens, the relevant scene information is already available for immediate analysis, eliminating the need to rely on incomplete traditional trace files that are generated after the fact.
2Measurement precision
If detailed scene information is recorded during application operation, then abnormality analysis precision is improved, but system resource consumption increases
Solution Approach 1:
The patent applies local quality by selectively recording scene information based on its relevance to abnormality analysis. Rather than uniformly recording all possible data, the system focuses on capturing life cycle nodes, event information, and service interactions that are locally significant to understanding application behavior. This targeted approach maintains high analysis precision while minimizing unnecessary resource consumption.
Solution Approach 2:
The patent dynamically adjusts the recording parameters of scene information based on application state and abnormality risk. When the application is in critical sections or experiencing stress, the system increases recording detail; during normal operation, it reduces overhead. This parameter adjustment strategy balances analysis precision requirements with resource conservation.
3Difficulty of detecting and measuring
If comprehensive scene information including life cycle nodes and service interactions is recorded, then the ability to pinpoint abnormality causes is improved, but data processing complexity increases
Solution Approach 1:
The patent extracts and isolates the critical scene information elements needed for abnormality analysis, separating them from unrelated data. By taking out only the essential components (life cycle nodes, event information, service interactions) and organizing them into a structured format, the system reduces data processing complexity while maintaining comprehensive abnormality detection capability. The extracted scene information is presented in a way that facilitates direct correlation with abnormality events.
Data Source
AI summary
Embodiments of this application provide a method performed at a computing device for dealing with an abnormality of an application program. The method includes: recording scene information obtained during operation of an application program; receiving abnormality indication information indicating that an abnormality occurs to the application program; in response, extracting, from the recorded scene information, corresponding scene information, where the application program is divided into a plurality of functional modules, each of the functional modules has a plurality of user interfaces (UI), each functional module having a respective unique tag and each US having a respective unique identifier ID, and the recorded scene information including a life cycle node of an operation page that exists during the operation of the application program, and a TAG and an ID corresponding to the operation page; and performing abnormality analysis according to the extracted scene information.