Code-Tracing Data Management via Property-Based Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing code-tracing data poses challenges due to the large volume of data generated, which consumes storage space and slows down software operations, making it difficult to identify relevant data for error analysis.
Innovation Solution
A method and system for managing code-tracing data involve receiving data from executing software components, adding it to temporary storage, detecting properties, copying to non-volatile storage when a defined trigger property is met, and deleting data based on a caching policy to maintain manageable storage sizes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If code-tracing data is stored in a log file for fault detection, then fault detection capability is improved, but storage space consumption increases and software performance deteriorates
Solution Approach 1:
The patent extracts only the essential information from code-tracing data by detecting specific properties (such as error categories, function entry/exit points, or custom metadata) and storing only those filtered records in the log file, rather than storing all generated trace data. This reduces storage consumption and improves write performance while maintaining fault detection capability for relevant events.
Solution Approach 2:
The patent applies preliminary filtering action by detecting properties of code-tracing data before storing it in the log file. The system evaluates each trace record against predefined criteria (such as category, source, or condition flags) and only persists records that match the criteria, preventing unnecessary data from consuming storage space and slowing down the logging operation.
2Loss of information
If all code-tracing data is stored for analysis, then comprehensive error analysis is improved, but difficulty in identifying relevant data increases
Solution Approach 1:
The patent extracts and stores only the most relevant code-tracing data by detecting specific properties such as error categories, warning levels, or custom metadata tags. This selective extraction ensures that the log file contains only information needed for fault detection, making relevant data easy to identify through property-based filtering while maintaining comprehensive analysis capability for the extracted subset.
Solution Approach 2:
The patent applies different storage qualities to different types of code-tracing data based on their importance. By detecting properties such as error severity, category, or source priority, the system stores high-priority data (errors, warnings) while filtering out low-priority data (informational messages), creating a log file with varying quality levels that optimizes both comprehensiveness and identifiability.
3Reliability
If code-tracing data is written to disk continuously, then real-time fault detection is improved, but storage space consumption increases
Solution Approach 1:
The patent extracts only essential information from continuous code-tracing data streams by detecting specific properties before disk writing. This property-based filtering ensures that only relevant trace records are persisted to disk, reducing storage space consumption while maintaining real-time fault detection capability through continuous property evaluation and selective persistence.
Data Source
AI summary
A method of managing code-tracing data is described. The method comprises the steps of: receiving code-tracing data from an executing software component; adding the received code-tracing data to a log stored in temporary storage; detecting a property of the received code-tracing data; copying the log to non-volatile storage when the property matches a defined trigger property; and deleting contents of the log according to a caching policy when the property does not match a defined trigger property.