Selective Component Logging for Edge Device Issue Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing log collection in cloud-based information processing systems with virtual resources is challenging due to resource constraints and the need for efficient identification and resolution of issues on edge devices, which can impact performance and data integrity.
Innovation Solution
An apparatus and method for selecting target components on edge devices for log collection using similarity analysis and clustering algorithms to identify relevant logs for issue resolution, minimizing resource impact and improving diagnostic efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If logs are collected from all components of computing devices, then complete diagnostic information is obtained, but resource overhead and performance impact increase significantly
Solution Approach 1:
The patent segments the computing device into multiple components (e.g., virtualization layer, container layer, host operating system layer, hardware layer) and selectively collects logs only from relevant components based on the identified issue type, rather than collecting from all components uniformly
Solution Approach 2:
The patent applies local quality by tailoring the log collection scope to the specific diagnostic need - different issue types (virtualization issues, container issues, host OS issues) trigger collection from different component subsets, optimizing resource usage for each diagnostic scenario
2Measurement precision
If logs are collected from multiple components for comprehensive diagnosis, then root cause identification accuracy improves, but time required for log collection and processing increases
Solution Approach 1:
The system performs preliminary classification of the issue type before log collection, using this classification to pre-determine which components are relevant, thereby avoiding unnecessary log collection from unrelated components and reducing overall diagnostic time
Solution Approach 2:
The patent applies partial action by collecting logs only from the minimum necessary components required to diagnose the specific issue type, rather than performing excessive full-system log collection, thus achieving sufficient diagnostic accuracy with reduced time overhead
3Productivity
If selective log collection is performed based on issue type, then resource efficiency improves, but complexity of identifying relevant components increases
Solution Approach 1:
The system performs preliminary issue classification into predefined categories (virtualization, container, host OS, hardware) before log collection, which simplifies the subsequent component selection process by mapping issue types to known relevant components, reducing the complexity of real-time decision-making
Solution Approach 2:
The patent introduces an intermediary classification layer that bridges the issue detection and log collection stages, translating diverse issues into standardized categories that map to predetermined component sets, thereby managing complexity through abstraction
4Reliability
If comprehensive logs are collected for all computing devices in a cluster, then cluster-wide issue patterns are detected, but data transmission overhead and processing load increase
Solution Approach 1:
The patent segments the cluster into different computing devices and further segments log collection within each device based on relevance, transmitting only the necessary subset of logs from each device to the centralized system, reducing overall data transmission overhead while maintaining cluster-wide issue detection capability
Solution Approach 2:
The system performs partial log collection at each computing device based on the identified issue type and relevant components, rather than collecting and transmitting excessive comprehensive logs from all devices, achieving sufficient reliability for cluster-wide pattern detection with reduced transmission overhead
Data Source
AI summary
An apparatus comprises a processing device configured to detect a given issue encountered on a given computing device, to identify a given cluster of computing devices to which the given computing device belongs, and to determine a similarity between the given issue encountered on the given computing device and one or more historical issues encountered on one or more other computing devices belonging to the given cluster. The processing device is also configured to select, based at least in part on the determined similarity between the given issue and the one or more historical issues, a subset of a plurality of components of the given computing device as target components for log collection. The processing device is further configured to collect logs from the target components and to perform remedial actions determined utilizing the collected logs on the given computing device to resolve the given issue.