Disaster Recovery Metadata Analysis for Failure Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disaster recovery management systems (DRMs) face challenges in efficiently managing failures within data centers and DRMs, leading to potential catastrophic issues due to prolonged inactivity, resulting in a lack of effective orchestration of failure flows.
Innovation Solution
The proposed solution involves a method and system that extract insights from metadata of data centers and DRMs, identify device states, and take proactive actions to manage health and resilience, including generating virtual DRMs for continuous operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional disaster recovery management systems are used, then basic failure management is provided, but resource-intensive efforts are required for metadata refinement and insight extraction
Solution Approach 1:
The system automatically extracts insights from metadata and generates virtual DRMs without requiring manual intervention or resource-intensive refinement processes. The anomaly detection system autonomously monitors device states and triggers appropriate responses, enabling the system to serve itself rather than requiring continuous human oversight for metadata management.
Solution Approach 2:
The patent replaces manual metadata refinement processes with automated machine learning-based anomaly detection systems. Instead of requiring human resources to refine metadata and extract insights manually, the system uses AI/ML algorithms to automatically process metadata, identify anomalies, and generate virtual DRMs, significantly reducing resource consumption while maintaining or improving management effectiveness.
2Reliability
If proactive actions are taken to manage device states, then resilience is improved, but system complexity increases
Solution Approach 1:
The system segments the disaster recovery management function by creating separate virtual DRMs for different device states or failure modes. Each virtual DRM can be independently configured and activated based on detected anomalies, allowing the system to manage complexity through modular virtual instances rather than a monolithic complex system.
Solution Approach 2:
The patent creates virtual copies of disaster recovery management functions that can be instantiated based on detected device states. Instead of maintaining a single complex active DRM system, the system creates virtual replicas that are activated only when needed, reducing overall system complexity while maintaining resilience capabilities.
3Ease of operation
If continuous monitoring and management is implemented, then user experience is improved, but time consumption increases
Solution Approach 1:
The system implements periodic anomaly detection cycles rather than continuous intensive monitoring. The anomaly detection system operates at defined intervals or triggered by specific events, maintaining user experience through regular checks while reducing time consumption compared to continuous real-time analysis of all device states.
Solution Approach 2:
The system uses feedback mechanisms where anomaly detection results automatically trigger appropriate responses without requiring continuous user intervention. The system monitors device states, detects anomalies, and activates virtual DRMs based on feedback from the monitoring process, improving user experience through automated responses while minimizing time consumption by eliminating manual intervention loops.
Data Source
AI summary
A method for managing a disaster recovery module (DRM) includes: obtaining data center (DC) metadata from a DC; obtaining DRM metadata from the DRM; analyzing the DC metadata and the DRM metadata to extract relevant data; making, based on the relevant data and a first device state chain, a first determination that a failure score of the DRM is greater than a first predetermined failure score, in which the first determination indicates that the DRM is unhealthy; making, based on the first determination, the relevant data, and a second device state chain, a second determination that a failure score of the DC is less than a second predetermined failure score, in which the second determination indicates that the DC is healthy; and sending, based on the second determination, a recommendation to an administrator of the DRM to manage the health of the DRM.