Host Recommendation Service Using Historical Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for selecting physical hosts for virtual machines lack access to historical data, making it difficult to evaluate workload trends and predict future requirements, leading to suboptimal host selection.
Innovation Solution
A method and system that utilize an analysis engine to retrieve and analyze historical metrics, including current and past load levels of physical hosts and virtual machines, to assign scores and recommend optimal hosts for virtual machine execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If only current status metrics are used for host selection, then the selection process is simple and fast, but the ability to evaluate workload trends and predict future requirements is compromised
Solution Approach 1:
The system performs preliminary actions by collecting and storing historical workload metrics for both physical hosts and virtual machines before the host selection decision is needed. This historical data is prepared in advance and made available when selection algorithms need to evaluate potential hosts, allowing trend analysis without delaying the actual selection process.
Solution Approach 2:
The patent introduces an intermediary component (the host selection system with analysis capabilities) that bridges the gap between raw historical metrics and the final host selection decision. This intermediary processes historical data, identifies trends, and provides informed recommendations, thereby preserving information value while maintaining efficient decision-making.
2Reliability
If historical metrics are collected and analyzed, then workload trends can be evaluated and future requirements predicted, but the system complexity increases
Solution Approach 1:
The host selection system is designed with multi-functionality, serving both as a simple current-status evaluator and as a comprehensive historical trend analyzer. The same system infrastructure handles both real-time metrics and historical data processing, reducing the need for separate complex subsystems while improving selection reliability through enhanced analytical capabilities.
3Measurement precision
If comprehensive historical data is accessed, then better-informed host selection is achieved, but the data retrieval and processing time increases
Solution Approach 1:
The system extracts only the relevant historical metrics needed for accurate host selection rather than processing all available historical data. By selectively extracting pertinent information about workload patterns, resource utilization trends, and host performance history, the system maintains high evaluation accuracy while minimizing unnecessary data processing time.
Data Source
AI summary
A host recommendation service identifies a virtual machine. An analysis engine retrieves a plurality of metrics for each of a plurality of physical hosts available for executing the virtual machine. The plurality of metrics includes a current level of load on each of the plurality of physical hosts and a level of load on each of the plurality of physical hosts during a prior time period. The analysis engine also identifies a level of load placed on a physical host by the virtual machine during a prior time period. The analysis engine determines a level of priority associated with the metrics and assigns a score to each of the plurality of physical hosts. The host recommendation service transmits an identification of one of the plurality of physical hosts on which to execute the virtual machine.