Virtual Machine Resource Consumption Metric Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cloud computing environments, accurately estimating computing resource consumption is challenging, especially when virtual machines are first deployed, leading to potential over- or under-billing due to lack of representative data, requiring coordination among multiple teams and resulting in inefficiencies and increased work hours.
Innovation Solution
A method that collects data from hypervisors and virtual machines, incorporating a stability factor to generate a custom consumption metric, which corrects default consumption values and provides a more accurate reflection of resource usage, thereby avoiding unexpected peaks and valleys in licensing fees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If default consumption metrics are used for billing, then billing can be processed quickly without extensive data collection, but the billing accuracy deteriorates leading to over- or under-billing
Solution Approach 1:
The system performs local scans at each virtual machine to collect computing resource consumption data in advance before billing is needed. This preliminary data collection enables accurate custom consumption metrics to be generated when hypervisor data becomes available, rather than relying on default metrics that may be inaccurate.
Solution Approach 2:
The system generates a stability factor based on hypervisor resource consumption data and compares it with default consumption metrics. When the default metric exceeds a threshold, the system provides feedback to generate a custom consumption metric instead, creating a closed-loop system that adapts to actual resource usage patterns.
2Measurement precision
If local scans are performed at each virtual machine to collect consumption data, then measurement precision improves, but device complexity and coordination requirements increase
Solution Approach 1:
The system divides the data collection process into segments: local scans are performed independently at each virtual machine to collect computing resource consumption data, while the hypervisor separately provides resource consumption data. This segmentation allows each component to operate independently, reducing coordination complexity while maintaining measurement precision.
Solution Approach 2:
The system introduces an intermediary mechanism that receives data from both local virtual machine scans and hypervisor monitoring, generates a stability factor, and determines whether to use default or custom consumption metrics. This intermediary layer simplifies the overall coordination by centralizing the decision-making process.
3Measurement precision
If custom consumption metrics are generated based on collected data, then billing accuracy improves, but processing time and computational resources increase
Solution Approach 1:
Local scans are performed in advance to collect computing resource consumption data, and this data is stored for later use. When hypervisor data becomes available, the custom consumption metric can be generated quickly by combining pre-collected local scan data with hypervisor data, rather than collecting all data at the last moment.
Solution Approach 2:
The system performs local scans at each virtual machine that collect more data than immediately needed, storing this data for future metric generation. This partial action of collecting data in advance reduces the processing time required when actual billing metrics need to be generated.
Data Source
AI summary
A set of virtual machines is deployed on a hypervisor. At each virtual machine, one or more local scans is performed to generate a set of computing resource consumption data. In response to receiving a set of hypervisor resource consumption data, a stability factor is generated. Based on the set of resource consumption data, the set of hypervisor resource consumption data, and the stability factor, a determination is made that a default consumption metric exceeds a threshold. In response to the determination, a custom consumption metric is generated, based on at least the set of computing resource consumption data. A user is notified of the custom consumption metric.