Virtual Machine Energy Estimation via Hardware Functional Unit Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating virtual machine energy consumption are inaccurate due to their reliance on coarse CPU-based metrics, failing to account for the complex energy dynamics within computer systems.
Innovation Solution
A method and apparatus that divide system hardware resources into functional units, collect event information, and map it to virtual machines using energy consumption coefficients to estimate energy consumption more accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If CPU runtime-based estimation method is used, then the estimation process is simple, but the measurement precision of virtual machine energy consumption is low
Solution Approach 1:
The patent segments the system hardware resources into multiple functional units (CPU, memory, storage, network, etc.) and divides energy consumption estimation into separate components. Each functional unit's energy consumption is estimated independently using specific event information and coefficients, then aggregated to obtain total virtual machine energy consumption. This segmentation approach maintains operational simplicity while significantly improving measurement precision by accounting for multiple energy-consuming components rather than relying solely on CPU runtime.
2Measurement precision
If multiple functional units and event information are considered, then the measurement precision of virtual machine energy consumption is improved, but the device complexity increases
Solution Approach 1:
The patent introduces an intermediary estimation model that acts as a mediator between hardware resources and virtual machines. The model includes functional unit event reading components that collect event information, energy consumption coefficient components that store pre-determined coefficients, and calculation components that aggregate results. This intermediary layer simplifies the complex relationship between multiple hardware components and virtual machines by providing a structured framework for data collection, processing, and aggregation, thereby improving measurement precision without proportionally increasing system complexity.
Solution Approach 2:
The patent changes the parameters used for energy consumption estimation from a single CPU runtime parameter to multiple parameters including event information from different functional units (CPU events, memory access events, storage I/O events, network transmission events) and corresponding energy consumption coefficients. By changing the parameter set and using pre-determined coefficients for each parameter, the system achieves higher measurement precision while managing complexity through systematic parameter organization and calculation aggregation.
Data Source
AI summary
A method and apparatus for estimating virtual machine energy consumption, and in particular, a method and apparatus for estimating virtual machine energy consumption in a computer system. The method includes: obtaining system energy consumption of the system hardware resources; obtaining event information of a plurality of functional units into which the system hardware resources are divided, and mapping the event information to the respective virtual machines; and calculating energy consumption of the virtual machines according to a plurality of energy consumption coefficients corresponding to the plurality of functional units and according to the event information mapped to the functional units of the respective virtual machines.