Detecting an error in a prediction of resource usage of an application in a virtual environment

Abstract

Described herein is a method for detecting an error in a prediction of resource usage of an application running in a virtual environment, comprising: providing a plurality of benchmark sets, executing the plurality of benchmark sets in a native hardware system in which the application natively resides, executing the plurality of benchmark sets in the virtual environment, collecting first traces of first resource utilization metrics in the native hardware system based on the execution of each of the plurality of benchmark sets in the native hardware system, collecting second traces of second resource utilization metrics in the virtual environment based on the execution of each of the plurality of benchmark sets in the virtual environment, generating at least one initial prediction model that maps the first traces of first resource utilization metrics to the second traces of second resource utilization metrics, computing a plurality of mean squared errors (MSE's) based on the at least one initial prediction model, each of the MSE's is further based on and associated with the collected first and second traces for a different one of the plurality of benchmark sets, and determining whether to use the initial prediction model to predict a resource usage of the application running in the virtual environment based on the plurality of MSEs.