In English

Optimera användningen av virtuella maskiner i Azure med maskininlärning

KEVIN HEDBERG GRIFFITH ; ERIK NGUYEN
Göteborg : Chalmers tekniska högskola, 2016. 39 s.
[Examensarbete på grundnivå]

This report describes a project that is about examining the possibilities to optimize the utilization of virtual machines in Microsoft Azure using machine learning. The thesis has been done at the company Atea Global Services (AGS). Virtual machines is a Azure service that AGS uses. However could these virtual machines be running without really being used. Services in Azure are not for free and when using a service like virtual machines companies are being charged be per minute. This means that AGS pays unnecessary expenses for the virtual machines that are running when they are not being used. Using an Azure service called Machine Learning Studio, a user pattern for when a virtual machine was being used was developed. An application has been developed that turns on or off a virtual machine based on user patterns from Machine Learning Studio. AGS can choose whether they want to continue working on the project or to take advantage of it right away to cut back on costs.



Publikationen registrerades 2017-10-11. Den ändrades senast 2017-10-11

CPL ID: 252442

Detta är en tjänst från Chalmers bibliotek