Cloud cost optimization : Finding unused cloud resources using machine learning and heuristics

University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)

Author: Sonja Ericsson; [2020]

Keywords: ;

Abstract: Data shows that organizations’ IT infrastructure suffer from severe underutilization and large amounts of unused resources. This project investigates ways to detect unused and rarely used cloud resources automatically by using unsupervised learning and heuristics on monitored metrics and metadata. The results indicate the usefulness of using both of these methods to detect such resources. Limited research has been conducted in this area and not much support is provided natively through public providers despite data indicating significant waste. As the public cloud industry continues to grow, further research is needed to support efficient usage of cloud resources not only to avoid overspending in the cloud, but in particular due to the environmental footprint of the cloud computing industry. Conclusively, big cost savings can be made from monitoring for unused cloud resources. 

  AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)