Autoscaling Rules Evaluation with the Autoscaling Performance Measurement Tool
Year: 2018
Type: dissertation
Abstract: One of the keys to cloud computing has been the possibility to deliver consistent performance for reduced costs compared to traditional static implementations. With more and more autoscalers becoming available from different providers, it can be difficult to determine which of the implementations best suits one's needs. This raises the question on how to reliably compare autoscalers, since [...] it is hard to distinguish [an overall] winner [3, Chapter 6].
This thesis will tackle this problem by providing customers with an addition to the Auto-scaling Performance Measurement Tool (APMT) that includes the ability to measure various autoscalers using experimental metrics proposed by [3]. This offers customers a wide range of generalised metrics that they can use to determine which autoscaler best covers their Quality of Service (QoS) requirements.
Author: Walentin Lamonos
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2945762417