Abstract
Autoscaling system can reconfigure cloud-based services and applications, through various configurations of cloud software and provisions of hardware resources, to adapt to the changing environment at runtime. Such a behavior offers the foundation for achieving elasticity in a modern cloud computing paradigm. Given the dynamic and uncertain nature of the shared cloud infrastructure, the cloud autoscaling system has been engineered as one of the most complex, sophisticated, and intelligent artifacts created by humans, aiming to achieve self-aware, self-adaptive, and dependable runtime scaling. Yet the existing Self-aware and Self-adaptive Cloud Autoscaling System (SSCAS) is not at a state where it can be reliably exploited in the cloud. In this article, we survey the state-of-the-art research studies on SSCAS and provide a comprehensive taxonomy for this field. We present detailed analysis of the results and provide insights on open challenges, as well as the promising directions that are worth investigated in the future work of this area of research. Our survey and taxonomy contribute to the fundamentals of engineering more intelligent autoscaling systems in the cloud.
Highlights
Modern IT companies, from small business to large enterprises, increasingly leverage cloud computing to improve their profits and reduce costs
Comparing with other cloud resources management sytems in general, the autoscaling system has been designed for (i) scaling cloud-based applications in response to dynamic changes in load, uncertainties in operations, handling multitenancy while ensuring Service Level Agreement compliance and so on; (ii) adapting both the software configurations and hardware resources that span over all SaaS, PaaS, and IaaS levels, whereas most of the other resource management considers hardware resources and IaaS only; and (iii) taking the Quality of Service (QoS) for cloud-based applications/services at the centre of the concerned objectives while the other resource management tasks often focus on resource utilization
4 TAXONOMY AND SURVEY RESULTS FOR Self-adaptive Cloud Autoscaling System (SSCAS) we present a taxonomy and survey results for the state-of-the-art SSCAS research obtained from our review process
Summary
Modern IT companies, from small business to large enterprises, increasingly leverage cloud computing to improve their profits and reduce costs. For many other cases, for example, unexpected workload changes, elasticity can be only enabled by runtime automatic scaling or, autoscaling: a dynamic process, often operating on a Physical Machine (PM), that adapts software configurations (e.g., threads, connections and cache, etc.) and hardware resources provisioning (e.g., CPU, memory, etc.) on demand, according to the time-varying environmental conditions. Comparing with other cloud resources management sytems in general, the autoscaling system has been designed for (i) scaling cloud-based applications in response to dynamic changes in load, uncertainties in operations, handling multitenancy while ensuring Service Level Agreement compliance and so on (in contrast, other resource management tasks, e.g., resource scheduling, often work on planned and deterministic sequence of resource demand); (ii) adapting both the software configurations and hardware resources (and their interplays) that span over all SaaS, PaaS, and IaaS levels, whereas most of the other resource management considers hardware resources and IaaS only; and (iii) taking the QoS for cloud-based applications/services at the centre of the concerned objectives (explicitly or implicitly) while the other resource management tasks often focus on resource utilization
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
