Share:
The cloud workload refers to charge made by a huge diversity of independent services and applications located on cloud infrastructures. Workload characteristics can be defined by tasks or Virtual Machines (VMs). Then again, tasks/VMs scheduling, allocation, workload predictions are important topics, which are gaining steadily increasing attention particularly in the past few years. In all these fields, clustering methods are often used to recognize groups of workload components characterized by comparable behaviors. In order to attain an effective clustering, the appropriate clustering technique needs to be selected. This choice is vital especially when there are vast choices that provide different results. To address such issue, this paper presents a comprehensive review of clustering categories application in cloud workload clustering. This paper also proposes a novel systematic framework to select the suitable tasks/VMs clustering method in large-scale data centers based on clustering purpose, validation indices and comparison of results. © 2019 IEEE.
