QOS-AWARE TASK SCHEDULING IN CLOUD-EDGE ENVIRONMENT

In Cloud frameworks, Virtual Machines (VMs) are planned to have as per their moment asset use (for example to has with most accessible RAM) disregarding their generally speaking and long haul use. Likewise, generally speaking, the planning and situation processes are computationally costly and influence the execution of conveyed VMs. In this work, a Cloud VM planning calculation that considers previously running VM asset use over the long run by dissecting past VM usage levels to plan VMs by advancing execution by utilizing knn and Naive Bayes characterization procedure. The Euclidean distance of knn is estimated and afterward, virtual machine is booked on the actual machine. The Cloud the board processes, as VM situation, influence previously sent frameworks so the point is to limit such execution corruption. In addition, over-burden VMs will quite often take assets from adjoining VMs, so the work boosts VMs genuine CPU usage. The outcomes show that our answer refines conventional Instant-based actual machine determination as it learns the framework conduct as well as it adjusts over the long run. The idea of VM booking as per asset observing information separated from past asset uses (counting PMs and VMs).The count of the actual machine gets diminished by four utilizing K-NN and NB classifier than Support Vector Machine (SVM) classifier. The errand performed by 28 actual machine while utilizing SVM is diminished by 24 actual machine by utilizing knn &nb classifier calculation additionally the mistake rates gets diminished by 0.025%.