Fig. 1
From: Efficient resource provisioning for elastic Cloud services based on machine learning techniques
Elastic cloud service. This figure represents a typical elastic web server cluster, consisting on a service front-end, acting as load balancer, and a variable number of backend servers that process users requests. The cloud auto-scaling mechanisms allow to dynamically adapt the system to workload changes, by autonomously provisioning and deprovisioning resources (i.e., back-end servers), so that at each point in time, the available resources match the current service demand as closely as possible