Cloud Data Centre Optimisation for Various Client Classes

Cloud computing is the pay-per-use, dynamic, virtual provisioning of computer resources by service providers via the internet, including processing power, storage, networking, and IT infrastructure. The effective administration and processing of various client requests in a cloud data centre presents a number of difficulties in the modern cloud computing environment. Optimising resource allocation, improving system performance, and guaranteeing a flexible and responsive cloud infrastructure are the goals. This article deals with multi-class client requests that have no defined priority. One or more VMs serve any class of incoming client requests, and the likelihood of service is the same for every customer class. Assuming exponential distributions for arrival and service times, analytical explicit solutions are derived. To find the ideal system capacity and virtual machine (VM) count, Atom search optimisation (ASO) is used to build and optimise a cost model. The ideal number of virtual machines (VMs), the optimal system capacity, the minimal estimated cost, and other system performance metrics are computed. A range of numerical outcomes are shown in order to validate the system.