Priority-Aware Resource Allocation Algorithm for Cloud Computing
Kushagra Kinger, Ajeet Pratap Singh, Sanjaya Kumar Panda · Proceedings of the 2022 Fourteenth International Conference on Contemporary Computing · 2022
Cloud is an innovative model of the computing paradigm, where servers, networks, storage, development tools, and applications are enabled to the users through the Internet in the form of services. These services are provided by the cloud service providers (CSPs) by deploying virtual machines (VMs) as per the resource request received from the users. These VM requests are of various types and priorities, and therefore have varying impacts in terms of importance, criticality, accountability, and domains of applications. On the other hand, the resources/hosts are of varying sizes and assigned to the requests based on their capability, availability, and feasibility. Therefore, it is an exigent issue to map the VM requests with the hosts to maximize the allocation rate with respect to their priority. One of the recent solutions is to assign the high-priority VM requests in close proximity, and other requests can be assigned using first fit decreasing (FFD). However, this solution does not partition the hosts to discriminate the type of requests. This paper introduces a priority-aware resource allocation (PARA) algorithm to consider three types of VM requests: highly critical, critical, and normal, and partition the hosts into three levels. PARA enables the high-priority VM requests to a relatively large pool of hosts compared to low-priority requests. The proposed algorithm is simulated by considering 10000 to 100000 VM requests and 20 to 30 hosts and compared with a non-partitioned pool of hosts with FFD to show effectiveness in allocation rate and weighted score concerning the serving requests.