A Harris Hawk Optimisation system for energy and resource efficient virtual machine placement in cloud data centers

PLOS ONE RESEARCH ARTICLE A Harris Hawk Optimisation system for energy and resource efficient virtual machine placement in cloud data centers Madhusudhan H. S. ID1, Satish Kumar T.2, Punit Gupta ID3*, Gavin McArdle3 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: H. S. M, T. SK, Gupta P, McArdle G (2023) A Harris Hawk Optimisation system for energy and resource efficient virtual machine placement in cloud data centers. PLoS ONE 18(8): e0289156. https://doi.org/10.1371/journal. pone.0289156 Editor: Ali Safaa Sadiq, Nottingham Trent University School of Science and Technology, UNITED KINGDOM Received: November 17, 2022 Accepted: July 12, 2023 Published: August 11, 2023 Copyright: © 2023 H. S et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: https://www.cs.huji. ac.il/labs/parallel/workload/ The complete simulation data and sample of dataset is shared in the supporting information section. Results: Energy Consumption (kwh) HHO PSO PABFD W1 101.99 110.26 128.94 W2 117.3 126.42 184.28 W3 118.73 130.32 213.4 W4 114.08 129.56 145.6 W5 105.75 117.26 121.28 CPU Utilization HHO PSO PABFD W1 74.13 71.45 62.23 W2 75.35 73.28 65.3 W3 77.25 74.21 66.78 W4 73.09 73.54 64.78 W5 74.61 72.87 64.21 VM Migrations HHO 1 Department of Computer Science & Engineering, Vidyavardhaka College of Engineering, Mysuru, Karnataka, India, 2 Department of Computer Science & Engineering, BMS Institute of Technology & Management, Bengaluru, Karnataka, India, 3 School of Computer Science, University College Dublin, Dublin, Ireland * Abstract Virtualisation is a major technology in cloud computing for optimising the cloud data centre’s power usage. In the current scenario, most of the services are migrated to the cloud, putting more load on the cloud data centres. As a result, the data center’s size expands resulting in increased energy usage. To address this problem, a resource allocation optimisation method that is both efficient and effective is necessary. The optimal utilisation of cloud infrastructure and optimisation algorithms plays a vital role. The cloud resources rely on the allocation policy of the virtual machine on cloud resources. A virtual machine placement technique, based on the Harris Hawk Optimisation (HHO) model for the cloud data centre is presented in this paper. The proposed HHO model aims to find the best place for virtual machines on suitable hosts with the least load and power consumption. PlanetLab’s realtime workload traces are used for performance evaluation with existing PSO (Particle Swarm Optimisation) and PABFD (Best Fit Decreasing). The performance evaluation of the proposed method is done using power consumption, SLA, CPU utilisation, RAM utilisation, Execution time (ms) and the number of VM migrations. The performance evaluation is done using two simulation scenarios with scaling workload in scenario 1 and increasing resources for the virtual machine to study the performance in underloaded and overloaded conditions. Experimental results show that the proposed HHO algorithm improved execution time(ms) by 4%, had a 27% reduction in power consumption, a 16% reduction in SLA violation and an increase in resource utilisation by 17%. The HHO algorithm is also effective in handling dynamic and uncertain environments, making it suitable for real-world cloud infrastructures. Introduction Cloud computing is a paradigm for providing on-demand computational services and resources through the internet, such as data storage and computing power [1]. Cloud computing offers customers on-demand resources in the form of virtual machines (VMs) and PLOS ONE | https://doi.org/10.1371/journal.pone.0289156 August 11, 2023 1 / 27 PLOS ONE PSO PABFD W1 18590 21457 22100 W2 24109 25632 26234 W3 25696 26543 28256 W4 24427 25387 26678 W5 24145 25286 25875 SLA violations HHO PSO PABFD W1 14.58 15.02 17.3 W2 15.24 17.21 18.25 W3 13.71 16.54 18.43 W4 15.03 16.85 17.76 W5 16.07 17.65 17.42 Sample dataset:; MaxJobs: 76872; MaxRecords: 76872; Preemption: No; UnixStartTime: 788722174; TimeZone: 0; TimeZoneString: US/Pacific; StartTime: Thu Dec 29 09:29:34 PST 1994; EndTime: Sat Dec 30 23:54:09 PST 1995; MaxNodes: 400 (48 interactive, 352 batch, 6 service, 10 I/O); MaxProcs: 400; Note: service and I/O partitions are not used to run jobs; MaxQueues: 37; Job Number – a counter field, starting from 1.; Submit Time – in seconds. The earliest time the log refers to is zero, and is usually the submittal time of the first job. The lines in the log are sorted by ascending submittal times. It makes sense for jobs to also be numbered in this order.; Wait Time – in seconds. The difference between the job’s submit time and the time at which it actually began to run. Naturally, this is only relevant to real logs, not to models.; Run Time – in seconds. The wall clock time the job was running (end time minus start time).; Number of Allocated Processors – an integer. In most cases this is also the number of processors the job uses; if the job does not use all of them, we typically don’t know about it.; Average CPU Time Used – both user and system, in seconds. This is the average over all processors of the CPU time used, and may therefore be smaller than the wall clock runtime. If a log contains the total CPU time used by all the processors, it is divided by the number of allocated processors to derive the average.; Used Memory – in kilobytes. This is again the average per processor.; User ID – a natural number, between one and the number of different users.; Group ID – a natural number, between one and the number of different groups. Some systems control resource usage by groups rather than by individual users.; Executable (Application) Number – a natural number, between one and the number of different applications appearing in the workload. in some logs, this might represent a script file used to run jobs rather than the executable directly; this should be noted in a header comment.; Queue Number – a natural number, between one and the number of different queues in the system. The nature of the system’s queues should be explained in a header comment. This field is where batch and interactive jobs should be differentiated: we suggest the convention of denoting interactive jobs by 0.; Partition Number – a natural number, between one and the number of different partitions in the systems. The nature of the system’s partitions Hawk Optimization-VMP accomplishes their tasks while meeting Quality of Service (QoS) requirements. Each VM is designed to target a certain computing resource capability (e.g., the number of CPUs, I/O bandwidth and memory capacity). Using a physical machine (PM) or host to run several VMs, Virtualisation technology increases a data centr (...truncated)