PerficientCloudSim: a tool to simulate large-scale computation in heterogeneous clouds
- PDF / 3,026,550 Bytes
- 55 Pages / 439.37 x 666.142 pts Page_size
- 22 Downloads / 185 Views
PerficientCloudSim: a tool to simulate large‑scale computation in heterogeneous clouds Muhammad Zakarya1,2 · Lee Gillam1 · Ayaz Ali Khan2 · Izaz Ur Rahman2 Accepted: 31 August 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract The major reason for using a simulator, instead of a real test-bed, is to enable repeatable evaluation of large-scale cloud systems. CloudSim, the most widely used simulator, enables users to implement resource provisioning, and management policies. However, CloudSim does not provide support for: (i) interactive online services; (ii) platform heterogeneities; (iii) virtual machine migration modelling; and (iv) other essential models to abstract a real datacenter. This paper describes modifications needed in the classical CloudSim to support realistic experimentations that closely match experimental outcomes in a real system. We extend, and partially re-factor CloudSim to “PerficientCloudSim” in order to provide support for large-scale computation over heterogeneous resources. In the classical CloudSim, we add several classes for workload performance variations due to: (a) CPU heterogeneities; (b) resource contention; and (c) service migration. Through plausible assumptions, our empirical evaluation, using real workload traces from Google and Microsoft Azure clusters, demonstrates that “PerficientCloudSim” can reasonably simulate largescale heterogeneous datacenters in respect of resource allocation and migration policies, resource contention, and platform heterogeneities. We discuss statistical methods to measure the accuracy of the simulated outcomes. Keywords Clouds · Datacenters · Simulations · Modelling · Performance · Heterogeneity * Muhammad Zakarya [email protected] * Lee Gillam [email protected] Ayaz Ali Khan [email protected] Izaz Ur Rahman [email protected] 1
Department of Computer Science, University of Surrey, Guildford, UK
2
Department of Computer Science, Abdul Wali Khan University, Mardan, Pakistan
13
Vol.:(0123456789)
M. Zakarya et al.
1 Introduction Quantifying the performance of resource provisioning policies in a real cloud platform for different workload models under transient conditions is challenging due to, at least, three reasons: (i) clouds exhibit varying demands, system sizes and hardware resources; (ii) cloud users have heterogeneous and competing QoS requirements; and (iii) workloads have varying performance needs [1]. Furthermore, the use of real IaaS (Infrastructure as a Service) clouds to benchmark the workload performance and infrastructure energy consumption under these variable conditions is constrained by real test-bed availability. Consequently, it is difficult to reproduce verified results and findings that can be trusted. In addition, it would be time-consuming and, therefore, costly (in terms of efforts) to re-configure benchmarking parameters across a large-scale real IaaS cloud for multiple runs and experimentation. Therefore, in large-scale real IaaS clouds, it is not reasonable to condu
Data Loading...
