DEVELOPMENT OF A HIGH PERFORMANCE TOOLKIT FOR MODELING AND SIMULATING CLOUD COMPUTING ENVIRONMENT AND APPLICATION

Abstract

In cloud computing, computing resources and applications need to be simulated so that providers or developers will achieve high network performance and meet user’s application performance requirements with minimum expenditure before implementing it. The fact that none of the current distributed (including grid and network) system simulators offer the environment which can be directly used for modeling cloud computing environments and applications with high performance rate, and maximum resource utilization shows the need for a cloud network simulator that can perform these functionalities. To remedy this challenge this research work presents iCloud: a new simulation framework that allows seamless modeling, simulation, and experimentation of emerging cloud computing infrastructures and application services. The developed system is a Cloudsim-based simulator that models cloud networks with minimum processing time and maximum resource utilization ratio. This research focuses on the resource utilization using optimize execution time algorithm for service broker policy. It also takes into consideration task migration approach for load balancing algorithm used in the execution of tasks. The system was able to model cloud networks and application with high performance metrics. The experimental results showed a better performance with a response time of 28.46 milliseconds when 10000 users accessed an application on the modeled cloud network. Also a response time of 78.27 milliseconds was recorded instead of 124.07 milliseconds when the metrics of CloudAnalyst was inputted into the developed system. A response time of 31.14 milliseconds was recorded instead of 60.73 milliseconds when the metrics of iCanCloud was inputted in to the developed system. Also a response time of 23.84 milliseconds was recorded instead of 45.6 milliseconds when the metrics of CloudSim was inputted into the developed system. The results obtained from the simulation of existing works showed that the developed system performed better in terms of response time, execution time, makespan time and resource utilization ratio compared to CloudAnalyst, iCanCloud and CloudSim.