IJCATR Volume 7 Issue 11

Performance Analysis of Load Balancing and Physical Machines in Cloud Computing Centers with General Service Time

Ashri Shabrina Afrah, Moechammad Sarosa, Sholeh Hadi Pramono
10.7753/IJCATR0711.1001
keywords : Data Center; Cloud Computing; Queuing Theory; Utilization; Virtual Machine; General Service Time

PDF
The number of Virtual Machines in the Cloud Data Center may affect the value of Cloud Data Center’s utilization. When there are too many Virtual Machines dedicated in the Physical Machines, the utilization is tend to be low. As the result, some idle Virtual Machines will consume much electricity [12]. However, the adequate number of Virtual Machines should be used to maintain the data center’s performance, which is indicated by some QoS parameters, such as queue length, system response time, and drop rate. The system model which is observed this research insists of a Load Balancing and some Physical Machines which contain a number of Virtual Machines in each. However, the commonly used Markovian Model of Queuing Theory will be replaced by The General Model, which is more suitable to the characteristics of the Cloud Data Center. This research is aimed to get the optimum number of Virtual Machines in a Cloud Data Center to improve its performance. The simulation results show that 25 Virtual Machines in each Physical Machine are needed to obtain the optimum level of utilization and other indicator parameters.
@artical{a7112018ijcatr07111001,
Title = "Performance Analysis of Load Balancing and Physical Machines in Cloud Computing Centers with General Service Time",
Journal ="International Journal of Computer Applications Technology and Research(IJCATR)",
Volume = "7",
Issue ="11",
Pages ="407 - 411",
Year = "2018",
Authors ="Ashri Shabrina Afrah, Moechammad Sarosa, Sholeh Hadi Pramono"}
  • .