scholarly journals PEMILIHAN DAN MIGRASI VM MENGGUNAKAN MCDM UNTUK PENINGKATAN KINERJA LAYANAN PADA CLOUD COMPUTING

Author(s):  
Abdullah Fadil ◽  
Waskitho Wibisono

Komputasi awan atau cloud computing merupakan lingkungan yang heterogen dan terdistribusi, tersusun atas gugusan jaringan server dengan berbagai kapasitas sumber daya komputasi yang berbeda-beda guna menopang model layanan yang ada di atasnya. Virtual machine (VM) dijadikan sebagai representasi dari ketersediaan sumber daya komputasi dinamis yang dapat dialokasikan dan direalokasikan sesuai dengan permintaan. Mekanisme live migration VM di antara server fisik yang terdapat di dalam data center cloud digunakan untuk mencapai konsolidasi dan memaksimalkan utilisasi VM. Pada prosedur konsoidasi vm, pemilihan dan penempatan VM sering kali menggunakan kriteria tunggal dan statis. Dalam penelitian ini diusulkan pemilihan dan penempatan VM menggunakan multi-criteria decision making (MCDM) pada prosedur konsolidasi VM dinamis di lingkungan cloud data center guna meningkatkan layanan cloud computing. Pendekatan praktis digunakan dalam mengembangkan lingkungan cloud computing berbasis OpenStack Cloud dengan mengintegrasikan VM selection dan VM Placement pada prosedur konsolidasi VM menggunakan OpenStack-Neat. Hasil penelitian menunjukkan bahwa metode pemilihan dan penempatan VM melalui live migration mampu menggantikan kerugian yang disebabkan oleh down-times sebesar 11,994 detik dari waktu responnya. Peningkatan response times terjadi sebesar 6 ms ketika terjadi proses live migration VM dari host asal ke host tujuan. Response times rata-rata setiap vm yang tersebar pada compute node setelah terjadi proses live migration sebesar 67 ms yang menunjukkan keseimbangan beban pada sistem cloud computing.

Author(s):  
Arif Ullah ◽  
Nazri Mohd Nawi ◽  
Hairulnizam Bin Mahdin ◽  
Samad Baseer ◽  
Mustafa Mat Deris

In modern data centres of cloud computing contains virtualization system. In order to improve network stability, energy efficiency, and makespan proper virtualization need. The virtual machine is one of the examples of virtualizations. Cloud computing data centres consist of millions of virtual machine to manage load balancing. In this study check the different number of virtual machine role in data centres, for that purpose, we established a network with the help of cloudsim and compare different data centres at each zones taking a different number of the virtual machine with different paramater and network banwith.After the simulation the result shows that increasning in the number of VM can affect the netwok accuracy in term of energy ,processing time ,coast and network stabality . 


Author(s):  
Sourav Kanti Addya ◽  
Bibhudutta Sahoo ◽  
Ashok Kumar Turuk

The data center is the physical infrastructure layer in cloud architecture. To run a large data center requires a huge amount of power. A proper strategy can minimize the number of servers used. Minimization of active servers caused minimization of power consumption. But the maximum number of virtual machine placement will be a monetary benefit for cloud service providers. To earn maximum revenue, the CSP is to maximize resource utilization. VM placement is one of the major issues to achieve minimum power consumption as well as to earn maximum revenue by CSP. In this research chapter, we have formulated an optimization problem for initial VM placement in the data center. An iterative heuristic using simulated annealing has been used for VM placement problem. The proposed heuristic has been analysis to be scalable and the coding scheme shows that the proposed technique is outperforming traditional FFD on bin packing technique.


2016 ◽  
pp. 783-808
Author(s):  
Sourav Kanti Addya ◽  
Bibhudatta Sahoo ◽  
Ashok Kumar Turuk

The data center is the physical infrastructure layer in cloud architecture. To run a large data center requires a huge amount of power. A proper strategy can minimize the number of servers used. Minimization of active servers caused minimization of power consumption. But the maximum number of virtual machine placement will be a monetary benefit for cloud service providers. To earn maximum revenue, the CSP is to maximize resource utilization. VM placement is one of the major issues to achieve minimum power consumption as well as to earn maximum revenue by CSP. In this research chapter, we have formulated an optimization problem for initial VM placement in the data center. An iterative heuristic using simulated annealing has been used for VM placement problem. The proposed heuristic has been analysis to be scalable and the coding scheme shows that the proposed technique is outperforming traditional FFD on bin packing technique.


2018 ◽  
Vol 12 (4) ◽  
pp. 3509-3518
Author(s):  
Sourav Kanti Addya ◽  
Ashok Kumar Turuk ◽  
Bibhudatta Sahoo ◽  
Anurag Satpathy ◽  
Mahasweta Sarkar

Sign in / Sign up

Export Citation Format

Share Document