ESTIMATION OF LOAD BALANCING ALGORITHM USING CLOUDSIM / OGHOGHO AGBAMAN NAPOLEON; SUPERVISED BY ASSOC. PROF. DR. MARY AGOYI

Yazar: Katkıda bulunan(lar):Dil: İngilizce 2022Tanım: 62 sheets; 31 cm. Includes CDİçerik türü:
  • text
Ortam türü:
  • unmediated
Taşıyıcı türü:
  • volume
Konu(lar): Tez notu: Thesis (MSc) - Cyprus International University. Institute of GraduateStudies and Research Information Technology Department Özet: ABSTRACT Cloud Computing is a set of resources of which users’ requests on cloud services are increasing on a day-to-day basis, and for better performance it should be balanced such that all the requests should have minimum response time. Therefore, Load Balancing is required and it is one of the major Issues in Cloud Computing. An essential task for any cloud provider is task scheduling, wherein tasks are assigned to computing resources within the cloud system by a broker. There is a lot of load balancing algorithm implemented in cloud computing till today. The aim of this thesis is to test the performance of some algorithms, namely Least-connection, Shortest Remaining Time and Shortest Job First, in order to know which of them do better when there are large cloudlets or when there are more VM, as well as when the CPU is higher or lower. We successfully simulated the various algorithm using Cloudsim 3.0.3 and the end-result shows that Least Connection Algorithm perform best when there are less cloudlets with average turn-around time of 7.6 and average waiting time of 3.1 when there are less cloudlets (tasks) in the system while in lower CPU, the result shows that Shortest Remaining Time algorithm perform better with average turn-around time of 5.8 and average waiting time of 1.4. Also, the result shows that Least Connection algorithm perform better than others with average turn-around time of 5.0 and average waiting time of 8.0 when the system has less VM’s and Shortest Remaining Time Algorithm perform best when there are more cloudlets of about 20 with average turn around time of 30.08 and average waiting time of 24.31 and there is no algorithm that is considered to be the best because different algorithm does better than others at different conditions/situations. Key Words: Cloud Computing, Algorithms, Cloudlets, Task-scheduling, Brokers, Data Center, Cloudsim
Materyal türü: Thesis
Mevcut
Materyal türü Geçerli Kütüphane Koleksiyon Yer Numarası Durum Notlar İade tarihi Barkod Materyal Ayırtmaları
Thesis Thesis CIU LIBRARY Tez Koleksiyonu Tez Koleksiyonu YL 2666 N27 2022 (Rafa gözat(Aşağıda açılır)) Kullanılabilir Information Technologies Department T3025
Suppl. CD Suppl. CD CIU LIBRARY Görsel İşitsel YL 2666 N27 2022 (Rafa gözat(Aşağıda açılır)) Kullanılabilir Information Technologies Department CDT3025
Toplam ayırtılanlar: 0

Thesis (MSc) - Cyprus International University. Institute of GraduateStudies and Research Information Technology Department

Includes bibliography (sheets 58-62)

ABSTRACT
Cloud Computing is a set of resources of which users’ requests on cloud services are
increasing on a day-to-day basis, and for better performance it should be balanced such
that all the requests should have minimum response time. Therefore, Load Balancing
is required and it is one of the major Issues in Cloud Computing. An essential task for
any cloud provider is task scheduling, wherein tasks are assigned to computing
resources within the cloud system by a broker. There is a lot of load balancing
algorithm implemented in cloud computing till today. The aim of this thesis is to test
the performance of some algorithms, namely Least-connection, Shortest Remaining
Time and Shortest Job First, in order to know which of them do better when there are
large cloudlets or when there are more VM, as well as when the CPU is higher or
lower. We successfully simulated the various algorithm using Cloudsim 3.0.3 and the
end-result shows that Least Connection Algorithm perform best when there are less
cloudlets with average turn-around time of 7.6 and average waiting time of 3.1 when
there are less cloudlets (tasks) in the system while in lower CPU, the result shows that
Shortest Remaining Time algorithm perform better with average turn-around time of
5.8 and average waiting time of 1.4. Also, the result shows that Least Connection
algorithm perform better than others with average turn-around time of 5.0 and average
waiting time of 8.0 when the system has less VM’s and Shortest Remaining Time
Algorithm perform best when there are more cloudlets of about 20 with average turn around time of 30.08 and average waiting time of 24.31 and there is no algorithm that
is considered to be the best because different algorithm does better than others at
different conditions/situations.
Key Words: Cloud Computing, Algorithms, Cloudlets, Task-scheduling, Brokers,
Data Center, Cloudsim

Araştırmaya Başlarken  
  Sıkça Sorulan Sorular