|
Document Type |
: |
Thesis |
|
Document Title |
: |
Toward Efficient Cloud Services: An Energy-Aware Hybrid Load Balancing Approach نحو خدمات سحابية تتسم بالكفاءة: نهج موازنة التحميل الهجين الموفر للطاقة |
|
Subject |
: |
Faculty of Computing and Information Technology |
|
Document Language |
: |
Arabic |
|
Abstract |
: |
In recent years, there has been keen research interest in cloud-based load balancing and task scheduling. Load balancing of tasks on virtual machines (VMs) using nature-inspired algorithms has become a particular area of focus. Load balancing between VMs is important in order to avoid overloading and underloading of VMs, which causes system failure due to issues such as high power consumption, increased execution time, and increased response time.
Previous research has presented a load-balancing algorithm based on honeybee behavior (LBA-HB) to decrease processing time and overall response time. It selects a host based only on processing time, although other factors affect its performance, including the load and capacity of the host. The VM used to accept the task is reliant on a single element (the number of tasks handled by the VMs), although other crucial factors should be considered in the context of load balancing, such as power consumption and cost. In addition, no previous algorithm has taken into account the integration of quality of service (QoS) factors and power consumption in determining the host and the appropriate VM to receive the incoming task.
This research proposes an improved version of LBA-HB (ILBA-HB) that considers various QoS parameters, including processing time, load, and capacity, and estimates the power consumption of VMs on a host to enhance load balancing. Clarifying iv the correlation between the VM’s power consumption and QoS factors (cost and processing time) was an important aim of this research. When a given situation is uncertain, fuzzy logic can be applied at various levels of input to achieve the desired tradeoff between power consumption and QoS.
CloudSim is used to simulate the ILBA-HB algorithm. The performance of ILBA-HB is compared to the LBA-HB and honeybee-behavior–based load balancing (HBB-LB) algorithms in terms of makespan (M), average response time (AVG -RT), and degree of imbalance (DI). The results show that ILBA-HB improved AVG-RT by 3.90% over LBA-HB and 52.44% over HBB-LB; M by 3.55% over LBA-HB and 78.30% over HBB-LB; and DI by 49.29% over HBB-LB. ILBA-HB did not improve on LBA-HB’s DI performance, but it nevertheless generated promising results in terms of DI. |
|
Supervisor |
: |
Dr. Sanaa Abdullah Sharaf |
|
Thesis Type |
: |
Master Thesis |
|
Publishing Year |
: |
1443 AH
2022 AD |
|
Added Date |
: |
Wednesday, January 25, 2023 |
|
Researchers
| هند سالم العطوي | Alatawi, Hind Salem | Researcher | Master | |
|