الفهرس 
Chapter 1 IntroductionOnly 14 pages are availabe for public view
 |  | from | 94 |  |  |
| | | from | 94 | | |
Abstract
Cloud computing is the most recent technology, which mainly based on pay-as-you-go principle. The core of the cloud computing is the virtualization, which allows multi-instance of the resources, Virtual Machine (VM), to be created on the same physical machine (Server). VMs are worked in isolated manner, where VMs are running on different servers. Now a day, most of big companies move toward cloud computing due to its flexibility and scalability features. It reacts faster to the needs of business and the resources can be scale up/down according to the load beaks.
Two important issues arise in cloud computing; task scheduling and resource provisioning. Task scheduling is the allocation of the requested tasks to the available resources (VMs) aiming to reduce the overall time to finish the tasks. Resource provisioning is the allocation of a cloud service provider’s resources to a customer. from the practical point of view, it means creating of VMs and allocating virtual resources instances on the physical resources (Servers) aiming to increase the resource utilization and increase the revenue coming from the resources.
The main objective of this thesis is to introduce and implement task scheduling algorithms in the cloud environment. There are two types of task scheduling in cloud computing; static task scheduling, and dynamic task scheduling. Static scheduling requires information and pipelining different stages of task execution. Static scheduling imposes less runtime overhead. In case of dynamic scheduling, information of the job components/task is not known beforehand.
In this thesis, the first contribution addresses the static task scheduling. The proposed algorithm is based on calculating the total processing power of the available resources (i.e., VMs) and the total requested processing power by the users’ tasks. Then, a power factor of each VM (the ratio of its processing power to the total processing power of all VMs) is calculated. Finally, a task or a group of tasks allocated to VM if they have processing power near to the power factor of the VM. The second contribution in this thesis, addresses the dynamic task scheduling, The proposed algorithm is based on measuring the availability level of the resources, then choice the best resource (VM) to the incoming task (measured by its processing power, price of execution, and the number of running tasks on it).
The proposed algorithms are implemented using java programming language and evaluated by using CloudSim simulator.