Abstract:
The growth of information technology has led to the increasing need
of computing and storage. Cloud services are the technologies with high
demand, consisting of a set of virtualized resources that serve the users as
per demand and on a pay-per-use basis via the internet. So, it must have the
ability to meet all the users’ requests with high performance and efficiency.
Task scheduling is a very important aspect to improve the overall
performance of the cloud computing. There are various types of scheduling
algorithms for resource utilization. The poor task scheduling resources are
not utilized efficiently. In order to address these weaknesses, this work
proposes a novel approach to map groups of tasks to customized virtual
machine types. Mapping of the tasks is based on task usage patterns like
length, file size and bandwidth. The proposed system uses K-means
clustering algorithm to generate the task clusters .The tasks are grouped base
on their “Tasks Length and Deadline”. After clustering, the individual task
in each cluster is scheduled to appropriate VMs. The proposed system
performs multi-objective task scheduling with an aim in minimizing the
makespan and execution time. The experimental results show that the
proposed method produces better results in terms of execution time,
makespan than the traditional algorithm (FCFS). In this system, CloudSim
simulator version 3.0.3, one of the open source frameworks is used .It is
written in Java programming language.
