Investigations ontask scheduling algorithms in cloud computing

Abstract

Cloud computing has been in use for some time now. Differentservices such as Software as a Service, Platform as a Service, Infrastructure asa Service and Business Process as a Service are being offered by variousservice providers. Using cloud services, users can have access to informationat anytime from anywhere using any device. Due to its popularity manycustomers are switching on to the storage of information in cloud.Such proliferation of cloud has led to some technical issues.Balancing the load across the cloud is one of the major challenges in cloud.Normally service providers create and use Virtual Machines [VM] in theavailable servers, in order torespond to the requests from the users. As thenumber of requests from users keeps increasing, the VMs in servers located atvarious geographical locations are allocated to different users for keeping allthe VMs occupied to the extent possible. However, there is always a situationwhere the distribution of the total load across the cloud is not even. Theamount of information transferred in the network increases the cost. The onusis on balancing the load and ensuring other parameters such as response time,throughput and migration time.This problem of load balancing ensuring the other parameters beingwithin acceptable limits is gaining attention in recent time. Many algorithmshave been proposed by many researchers. Some of the algorithms areheuristic and some others are nature inspired. Some of the popular algorithmsinclude First Come First Served (FCFS), Min-Min, Max-Min, MinimumCompletion Time (MCT), Minimum Execution Time (MET), OpportunisticLoad Balancing (OLB), Equally Spread Current Execution Time (ESCE). newline