IUP Publications Online
Home About IUP Magazines Journals Books Archives
     
Recommend    |    Subscriber Services    |    Feedback    |     Subscribe Online
 
The IUP Journal of Information Technology
A Dynamic Load Balancing and Task Scheduling Strategy in Cloud Computing†
:
:
:
:
:
:
:
:
:09
 
 
 
 
 
 
 

Instead of upfront investment in infrastructure, many organizations have moved towards cloud computing to fulfill their computational resource needs, as cloud computing model provides on-demand, scalable resources on pay-as-you-go basis. In cloud computing, it is not possible to predict the volume of user requests in the future, so cloud computing models must be capable of handling thousands of concurrent user requests. An efficient load balancing technique plays an important role in distributing the workload evenly among the servers, so that no server is overloaded or underloaded. The paper proposes a load balancing and task scheduling algorithm, Dynamic VM Load Balancer (DVLB). While making a load balancing decision, DVLB takes both the server processing power and request (task) processing requirement into consideration. The algorithm is implemented using cloudsim 3.0.2 tool, and its performance is compared with two existing load balancing methods, Round Robin and Equally Spread Current Execution (ESCE) algorithm.

 
 

Cloud computing provides an environment where Web Services can realize their initially promised potential (Randles, 2010). In this era of supercomputers, when researchers aim to build more and more powerful systems with maximized resource utilization and minimized energy consumption, still most of the computing resources remain unused. A case study by IBM in 2000 states that average utilization of resources is 5-10% for PCs and 30-35% for servers. These observations are true even today (Buyya et al., 2009).

Cloud computing is a network-based service consisting of a network of commodity computers placed at the same or different geographical locations to serve the needs of a number of customers with different workload on demand by using virtualization technique. Virtualization allows to share a single physical instance of a resource among multiple customers and organizations. Virtualization hides the physical characteristics of resources of computing from end users. The gang scheduling algorithm manages to make all the processes of a job progress together so that one process will not be in sleep state when another process needs to communicate with it (Liu et al., 2013).

 
 

Information Technology Journal, Dynamic VM Load Balancer (DVLB ), Equally Spread Current Execution (ESCE), Round Robin (RR), Makespan, Cloudsim, Virtualization, Dynamic task scheduler, Service model, Deployment model in cloud computing.