Scalability through Application Delivery Cont...

Scalability through Application Delivery Controllers(ADC) for Load balancing in Cloud Computing

Author Name : Aruna Mailavaram

ABSTRACT

One of the major issues in cloud computing is Load balancing[1]. Efficiently, managing the resources in cloud is done by proper utilization of virtual machines and optimal load balancing techniques. Performance of the cloud mainly depends upon a task scheduling and load balancing. Many load balancing approaches in cloud computing has been proposed by researchers to optimize different load performance parameters. The present load balancing algorithms support efficiently allocating the request and also ensuring the resources in an optimized way so that under loaded and overloaded situations will be balanced. Many proposed algorithms will try to decrease makespan time and enhance the utilization of resources. Load balancing software or algorithms were embedded with the operating system, all the incoming task when approach the operating system the load balancing algorithms used to distribute the task among all the servers efficiently. But many algorithms fail to achieve all the performance metrics of load balancing so instead of embedding the algorithms with the OS we can use a hardware called ADC[4][5] for every data center which receives the incoming request and distributes to the servers by checking the health of each server. If there is no proper response from the links, it means the server is down and ADC will not send a request to that particular server and directs that to a healthy server. So, in order to deliver applications to millions of servers around the world, we need an application delivery controller. Application Delivery Controller(ADC) is a server load balancer with advanced features layered on top of that core.   

Keywords: Load balancing, Virtual machines[3], makespan time[2], cloud computing, task scheduling, Application delivery controller.