An Efficient Service Broker Policy for Cloud Computing Environment
Kunal Kishor, Vivek Thapar · 2014
As almost all the applications are being provided over the internet, the need for computing resources is shifting from the user’s location to the service provider. The concept of services has gained popularity with the widespread use of the term “cloud computing”, which is a new paradigm that has been defined to address user requests on a pay-per-use basis. With the greatest benefit being elasticity in terms of increase or decrease of computing resources like computation power, storage and bandwidth, cloud is providing better computing solutions to the users of its services. But the success of these solutions lies in the use of efficient policies and algorithms that govern the underlying concept of cloud computing. These policies involve service brokerage, load balancing, virtual machine management and service level agreements. The broker is an intermediary between the client and the cloud service provider and hence the role of a broker is quite significant. Service broker policies are used by a service broker of a cloud solution in order to decide the data center to which the requests have to be routed. Testing such policies in different situations is very necessary for the best results. But carrying out this analysis for a large architectural setup in a real environment is not feasible. Therefore, we make use of simulation tools, which provide a modeling environment and provisioning facilities for analyzing the new algorithms in a successful manner. Cloud Analyst is a graphical cloud simulation tool that provides the necessary simulation environment for executing and analyzing various cloud scenarios. It also provides facilities to implement new policies and algorithms. The proposed broker policy makes use of a proportion weight to decide