This is a Custom Template Java Struts 2 web project and the files uploaded here are the eclipse project code. As said we have used Struts 2 framework for the implementation of this web project. Struts 2 is a java MVC framework for enterprise web applications.

For the client side i.e. user interface has been developed using JSP and CSS Bootstrap has been used for the look and feel of the application. Client side validations are implemented using java script and server side validations is java code written in the presentation layer of the project.

User uses the GUI for requirement specification and up on submission of user requirements the data is transfered to the business layer which has the logic for comparing the user requirements with available cloud resources from multiple vendors and resources are recommended for the users.

Like any other enterprise web application our's is multi tier web application with the following layers/tiers

a) Client/Presentation GUI Layer. In struts 2, each JSP can have a backend java class which controls the presentation of data on page. Java Script is used for client validations and java backend class handles the server validations.

b) Business Logic Layer. This is the core of the project which has the logic to handle the data. The user inputs are compared with the RSpace and eligible resources are identified in each component i.e network, storage and compute. These are called MacOps and then solution with all possible combinations of the MacOps from each component is constructed and are inserted into the Catalog(repository of the templates). Here, before insertion, we check whether the template is already present in the catalog are not and only unique ones are inserted. Then, we use KNN algorithm to get the 3 closest templates to the user requirements from the catalog and present them for the user. KNN uses the Euclidean distance to check for the closeness of the templates in the catalog and the user requirement. We also use weights derived from user preferences using AHP while calculating the distance. Up on user selecting a template the deployment manager will deploy the resources on GENI.

c) Data Layer. This layer is responsible for getting data from the MySQL tables and present it to the business layer. This layer handles all the database connections and other stuff.

RSpace is the resource pool which is basically all the available cloud resources from different cloud service providers abstracted and stored in MySQL. Information about the resources like their hardware configurations, billing and pricing etc.. are stored in MySQL tables. The folder "SQL Resources" under the root folder of the project has the sample RSpace data that we have used for the project. The folder contains the CSV file of the MySQL tables that we used for storing the RSpace. There are three CSV file one each for network, storage and compute. There is also a Database schema generation file.

GENI is public available distributed testbed and experiments are done here to mimic the multiple cloud service providers. Different aggregate managers are treated as different data center locations and also at each data center locations we assume that there are resources available from multiple cloud service providers. The project folder has the "geni_deployment.y" python script which is invoked after user selecting a template for the deployment. This python script will reserve the resource in GENI.