Tefis Use Case: Travel Ecommerce Experiment- through a Single-access Point to Different Testbed Resources via Web-services

Jérémie Leguay

Abstract:This Testbed use-case is a large scale SOA application for a huge travel eCommerce platform accessible both from websites and Web Services. This experiment demonstrates the capacity to deploy large scale experiment of the application on different testbeds accessible via the TEFIS[1] portal with the deployment of SOA components offering Services. TEFIS portal provides a single access point to different testing and experimental facilities for communities of software and business developers to test, experiment, and collaboratively elaborate knowledge. TEFIS will establish a connector model that makes it possible to interact with testbeds and their resources in a unified manner using Web services. Initially TEFIS platform integrates 7 complementary experimental facilities [2] , including network and software testing facilities, and user oriented living labs. This Testbed use-case uses in particular the TEFIS connectors to PACA Grid, ETICS, and PlanetLab testbeds illustrating Internet-scale service oriented computing and Service orchestration directions of the Future Internet application. Also involvement of the frontend is part of the use-case. For this purpose a wider set of testbed services provided through the TEFIS portal, in particular the IMS testbed for end user view on mobile device and the BOTNIA Living Lab including real end users are needed. The experiment takes place in SOA architecture and is the backend dedicated to support a huge travel eCommerce platform. This application integrates together a lot of independent services dealing with databases, log files and other dedicated tools. All these services are orchestrated in functions of the life of the frontend application, i.e. new features to be published, bug fixes to be built and identification of bugs. They are accessible either from websites or directly as Web Services and are distributed on dozens of servers. Some of the services create dynamically new instances using one or several machines according to the amount of computation expected, in this case we named these services parallel services. One of the aspects of this use case is to manage the distribution and the different layers of services existing in the application. To do so, an Enterprise Service Bus (ESB) implemented using a Scheduler and a Broker is used. This part of the application is in charge of finding the right node to deploy and execute a service. It must ensure that all the requiring tools and applications are available. A second characteristic of the use case is the processing of huge amount of technical and …

Computer Science,Business