Scalable Service Infrastructure
This Work Package aims at studying an infrastructure that brings GENESIS services and its corresponding Service Oriented Architecture (SOA) to a reasonable scale, based on design principles that provide required interoperability, openness and Quality of Service (QoS). The term scalable means in this context the possibility for the middleware (infrastructure) to take in charge an increase of the number of services to be managed.
The architecture framework has been thought in order to enable this scalability, allowing service infrastructure to take into account real usage of services (resisting to the load pressure). In particular, an Enterprise Service Bus (ESB) facing SOA issues and events management, that is to say, an event-driven based ESB is proposed. This solution provides brokering features for sensor alerts based on Web Services standards (as WS-Notification).
Schema representing the ESB with specific components (BC-XMPP and BC-SOAP for external communication, SE-NOTIFICATION for forward notifications) in the context of the Villarest resulting demo use case.
Achievements:
- WS-Notification library implementation (WS-Topic, WS-BaseNotification, WS-BrokeredNotification)
- Implementation of an EDA solution within the Petals ESB
- Using the WS-Notification library
- With implementation of the SE-NOTIFICATION component that plays the role of broker for alerts
Links and further information
A video was made in order to illustrate the Villarest use case.
The WS-Notification library and the SE-NOTIFICATION component is available on forge.ebmwebsourcing.com.
The WS-Notification is a standard of OASIS:
An explicative notice of the solution used for Villarest use case had been provided to GENESIS WP5000 partners.
The Petals ESB is a JBI based ESB, open source project hosted in OW2 consortium forge.
