The purpose of this document is to describe the components that provide the capabilities of data storage, object registration, stream composition, transformation and filtering in real time, and object actuation in COMPOSE. The architecture chosen looks for scalability and performance as a must, and therefore builds on top of highly scalable distributed components. The data processing infrastructure is accessed through a REST API that isolates the users and devices from the internals of the platform. All components have been released as open source projects within the servIoTicy platform (servioticy.com) that groups all developments performed in WP2 around the topic of data processing. The document is accompanied by a video that describes the use of the servIoTicy stream processing features, and by a fully featured Virtual Appliance that can be built from the servIoTicy portal to get a local deployment of the platform. An online service has been also deployed and is openly accessible to the consortium members and to external users willing to use it under request.
All the servIoTicy sources developed by BSC are located under GitHub in https://github.com/servioticy/servioticy. Basically, the servIoTicy project is composed of different subprojects, some of them related with the API and others related to the implementation of the Data Processing Pipelines. servIoTicy sources as well as all the other components developed by CREATE-NET in the field of Client Libraries are available at the COMPOSE repository: https://github.com/compose-eu
A virtual appliance is now offered as a vagrant instance, and therefore can be built from some description files: the users can just download the vagrant configuration files and run ‘vagrant up’ and a fully featured instance of servioticy is made available after a few minutes. The sources for the Vagrant image are also located in GitHub.
A video summarizing the major steps of the demo has been prepared and made available at http://www.servioticy.com/?page_id=304 as well as through the servIoTicy YouTube channel (http://goo.gl/L56Yds).