Automatic Migration and Deployment of Cloud Services for Healthcare Application Development in FIWARE Stelios Sotiriadis 1 , Lenos Vakanas 1 , Euripides Petrakis 1 , Paolo Zampognaro 2 , Nik Bessis 3 1 Technical Univeristy of Crete, Kounoupidiana Campus 73100, Crete, Chania 1 {s.sotiriadis, petrakis}@intelligence.tuc.gr, lenosvakanas@gmail.com 2 Engineering Ingegneria Informatica S.P.A, Napoli, Italy 3 paolo.zampognaro@eng.it 3 Edge Hill Univerity, Ormskirk, UK Nik.Bessis@edgehill.ac.uk Abstract— Over the latest years due to the emergency of new requirements in the area of healthcare domain, the software to data cloud model attracted significant attention by bridging the gap between cloud service deployment and sensitive data storage and manipulation. In particular, in FI-STAR FP7 we develop innovative cloud services for the healthcare domain using the FIWARE platform. This offers new opportunities for building innovative applications that utilize FIWARE Generic Enablers as the building blocks of Future Internet (FI) applications. In this work we focus on (a) the automatic migration requirements in FIWARE systems and (b) on the automatic shipment and deployment of FI-STAR services among service providers infrastructures that is the FIWARE cloud nodes and service consumer infrastructures (FI-STAR use cases). We describe the design and implementation enabling the automatic shipment and deployment to support the software to data paradigm and to allow marketplace development of the FI-STAR catalogue. The deployment automation solution offers less error-prone and highly repeatable processes, efficient deployment and management of updates and new releases. Keywords— Cloud Computing, Automatic service deployment, Cloud software to data, FIWARE, FI-STAR I. INTRODUCTION FI-STAR FP7 project 1 utilizes cloud computing and Internet of Things (IoT) concept to build innovative Future Internet services [1] for healthcare provision in the information and communication technology domain. FI-STAR experimentation sites and use case scenarios involve seven healthcare providing sites, servicing around 6 Million people throughout Europe. This encompasses various use cases including monitoring, rehabilitation and other of patients utilizing cloud services 2 . In general, cloud computing has been utilized with success in various areas such as industry, agriculture, energy and environment yet not in the healthcare area. Data that are stored in cloud systems and related with healthcare information are usually confidential and restricted for open access (e.g. ISO-25010). Since there are various 1 http://www.fi-star.eu 2 https://www.fi-star.eu/use-cases.html standards, regulations and recommendations described in [2], there are severe restrictions to data transfer and storage. To overcome this hurdle, FI-STAR develops cloud services based on the software to data [8] solution and according to an automatize shipment and deployment model. Specifically, based on [6] we aim to create a framework that allows GEs to be delivered to different physical locations. That is to say that the services are encapsulated as cloud Virtual Machines (VMs) in OpenStack environments [3] and [5], that could be transferred from a public cloud service provider to a private cloud system seamless and without any manual intervention by the user. The FI-STAR platform is developed following the FI-WARE conceptual model and offers a set of software modules, called as Specific Enablers (SEs) that allow flexible developments of healthcare applications. The proposed solution allows SEs to be transferred near to the data source and to be executed and managed automatically. The solution is developed based on a RESTFul architecture, wherein an agent allows two-fold communication and is responsible for updating and versioning features. Based on this discussion next we present the FI-STAR automatic deployment model. In this work we focus on (a) the automatic migration requirements in FIWARE systems and we present the VM migration tool among OpenStack cloud platforms and (b) on the automatic shipment and deployment of FI-STAR services among service providers infrastructures that is the FIWARE cloud nodes and service consumer infrastructures (FI-STAR use cases). The work is organized as follows. Section II present the requirements for software to data approaches, Section III the software to data service functionality and the Section IV FI-STAR automatic deployment process. II. REQUIREMENTS FOR SOFTWARE TO DATA APPROACHES A vital concept of the FI-STAR project is that aims to eliminate restrictions, which are responsible for the minimal usage of cloud technology by various application fields especially health care. Software to data refers to the idea of transferring the software to the data location, thus near to the source of data generation. In order to accomplish this goal,